T EC H N I Q U ES OF

HISTO- AND CYTOCHEMISTRY

G "/J

TECHNIQUES OF

HISTO- AND CYTOCHEMISTRY

A Manual of Morphological and Quantitative
Micromethods for Inorganic, Organic and
Enzyme Constituents in Biological Materials

By DAVID CLICK, Ph.D.

ASSOCIATE PROFESSOR OF PHYSIOLOGICAL CHEMISTRY
THE MEDICAL SCHOOL, UNIVERSITY OF MINNESOTA

With a Foreword by Robert R. RENSLEY, M. B., D. Sc.

PROFESSOR EMERITUS OF ANATOMY, UNIVERSITY OF
CHICAGO

1949

INTERSCIENCE PUBLISHERS, INC., NEW YORK
INTERSCIENCE PUBLISHERS LTD., LONDON

Copyright, 1949, by
Interscience Publishers, Inc.

All Rights Reserved

THIS BOOK or any PART THEREOF MUST NOT
BE REPRODUCED IN ANY FORM WITHOUT PER-
MISSION OF THE PUBLISHER IN WRITING. THIS
APPLIES SPECIFICALLY TO PHOTOSTATIC AND
MICROFILM REPRODUCTIONS.

PRINTED IN THE UNITED STATES OF AMERICA
BY MACK PRINTING CO., EASTON, PENNSYLVANIA

This Book is Dedicated to

KAJ LINDERSTR0M-LANG and HEINZ HOLTER

of the Carlsberg Laboratory, Copenhagen, in appreciation of
their scientific achievements, which have contributed gener-
ously to the development of histo- and cytochemistry, and in
appreciation of their fine human qualities of integrity, under-
standing, and a truly civilized sense of values and of humor.

^\GAi

X

FOREWORD

To the older biologist, microchemistn.- meant the application of
appropriate reagents to sections of tissue or to intact cells to enable
him to recognize mider the microscope the nature and localization
of compounds in li^-ing substance. To the chemist, on the other hand,
it signifies the application of instruments of great precision and
rigorous methods to the accurate determination of the composition
of extremely small amounts of material. To the modem biologist,
it includes all of those methods which may aid him to delineate in
the exiguous confines of the cell that elusive and mysterious
chemical pattern which is the basis of life. To the extent that it
requires isolation and purification of compounds, microchemistn.* is
but a significant extension of the usual methods of biochemistry-,
but with the discovery- of methods of isolating microscopic and
submicroscopic and even ultramicroscopic components of lining
substance, and the application of physical and chemical methods of
analysis to them, the new microchemistrf promises to become the
most important tool we possess for elucidation of the fundamental
chemical pattern of protoplasm.

In our enthusiasm for these methods, however, we must not forget
that by far the most sensitive instrument for microchemical analysis
is the li\'ing organism itself. The methods of immunology*, for
example, suffice to discriminate between compounds so closely
related that the chemist is at a loss to distinguish between them.
The genes revealed by genetic experiment exceed by an infinite
multiple the meager number of nucleoproteins revealed by bio-
chemical research. The bioassay methods depend on the exquisite
sensiti\'ity of the li^-ing organism to minute changes in its chemical
en\-ironment. These are also microchemical methods.

In this volume, the author has chosen to follow the historical
pattern and to present the methods of microscopical analysis first.
In this field, the difficulty is not so much to find suitable reagents
as to prepare material in a form susceptible to microscopic study.
Botanical material for a long time possessed definite advantages,
since the support aft'orded by cellulose walls, and. in higher plants,

Vlll FOREWORD

by the vascular bundles, enabled the worker to obtain sections with-
out previous preparation. The introduction of the freezing-drying
technique by Gersh removed to some extent this advantage and
enabled the worker to obtain sections of animal material without
previous chemical treatment. This valuable method has been wisely
chosen as the first subject of the first section.

This section is devoted to the methods for recognition of chemical
substances in microscopic preparations. The author has contented
himself with presenting accurately and without prejudice the many
methods so far suggested for the detection of various chemical sub-
stances in tissues. In this field there are three purposes to be served
— namely, recognition, localization, and quantitation. The first of
these may be accomplished equally well by macrochemical methods
and the third has only recently achieved importance through the
introduction of the newer physical methods of measurement. Accord-
ingly, localization becomes the chief function of microchemistry of
this order. On the user rests the responsibility for perceiving and
avoiding the many pitfalls which are inevitable. Gross blunders have
been made in the past and can only be avoided in the future by
meticulous care and high critical ability on the part of the worker.
The belief that in the complex colloidal matrix of protoplasm reac-
tions occur as they do in more simplified systems in vitro is respon-
sible for many mistakes. No assumption as to solubility or insolu-
bility is admissible. Otherwise soluble substances may be firmly
adsorbed to submicroscopic surfaces or incorporated into molecular
aggregates. In those methods which require fixing, imbedding, and
cutting in order to prepare the material for microchemical tests, the
disparity between the mass of material and the volume and variety
of solvents employed makes the data on insolubility equally untrust-
worthy. The worker must examine critically each step of the process
and seek to api)raise its effect on the final result. Reaction time, drift
of highly dispersed reaction products to other locations due to surface
charges, and translocation of reaction owing to differences of ion
mobility, all must be carefully considered.

Microchemical reactions yielding colorless products and requiring
the use of a second reagent to visualize them should be accepted only
with reservation and should be used only as a first approach, subject
to the results of later confirmatory tests. To this category belong,
for example, the molybdate reactions for the detection of phosphate,

FOREWORD IX

Macallum's reaction for potassium, and the popular methods for
the detection of phosphatases. In all of these, the possibility for the
adsorption of molybdic, cobalt, or lead ions, respectively, independ-
ent of the reactions supposed to occur, should be entertained.

Negative results should not be accepted; the limitations imposed
by the microscope as to the thickness of preparations both from the
standpoint of transparency and of dispersion of light make any
negative conclusions inadmissible.

It may be inferred from the preceding remarks that this writer
views microchemical methods of this category with suspicion. This
is not the case. He simply wishes to insist that the worker scrutinize
critically every phase of his technique and consider seriously what
the value of the method may be for recognition, quantitation, and
localization, respectively.

The second, third, and fourth parts of the volume are devoted to
the methods for accurate physical and chemical microanalysis as
applied to biological problems, and to the newer methods for the
mechanical separation of the morphological constituents of proto-
plasm. This writer has already indicated above his belief that in
these methods rests our chief hope of progress in the solution of the
mystery of life.

The road to this goal is a long and laborious one. It is the writer's
hope and belief that travelers along this difficult highway will find
their burdens lightened by the collection into one volume of so many
useful methods of investigation.

R. R. Bensley

PREFACE

"But not by nature is the man of science more critical
or careful than his colleagues are. What gives him an
advantage over them is that when issues rise in his
domain they can be settled with a sureness and
dispatch which elsewhere are unknown; for science
has a priceless touchstone here to seek out truth — the
technique of measurement."

Edmund W. Sinnott in Science and the Education of
Free Men, American Scientist 32: 209 (1944).

Like a lens gathering diverse rays and concentrating them in a
new beam which penetrates depths hitherto unilluminated, each
new border science reveals a new threshold of knowledge. We who
work in the life sciences stand at such a threshold today. The lights
of histology and cytology are being joined with those of chemistry,
and the brilliant new beams converging to a focus probe beyond
the old limitations. The day is past when our vision cannot penetrate
beyond the architecture of the cell. The wealth of knowledge that
has been, and can still be gleaned from purely descriptive microscopic
anatomy is not to be minimized, but under the new illuminations
we can begin to discern the chemical patterns in the cellular archi-
tecture. From this knowledge an understanding of the functions of
the patterns will follow. Thus, from histology and cytology, as we
have known them in the past, the new field of histo- and cyto-

XI

Xll PREFACE

chemistry is arising, and from this new field, a histo- and cyto-
physiology will develop — and so on in the expanding and exciting
quest into the nature of living processes. These new instruments and
techniques which carry our vision deep into the living unit, to the
molecules and the atoms — they include the beautiful ingenuities
which have been refined out of the mountain of past scientific
experience, and some which have been newly created for the purpose.
It is these instruments and techniques with which we shall be
concerned in this book.

In a discussion on cytological technique, J. R. Baker (1942) of
Oxford stated, "It was once remarked to the writer that biochemists
like to have their substances in test-tubes. The cytologist wants to
have his exactly where they were in life, and to know, as precisely
as he can, what they are. When substance and structure are known,
the way is clear for the elucidation of the main problem of cytology,
which is to discover what a cell does to keep alive and to perform
its functions for the body as a whole or for the next generation." It
isn't so much "that biochemists like to have their substances in test-
tubes" as it is that, until not long ago, biochemists had no means of
dealing with substances except in "test-tubes." That biochemists
have been fundamentally dissatisfied with this limitation is apparent
in their growing efforts to refine their techniques to enable investiga-
tions in situ. We can be sure that future campaigns designed to as-
sault the present horizons of cytology will follow the strategic lines
made possible by the development of equipment and procedures
which bring chemical investigations directly to the cell, and the parts
of the cell, existing in natural milieu. In the following pages we shall
examine the techniques and devices already elaborated for this
purpose, and no one with imagination will fail to be impressed and
excited by the possibilities engendered.

Lison's Histochemie Animate, a book dealing mainly with micro-
scopic techniques of chemical morphology, was published in Paris
in 1936. Many notable advances have occurred since then, and the
present volume has been designed to bring together in a compact
and readily available form detailed descriptions, not only of the
morphologic, but also of the quantitative techniques. Insofar as
it has been possible, the material presented has been brought up to
date as of January 1, 1947. Pertinent publications which have ap-
peared between January 1 and September, 1947, have not been

PBEFACE XIU

discussed, but a bibliography appendix containing these references
has been included at the end.

To those who treat with condescension all science that cannot be
quantitated, it might be said, "True, morphology is only a stepping
stone, and while as a stepping stone it is not a place to stop, it is
none the less basic." And to those whose comprehension is confined
to mere morphology, it might be said, "True, morphology is basic,
but it is only a stepping stone in science." And to both it should be
said, "All fields of science are stepping stones, and in science there
are no places to stop."

The author wishes to express his appreciation to the following for
critically reviewing various sections of this volume and offering many
helpful suggestions: Dr. R. R. Bensley, Dept. of Anatomy, University of
Chicago; Dr. W. L. Doyle, Dept. of Anatomy, University of Chicago;
Dr. 0. H. Lowry, Dept. of Pharmacology, Washington University; Dr.
G. H. Scott, Dept. of Anatomy, Wayne University; and Dr. J. M. Tobias,
Dept. of Physiology, University of Chicago. The encouragement in this
undertaking of Dr. M. B. Visscher, Dept. of Physiology, University of
Minnesota; and Dr. B. Sullivan, Director of Laboratories, Russell-Miller
Milling Co., Minneapohs, is also gratefully acknowledged. The author is
grateful for the invaluable assistance of the pubhsher's staff.

David Glick
October, 1948
Minneapolis

CONTENTS

Foreword vu

Preface • xi

Abbreviations xxiv

Microscopic Techniques 1

I. Freezing-Drying Preparation of Tissue 3

Parker-Scott Method for Freezing-Drying Tissues 5

II. Cliemical Methods U

A. Requirements 11

B. Inorganic Elements and Radicals 14

Potassium 14

Gersh Modification of Macallum Method for Potassium 14

Carere-Comes Siena Orange Method for Potassium 15

Calcium 16

Calcium Sulfate Test for Calcium in Plant Tissue 17

Cretin Color Test for Calcium and Other Minerals 17

Von Kossa Silver Test for Calcium 17

Magnesium ■ 18

Broda Method for Magnesium 18

Zinc 18

Mendel and Bradley Method for Zinc 19

Iron 19

Prussian-Blue Test for Ferric Iron and TurnbuU's Blue for

Ferrous Iron 20

Humphrey Dinitrosoresorcinol Test for Iron 21

Thomas and Lavollay Hydroxyquinoline Test for Iron 21

Nickel 22

Cretin and Pouyanne Method for Nickel 22

Lead and Copper 22

Mallorj' and Parker Hematoxylin Method for Lead and Copper 23
•• Mallory and Parker Methylene Blue Method for Lead and

Copper 24

Mercury 24

Method of Hand et al. for Mercurous and Mercuric Mercury . . 25

Silver 25

Okamoto et al. Procedure for Silver 26

Gold 27

Roberts Method for Gold 27

Elftman and Elftman Method for Gold 28

Okamoto et al. Method for Gold ; 28

Platinum 29

Palladium 29

Uranium 29

Arsenic 30

Castle Cupric Salt Method for Arsenic 30

Bismuth 30

Wachstein and Zak Method for Bismuth 31

Castel Method for Bismuth 32

XV

63013

Xvi CONTENTS

Microscopic Techniques (continued)

Chloride and Phosphate-Carbonate 32

Gersh Method for Chloride and Phosphate-Carbonate 33

Iodide 34

Phosphate 34

Serra and Queiroz Lopes Modification of the Molybdate

Method for Phosphate 34

Nitrate 35

Cramer Method for Nitrate 35

Sulfhydryl and Disulfide Groups 36

Nitroprusside Test of Rapkine 36

Bourne Nitroprusside Test 37

Hammett and Chapman Nitroprusside Test 37

Organic Substances and Groups 38

Lipids and Cholesterol 38

Kay and Whitehead Procedure for the Sudan IV Stain
for Lipids 39

Jackson Procedure for Lipids Using Acetic-Carbol-Sudan III . . 39

Telford Govan Technique for Sudan Dye Staining in
Aqueous Media 40

Schultz Cholesterol Test 41

Carotene, Carotenoids, and Vitamin A 42

Steiger Method for Carotene in Leaves 42

Riboflavin 43

Chevremont and Comhaire Method for Riboflavin 43

Polysaccharides in General 43

Method of Hotchkiss for Polysaccharides 44

Acid Polysaccharides— Hyaluronic Acid 45

Hale's Method for Acid Polysaccharides 45

Mucoproteins 46

Lison's Method for Polysaccharide Sulfate Compounds (after

Sylven) 47

Glycogen and Mucin 47

Bauer-Feulgen Stain for Glycogen (after Bensley) 48

Best Carmine Stain for Glycogen (after Bensley) 49

Gomori Procedure for Glycogen and Mucin 50

Starch 51

Milovidov Method for Starch 51

Cellulose 52

Post and Laudermilk (1942) Iodine Stain for Cellulose 52

Chitin 52

Murray Method for Softening Chitin 53

Diaphanol Method for Softening Chitin 53

Methods for Staining Chitin 54

Ascorbic Acid 54

Bourne Silver Stain for Ascorbic Acid 55

Protein Reactions 56

Arginine and Arginine-Containing Proteins 56

Serra Method for Arginine and Arginine-Containing Proteins. . 57

Thomas Method for Arginine and Arginine-Containing Proteins 58
Tryptophane in Proteins 58

Romieu Reaction for Tryptophane in Proteins 59

Tyrosine in Proteins 59

Millon Reaction for Tyrosine in Proteins (after Bensley

and Gersh) 59

a-Amino Acid Groups in Proteins 60

Berg Ninhydrin Test for o-Amino Acid Groups 60

CONTENTS XVll

Microscopic Techniques (continued)

Melanin 61

Dublin Application of the Bodian Method to Demonstration

of Melanin 61

Hemoglobin 62

Ralph Method for Hemoglobin 62

GouUiart Method for Hemoglobin 62

Dunn Method for Hemoglobin 63

Bile Pigments and Acids 63

Stein Test for Bile Pigments 63

Forsgren Test for Bile Acids 64

Aldehydes, Nucleic Acids, and "Plasmal" 65

Coleman Preparation of Feulgen Reagent 67

Whitaker Feulgen Technique for Plant Tissues 67

Cowdry Modification of Feulgen Reaction for Paraffin

Sections of Animal Tissues 68

Oster Modification of Feulgen Reaction for Fresh-Frozen

Sections of Animal Tissues 68

Method of Turchini et al. for Nucleic Acids 69

Water-Insoluble Carbonyl Compounds 69

Bennett Use of Phenylhydrazine Reaction for Water-Insolu-
ble Aldehydes and Ketones 70

Albert and Leblond Use of 2,4-Dinitrophenylhydrazine Re-
action for Water-Insoluble Aldehydes and Ketones 71

Bennett Use of Semicarbazide Reaction for Water-Insoluble

Aldehydes and Ketones 71

Purines - 72

Murexide Test for Certain Purines 72

Hollande Modification of the Courmont-Andre Method for

Uric Acid and Urates 73

Indole and Related Compounds 73

Phenols 74

Lison Modification of the Chromaffin Reaction 74

Urea 75

Sulfonamides 75

Method of MacKee et al. for Sulfonamides 76

D. Enzymes 77

Urease 77

Alkaline Phosphatase 78

Gomori Revised Method for Alkaline Phosphatase 79

Acid Phosphatase 80

Gomori Revised Method for Acid Phosphatase in Animal

Tissues 81

Glick and Fischer Adaptation to Grains and Sprouts of

Gomori Method for Acid Phosphatase 82

Other Phosphatases 85

Zymohexase (Aldolase plus Isomerase) 86

Allen and Bourne Method for Zymohexase 87

Lipase 88

Gomori Revised Method for Lipase 89

Peroxidase 90

McJunkin Method for Peroxidase in Tissue Sections 90

Armitage Method for Peroxidase in Blood or Bone Marrow

Smears 91

Dopa Oxidase 91

Laidlaw Method for Dopa Oxidase 92

Amine Oxidase 93

Oster and Schlossman Method for Amine Oxidase 93

Xviii CONTENTS

Microscopic Techniques (continued)

Cytochrome Oxidase 94

Graff Method for Cytochrome Oxidase in Fixed Tissue

("M. Nadi Oxidase") 94

Graff Method for Cytochrome Oxidase in Fresh Tissue

("G. Nadi Oxidase") 95

Loele Method for "a-Naphthol Oxidase" 96

Succinic Dehydrogenase* 96

Semenoff Method for Succinic Dehydrogenase 96

III. Physical Methods 99

A. Fluorescence Microscopy 99

1. Apparatus 99

2. Preparation of Tissues 102

3. Photomicrography , 102

4. Characterization of Substances 104

Direct Observation of Fluorescence 104

Spectroscopic Analysis of Fluorescence 108

B. Emission Histospectroscopy 109

1. PoHcard Technique 109

2. Scott and Williams Technique 110

C. Visible and Ultraviolet Absorption Histospectroscopy 113

1. Caspersson in Situ Technique 114

2. Norberg Technique 120

Apparatus 120

Manipulations 121

The Sample Slide for Absorption Measurements 122

Accessories for Norberg Technique 123

Methods 124

Phosphorus 124

Norberg Method for Phosphorus 124

3. Stowell Technique 125

D. Roentgen Absorption Histospectroscopy 127

1. Some Theoretical Aspects 128

2. Thickness of Sections 132

3. Apparatus 133

4. Measurement of Density of Photographic Film 138

Nomogram 140

E. Microincineration 140

1. Preparation for Incineration 141

2. The Incineration Furnace 142

3. Scott Incineration Procedure 143

4. Microscopic Examination and Interpretation 144

5. Quantitative Estimation of Ash 146

F. Analytical Electron Microscopy 147

The Scott-Packer Analytical Electron Microscope 148

Manipulation 151

G. Radioautography 152

Preparation of Radioautographs 156

Belanger and Leblond Technique 157

Discussion 160

Chemical Techniques 163

Introduction 165

I. General Apparatus and Manipulation 165

A. Vessels, Stoppers, Holders, etc 165

B. Microliter Pipettes 170

CONTENTS XlX

Chemical Techniques (continued)

C. Filters 176

D. Stirring Devices 178

E. Heating Devices 180

F. Moist Chambers 181

G. Electrodes 183

H. Conductivity Apparatus 188

I. Balances 189

II. Colorimetric Techniques 195

A. Capillary Tube Technique 195

1. Apparatus 196

2. Manipulations 197

3. Methods 198

Preparation of Protein-Free Supernatants 198

Tungstic Acid Supernatants 198

Trichloroacetic Acid Supernatants 199

Zinc Sulfate-Sodium Hydroxide Supernatants 200

Chloride 200

Westfall, Findley, and Richards Method for Chlorides 200

Sodium 203

Bott Method for Sodium 203

Phosphate 208

Walker Method for Phosphate 208

Phosphatase 209

Weil and Russell Method for Phosphatase 209

Reducing Substances 210

Walker and Reisinger Method for Reducing Substances 211

Creatinine 211

Method of Bordley et al. for Creatinine 211

Uric Acid 213

Bordley and Richards Method for Uric Acid 213

Urea 214

Walker and Hudson Method for Urea 214

Hydrogen Ion Concentration 216

B. Cuvette Technique 216

1. Apparatus '. 216

2. Methods 219

Calcium 219

Sendroy Method for Calcium 219

Chloride 224

Sendroy Method for Chloride 225

Phosphate and Phosphatase 226

Lundsteen and Vermehren Method for Inorganic Phosphate

and Phosphatase 227

Lowry and Lopez Method for Inorganic Phosphate in

Presence of Labile Phosphate Esters 228

Bessey, Lowry, and Brock Method for Phosphatase 229

Nitrogen and Ammonia 230

General ; 230

Digestion of Sample for Determination of Total Nitrogen ...234

Le\'y Nesslerization Method for Nitrogen 237

Russell Phenol-Hypochlorite Method for Ammonia 238

Uric Acid, Creatine, Creatinine, and AUantoin 239

Borsook Method for Uric Acid 240

Borsook Method for Creatine and Creatinine 242

Sure and Wilder Method for Creatine and Creatinine 243

Borsook Method for AUantoin 243

XX CONTENTS

Chemical Techniques (continued)

Ascorbic Acid 245

Lowry, Lopez, and Bessey Method for Ascorbic Acid 245

Glycogen 247

Boettiger Method for Glycogen 248

Van Wagtendonk, Simonsen, and Hackett Method for

Glycogen 249

Vitamin A and Carotene 250

Method of Bessey et al. for Vitamin A and Carotene 250

III. Titrimetric Techniques 255

A. Microhter Burettes 255

B. Photometric End Points 265

C. Methods 265

Sodium and Potassium (Combined) 265

Linderstr0m-Lang Method for Sodium and Potassium 266

Potassium 268

Norberg Method for Potassium 268

Sodium 270

Lindner and Kirk Method for Sodium 271

Calcium 272

Siwe lodometric Permanganate Method for Calcium 273

Siwe Acidimetric Method for Calcium 274

Lindner and Kirk Cerimetric Method for Calcium 275

Iron 276

Kirk and Bentley Method for Iron 277

Ramsay Method for Iron 278

Phosphorus 280

Chloride 281

Linderstr0m-Lang, Palmer, and Holter Electrometric

Method for Chloride 282

Nitrogen and Ammonia 283

Brliel, Holter, Linderstr0m-Lang, and Rozits Method for

Nitrogen and Ammonia 283

Urea 286

Kinsey and Robison Method for Urea 286

Urease 287

Amide, Peptide, and Nitrate Nitrogen 288

Borsook and Dubnoff Methods for Amide, Peptide, and

Nitrate Nitrogen 288

Acid, Alkah, Amino, and Carboxyl Groups 290

Lipid 291

Schmidt-Nielsen Method for Lipid 291

Extraction and Fractionation of Lipids 292

Schmidt-Nielsen Method for Extraction and Fractionation

of Lipids 293

Iodine Number of Lipids 294

Schmidt-Nielsen Method for Iodine Number 295

Reducing Sugars 296

Holter and Doyle Modification of Linderstr0m-Lang and
Holter lodometric Method for Reducing Sugars 296

Heck, Brown, and Kirk Cerimetric Method for Reducing

Sugars 297

Glycogen 298

Heatley Method for Total Glycogen 299

Heatley and Lindahl Method for Desmo- and Lyoglycogen 300
Ascorbic Acid 300

GHck Method for Ascorbic Acid 301

CONTENTS XXI

Chemical Techniques (continued)

Amylase 302

Proteolytic Enzymes 302

Linderstr0m-Lang and Holter Acidimetric Acetone Method
for Proteolytic Enzymes 303

Linderstr0m-Lang and Duspiva Alkalimetric Alcohol
Method for Proteolytic Enzymes 304

Weil Formol Method for Tryptic Activity 306

Arginase 306

LinderstrcSm-Lang, Weil, and Holter Methods for Arginase. .307
Esterases and Lipases 308

Glick Acidimetric Method for Esterase and Lipase 309

Glick Acidimetric Method for Cholinesterase 310

Catalase 310

Holter and Doyle Method for Catalase 310

IV. Gasometric Techniques 313

A. Volumetric 313

1. Capillary Respirometry 314

(a) Cunningham-Barth-Kirk Differential Respirometer 314

(b) Cunningham-Kirk Open-Tube Respirometer 319

(c) Tobias-Gerard Respirometer 322

(d) Scholander Micrometer-Burette Differential Respirometers 324

2. Gas Analysis 326

(a) Scholander Micrometer Burette Gas Analyzer 326

(b) Berg Simplified Gas Analyzer 328

(c) Scholander-Roughton Syringe Gas Analyzer 329

Oxygen 331

Roughton and Scholander Method for Oxygen 331

Carbon Monoxide 334

Scholander and Roughton Method for Carbon

Monoxide 334

Nitrogen 335

Edwards, Scholander, and Roughton Method for

Nitrogen 336

Carbon Dioxide 337

Scholander and Roughton Method for Carbon Dioxide. .337

3. Membrane Interferometer Volumetry 340

B. Manometric 342

1. Cartesian Diver Manometry 342

(a) Microliter Diver Technique 343

(b) 0.1 Microliter or Capillary Diver Technique 382

(c) Methods Other Than for Respiration 393

Cholinesterase 393

Lindestr0m-Lang and Glick Method for Cholinesterase . 393

Thiamine and Cocarboxylase 394

Westenbrink Method for Thiamine and Cocarboxylase 395

Diphosphopyridine Nucleotide 396

Anfinsen Method for Diphosphopyridine Nucleotide ...396

2. Optical-Lever Respirometry 399

C. Polarographic 404

Microelectrode Measurement of Local Oxygen Tension in Tissue 404

V. Dilatometric Techniques 413

Dilatometric Apparatus and Its Use 414

Peptidase 417

Method of Linderstr0m-Lang and Lanz for Peptidase 419

Density and "Reduced Weight" 420

XXll - CONTENTS

Chemical Techniques (continued)

VI. Determination of Amount of a Biological Sample 423

A. Preparation of Frozen Tissue Sections of Accurate Thickness 427

Linderstr0m-Lang and Mogensen Method for Accurate Cutting
and Special Handling of Frozen Tissue Sections 428

B. Microscopic Examination and Chemical Analysis of the Same

Tissue Section 430

Method of Anfinsen et al. for Microscopy and Analysis on
the Same Tissue Section 431

C. Volume of Irregularly Shaped, Small Biological Samples 431

Holter Method for Measurement of Volume 432

VII. Deductive Methods 435

Microbiological Techniques 43

Introduction 439

Riboflavin 439

Lowry and Bessey Method for Riboflavin 440

Mechanical Separation of Cellular Components 445

Introduction 447

I. Types of High-Speed Centrifuges 448

II. Separation of Components of A. punctulata Eggs (after Harvey) 452

III. Isolation of Cell Nuclei 454

A. Nuclei from Avian Erythrocytes 454

Laskowski Procedure for Isolation of Nuclei 454

Dounce and Lan Procedure for Isolation of Nuclei 455

B. Nuclei from Other Cells 456

Stoneburg Procedure for Isolation of Nuclei 456

Dounce Procedure for Isolation of Nuclei 456

Lazarow Procedure for Isolation of Liver Nuclei as Used by

Hoerr 458

Behrens Procedure for Isolation of Nuclei from Thymus and

Lymph Cells (as Modified by Gulick et al.) 459

IV. Isolation of Chromatin Threads from Cell Nuclei 461

Claude and Potter Procedure . .v 461

V. Isolation of Cytoplasmic Particulates 462

A. Mitochondria 462

Bensley and Hoerr Procedure for Guinea Pig Liver 462

Claude Procedure for Certain Neoplastic Cells of the Rat 463

Claude Procedure for Isolation of "Large Granules" from Liver 464

B. Submicroscopic Particulates 465

Lazarow Procedure for Separation and Isolation of Lipoprotein

and Glycogen Particles from Guinea Pig Liver 466

Claude Procedure for Isolation of "Microsomes" 466

VI. Isolation of Chloroplasts from Leaf Cells 468

Granick Procedure 469

Neish Procedure 469

VII. Isolation of Other Particulates from Cells 471

Bibliography 473

Bibliography Appendix 504

List of Manufacturers 508

Index 511

ABBREVIATIONS

1. liter

ml. 10~^ liter

lA. 10-6 liter

gal. gallon

g. gram

mg. 10~^ gram

[xg. 10-6 gram

m/xg. 10"*^ gram

lb. pound

oz. ounce

D.C. direct current

R.P.M. revolutions per minute

E.M.F. electromotive force

m. meter

cm. 10-^ meter

mm. 10"^ meter

fi 10-6 meter

mix. 10~^ meter

A 10-^0 meter

( angstrom unit)

X.U. 10"^ angstrom

ft.

foot

in.

inch

yr-

year

hr.

hour

min.

minute

sec.

second

amp.

ampere

/xamp.

10-6 ampere

/.F.

10-6 farad

Kev

10^ electron volts

M

molar

mil/

10-^ molar

N

normal

cone.

concentrated

dil.

dilute

soln.

solution

sp. gr.

specific gravity

C. P.

chemically pure

c.p.

candle power

m.p.

melting point

b.p.

boiling point

All temperatures are given in degrees centigrade.
If not otherwise indicated, all solutions are understood to be aqueous.
If not otherwise indicated, the term alcohol refers to 95% ethyl
alcohol.

xxiv

4

MICROSCOPIC TECHNIQUES

"She (Science) warns me to be careful how I adopt a
view which jumps with my preconceptions, and to
require stronger evidence for such a beUef than for
one to which I was previously hostile. My business
is to teach my aspirations to conform themselves to
fact, not to try and make facts harmonize with my
aspirations. — Sit down before fact as a little child,
be prepared to give up every preconceived notion,
follow humbly wherever and to whatever abysses
nature leads, or you shall learn nothing."

Thomas Huxley
in a letter to Charles Kingsley, September 23, 1860.

The microscopic techniques which are treated in the present vol-
ume are those designed to establish the distribution of elements,
groups, substances or activities in microtome sections of tissue by
means of examinations under some form of microscope. This re-
quires that the factor in question be made apparent by character-
istic optical or photographic properties such as a specific color,
fluorescence, or radiation. With the exception of absorption histo-
spectroscopy, these microscopic techniques are limited to observa-
tions which are essentially qualitative in nature. However, the
microscopic techniques permit a much greater degree of localization
of particular chemical constituents in histologically defined cells, or
cytologically defined parts of cells, than is possible by means of the
quantitative chemical techniques. Thus, one is often forced to
choose between degree of localization and quantitation. Obviously,
it would be preferable to establish both the cellular disposition of
biologically significant factors and their quantitative relationships.

/. FREEZING DRYING PREPARATION

OF TISSUE

Since the microscopic techniques that will be discussed are almost
all based on the use of microtome sections of tissue, it is pertinent
that the freezing-drying preparation for sectioning be described in
detail. This technique of sudden cooling to low temperatures and
rapid dehydration of the frozen material in vacuo has many advan-
tages over the usual histological methods employing fixing and
dehydrating solutions. The chief of these advantages are a minimum
of chemical change in the tissue (there is an almost instantaneous
cessation of metabolic activity and no chance for other chemical
changes to occur), a mimum of shifting of diffusible constituents
(fluid is not used and the fixation is immediate), a greater preserva-
tion of cytoplasmic inclusions than is possible with the use of fixing
solutions, the possibility of direct paraffin infiltration of dehydrated

4 MICROSCOPIC TECHNIQUES

tissue, and the absence of cell shrinkage. These are no inconsider-
able advantages, and the freezing-drying technique should be given
the preference wherever possible.

Scott ( 1943) has been careful to point out that, while distortion
of mineral distribution might be expected to occur as the result of
ice crystal formation during the freezing and that artifacts might
be occasioned by the paraffin infiltration, neither of these appears to
be a serious difficulty in the more recent improved techniques. As
Scott indicated, on the one hand ice crystal formation could be
readily recognized should it occur in a manner that might influence
interpretation, and on the other it has been impossible thus far to
find evidence of distortion resulting from the infiltration, although
various control experiments have been performed to test this
possibility.

Gersh (1932) extended the Altmann method of dehydrating tissue
in vacuo at liquid-air temperatures, and the improved procedure is
known as the Altmann-Gersh technique.* In a critical study Scott
(1933a) pointed out that the dehydration temperature of —20°
used by Gersh was not low enough to prevent a certain amount of
ion diffusion since this temperature is above the eutectic point of
certain naturally occurring salt systems. In the improved cryostat
of Packer and Scott (1942), to be described in detail later, the
temperature is maintained below — 54.9°, the eutectic point of
CaCl2.6H20. When he first indicated the desirability of using lower
temperatures, Scott ( 1933a) recommended the use of alcohol cooled
to — 177°, instead of liquid air, since the latter gives rise to a gas
envelope around the tissue which retards the rate of freezing. A
further improvement was effected by Hoerr ( 1936) , who found that
more rapid freezing (hence smaller ice crystals) was obtained by
placing tissue in isopentane cooled to — 160° to — 195° by means of
liquid nitrogen. Among others, Simpson (1941) confirmed the
advantages of the isopentane method and, in addition, pointed out
the desirability of employing small pieces of tissue for treatment
since the centers of larger pieces do not yield sections of the highest
quality.

After the appearance of the Gersh (1932) vacuum dehydrator,
other types were described by Goodspeed and Uber (1934) and
Scott and Williams (1936). However, since none of these was

* The Gersh apparatus is available from A. S. Aloe & Co.

FREEZING-DRYING PREPARATION OF TISSUE 5

wholly satisfactory, Packer and Scott (1942) developed a cryostat
of a new design that is the finest instrument yet devised for the
freezing-drying of tissues. An important feature of this apparatus
is that the frozen and thoroughly dried tissue can be brought
gradually to the temperature of the melted paraffin, and then it
can be embedded without contact with the moisture of the air.
Previous practice was to transfer the very cold tissue from the
cryostat to the air, and then plunge it directly into melted paraflEin,
thus subjecting it to a sudden temperature change of about 100°.
Sjostrand (1944) described a freezing-drying apparatus somewhat
simpler than the Packer-Scott instrument but it was not designed to
permit paraffin infiltration within the apparatus.

PACKER-SCOTT METHOD FOR FREEZING-
DRYING TISSUES

The Freezing-Drying Apparatus. In the diagram of the ap-
paratus, which is made of Pyrex glass (Fig. 1), the drying chamber
(C) is a 2.5 in. tube 12.5 in. long surrounded by a jacket of about
3.5 in. diameter that can be exhausted through stopcock B connected
by glass tubing to S. The 3 gal. Pyrex Dewar flask (Di), containing
solid carbon dioxide in butyl alcohol is used to cool the drying
chamber, and it is arranged so that it can be easily lowered away
from the apparatus. A commercially built refrigerator has also been
employed in place of solid carbon dioxide for the cooling by Hoerr
and Scott (1944). When a pressure of 1 mm. of mercury, or less,
is maintained in the space E, and paraffin is in tube C, the equi-
librium temperature over the paraffin is about — 66°. As used at
present, there is no occasion to employ temperatures higher than
— 66°, but Packer and Scott point out that a thermocouple sealed
in space E could be used to operate a thermostat which in turn could
control the current in the paraffin heater (D) in order to maintain
temperatures above — 66°. The heater (D) is required to melt the
paraffin in tube C so that the tissues held in the copper gauze basket
(Ci) can be embedded in vacuo.

The heater (D) is constructed by covering a thin-walled copper
cylinder with liquid porcelain (Insa-lute), and, after dry, winding
No. 18 Nichrome wire (about 70 ohms) over it and applying another
layer of liquid porcelain over the wire. A thin-walled sleeve of cop-

6 MICROSCOPIC TECHNIQUES

per is then fitted over the whole and, after testing the unit, elec-
trical connections to the outside are made through tungsten glass
seals. Small glass projections on the outside of the drying tube
near the bottom serve to support the heating unit in its proper
position.

Two glass boats {H) contain the phosphorus pentoxide used as
the drying agent. They are placed in tube G, which is 4 in. in di-
ameter, through the opening at J. The closures at A and J are
grease joints fitted with springs in the usual manner. All connect-
ing tubes on the low pressure side of the apparatus have a diameter
of 1.5 in. The vapor trap {K) has a 1.5 in. inner tube and a 0.5
in. annular space, and the inner tube projects about 5 in. below the
level of the solid carbon dioxide in butyl alcohol used as a re-
frigerant to surround the trap.

The two-stage diffusion pump (L) and the single-stage booster
pump {N) employ Octoil-S {Distillation Products Co.) instead of
mercury since the former has a very low vapor pressure (claimed
to be < 10"*^ mm. of mercury at 15°), effects a very high pumping
speed, and is much cheaper than mercury. The pump (L) has an
intake speed of 10 liters per second ; it was designed by Professor A.
L. Hughes of the Washington University Department of Physics.
The boilers of the vapor pumps were protected from drafts, etc. by
a covering of several layers of wet asbestos paper and over these
a thin aluminum cone (M) . Turned-under tabs of the cone support
a flat circular 300 watt heater ( Chromolox) . A small electric blower
is used to cool the upper chamber of the boiler of pump L for
maximum efficiency. A single-stage Welch mechanical pump is con-
nected by rubber pressure tubing at U to the booster pump {N).
The phosphorus pentoxide trap (0) prevents contamination of the
oil when the vacuum is broken by stopcock P. The stopcock Q
has a 1 cm. bore and is used to isolate the high-vacuum section
from the mechanical pump while the space E is being exhausted.
The two-way stopcock R connects to the air at T and to E through
S.

The type of support employed for the glass apparatus is shown in
Figure 2. The main standard is a rectangular aluminum box 4 ft.
high bolted to a % in. iron base plate 2 ft. square. The aluminum
box is made by welding together at W two heavy 8 in. aluminum
channels (V). The glass is supported by copper rings silver-

FREEZING-DRYING PREPARATION OF TISSUE 7

soldered to Vie in. brass rods which are threaded into the aluminum
standard. An instrument panel is mounted under G to carry the
rheostats and meters for controlling the heating current for the
diffusion pumps and the paraffin chamber.

Detail of
joints A and J

Detail of
paraffin heater D

Fig. 1, The Packer-Scott (1942) freezing-drying apparatus.

Fig. 2. Arrangement for supporting the Packer-Scott (1942)
freezing-drying apparatus.

In order to ascertain something of the state of dehydration of
the tissue the ionization pressure gauges F and I in Figure 1 are
fitted into tube G. When the rate of evaporation falls to a very

8 MICROSCOPIC TECHNIQUES

low value, the pressure gradient between F and / disappears. How-
ever, since water may diffuse out, it has been found desirable to
continue the pumping for a day or more after the pressures at
F and / are coincident and remain so.

The ionization gauges used, of the type described by Montgomery
and Montgomery ( 1938) , are No. 47 radio tubes of Radio Corpora-
tion of America, which are sealed to the vacuum system with black
vacuum wax, care being taken to avoid knocking off the filament
coating during the sealing. A power pack is employed consisting
of a half-wave rectifier, filter system, and power transformer. An
additional filament transformer is included to supply the filament
of the second gauge. In order that the same meters and galva-
nometer may be used for both gauges a double-throw triple-pole
switch is employed. For pressures less than 3 X 10"^ mm. of mer-
cury a student type of wall galvanometer is used to measure positive
ion current in the gauges; this current is directly proportional to
the pressure. For higher pressures a microammeter with a range
of 0-50 may be used. Since the usual calibration constant (1
yuamp. for 7 X lO"*' mm. of mercury) is for air, a different constant
would apply in the presence of water vapor, hence only relative
pressures are obtained.

PROCEDURE

1. Place paraffin in apparatus; melt and degas it in vacuo by
means of the mechanical pump alone.

2. Let the paraffin solidify and raise the cooling Dewar flask
around the drying chamber. When equilibrium is attained the
system is ready for the tissue.

3. Either freeze the tissues in liquid air or, preferably, in isopen-
tane at liquid air or liquid nitrogen temperatures. Violently agitate
the isopentane to speed the heat extraction.

4. Place frozen tissue in the copper gauze basket and transfer
immediately to drying chamber of apparatus.

5. Start the mechanical pump at once; then start the diffusion
pumps and run for 12 hr. before taking pressure readings. The
gauge filaments must be heated and gas allowed to escape for
several hours before reliable readings are possible. After the read-
ings of both gauges are equal, continue pumping for some time
depending on the size, shape, and character of the tissue. No rule

FREEZING-DRYING PREPARATION OF TISSUE 9

can be applied here since the time for total dehydration is a function
of a number of poorly defined variables.

6. When dehydration is considered complete, lower away the
Dewar cooling flask and allow the tissue to come to room tempera-
ture.

7. Start the paraffin heater and keep the temperature of the
paraffin just above its melting point. The top of the paraffin will
melt first; regulate the heating so that a portion of solid paraffin
remains at the bottom of the chamber. Conditions of — 66° and
4 X 10"^ mm. of mercury are obtained routinely during operation.
The tissue will sink into the melted paraffin without causing a single
bubble to rise if the tissue is properly dehydrated and the paraffin
completely degassed; otherwise bubbling will occur.

8. Break the vacuum after infiltration is complete and the par-
affin has been allowed to solidify; remove and block for cutting.

//. CHEMICAL METHODS

A. REQUIREMENTS

The microscopic technique employing chemical methods depends
in almost every case on the direct observation of an insoluble
product of a microchemical reaction between the substance or group
whose distribution is being investigated and a suitable reagent.
Since the whole purpose of these methods is to visualize the presence
of a cellular or intercellular constituent in situ, it is essential that
the tissue be handled in a manner that will not permit the con-
stituent to diffuse or change its anatomical position during the pro-
cedure. The minimum requirements of the chemical method then
may be listed as:

1. The preparation of microtome sections in which there has
been no significant alteration in the position of the constituent
being investigated.

2. A reagent which is specific for this tissue constituent.

3. A reaction between the reagent and constituent which is of
such a nature, and rapid enough, to obviate diffusion of the con-
stituent or of the reaction product.

4. A reaction product, thus trapped in situ, which is capable of
being visualized.

The frequency with which these requirements can be met is,
unfortunately, still very low. The problem is most difficult in the
case of those constituents which are diffusible in solution, e.g.,
inorganic ions. While it is possible to prepare tissue sections without
the use of solutions by means of the freezing-drying technique, the
chemical formation of a substance in these sections for purposes of
visualization involves the use of a reagent in solution. One might
imagine that, if the interaction of the reagent in solution with the
ion in the tissue were rapid, the ion would be bound as an insoluble

11

12 MICROSCOPIC TECHNIQUES

substance before serious diffusion could occur. Still, in the case
of the precipitation of tissue chloride by silver nitrate solution, Scott
and Packer (1939) pointed out that differences in ionic mobilities
and the effects of ionic charges at cellular interfaces might easily
produce precipitations in regions different from those in which
the chloride originally existed. On the other hand, Gersh (1941)
claimed that the results he obtained from chloride distribution,
using silver nitrate as the reagent, were valid as borne out by
related data obtained with entirely different biochemical methods.
Regardless of the merits in this particular instance, the dangers
indicated by Scott and Packer cannot be ignored, and no way has
yet been devised to really eliminate them ; they constitute a funda-
mental limitation in the application of the chemical methods of
microscopic technique.

In the special case of the localization of enzymes, the sites of
activity may be determined in tissue sections by immersing the
sections in a buffered substrate medium containing a reagent which
will bind one of the products of the enzymatic action in situ by
precipitation. In addition to the four requirements already listed
for determinations of the disposition of tissue constituents, it is also
necessary that the following be included for enzyme methods:

5. A reagent which when added to the buffered substrate will
react with one of the enzymatic products but not with the substrate
or buffer.

6. A reagent which will also have no untoward effect on the
enzyme.

7. If the enzymatic product which reacts with the reagent is a
substance pre-existing in the tissue, either the sites of enzyme action
must be different from those of the pre-existing substance, or the
increase in the amount of the visualized compound resulting from
the enzyme activity must be demonstrable, or, better yet, the sub-
stance must be removed in advance by a method which will not
take out the enzyme.

8. A control experiment in which either the substrate is omitted,
or a highly effective soluble enzyme inhibitor, such as fluoride, is
added (the inhibitor must not react with substrate, buffer, reagent,
or products") — the advantage of the inhibitor is that, in some cases,
naturally occurring substrate may be present with the enzyme and
thus give a false aspect to the nonenzyme control. No such control

CHEMICAL METHODS • 13

experiment is required if all substances pre-existing in the tissue and
capable of giving a positive reaction can be removed without seri-
ously reducing the enzyme activity. This has been accomplished
in certain tissues for the phosphatase test.

Many, if not most, of the tests described in the following pages
leave much to be desired. In some cases they have been developed
for particular tissues and cannot be adapted to others without a
certain amount of additional research. Most of the tests are clearly
not at all good. However, it is the purpose of the writer to present
the published methods for the localization of substances, groups,
and enzymes, even though they may be, and usually are, poor. In
this way the investigator who requires a particular method will
have at hand the procedures already developed, and, if they should
prove inadequate, at least he will have them as a basis from which
to work out improvements.

A word should be added concerning the mounting media employed
for tissue sections. The media given in the procedures that follow
are those used by the original authors. However, newer media are
available and they may be substituted for the balsam or damar
that have been employed in the past. A 60% solution of Clarite in
xylol appears to be superior to neutral balsam for mounting sections
since, according to Lillie (1941), Clarite does not promote the
fading of some stains to the degree that balsam does. Stowell and
Albers ( 1943) showed Clarite absorbs less visible light than balsam.
Tetrachloroethylene may also be used as a solvent for Clarite.
Clarite and Clarite X (also called Nevillite V and Nevillite No. 1,
respectively) are superior in all respects to balsam and damar,
according to Groat ( 1939) . A solution of 60% of the resin in 40%
of toluol by weight is recommended. The resins are clean, cheap,
water-white, inert, high-melting, absolutely neutral, and chemically
homogeneous. Clarite X undergoes a slight yellowing with age and
has a refractive index of 1.567 while Clarite is very color stable and
has a refractive index of 1.544.

The limited availability of certain reagents or enzyme substrates
may make it imperative to employ a considerably smaller volume
than is normally used in staining dishes and Coplin jars. The
hanging-drop technique (Glick and Fischer, 1945a) may be adopted
in these cases. The section on the slide is surrounded by a circle
of vaseline or stopcock grease, a drop of the reagent or substrate

14 MICROSCOPIC TECHNIQUES

solution is placed on the section, and to avoid evaporating when
prolonged contact between the liquid and the section is necessary,
a hanging-drop slide is placed over the section so that the drop is
enclosed by, but does not touch, the walls of the depression. The
two slides are bound together with a rubber band and carefully
inverted so that the section is covered by the hanging-drop.

B. INORGANIC ELEMENTS AND RADICALS

POTASSIUM

Macallum's original method for the histochemical detection of
potassium based on the precipitation of sodium potassium cobalti-
nitrite has been subject to modifications over the past thirty years.
The relatively recent modification of Gersh ( 1938) will be included
in the present work as well as the method of Carere-Comes ( 1938) ,
which depends on the development of an orange color with Siena
orange.

Gersh Modification of Macallum Method for Potassium

SPECIAL REAGENTS

Anhydrous Petroleum Ether freshly distilled over sodium {20-40°

b.p.).
Dried Paraffin {Grilbler, 50-52° ni.p.). Just before use heat at

100° or more in vacuo for about 15 min. or until bubbling stops.

12% Sodium Cobaltinitrite Solution. Dissolve 150 g. sodium nitrite

in 150 ml. hot water, cool to 40° (some crystals appear), add 50

g. cobalt nitrate crystals, while stirring rapidly add 50 ml. 50%

acetic acid in small portions, stopper, and shake well. Pass a

rapid stream of air through the solution and let stand overnight.

Siphon off clear hquid and filter. Add 200 ml. alcohol in small

portions to the filtrate with stirring. After a few hr. filter off

precipitate by suction. Wash precipitate four times with 25 ml,

portions of alcohol followed by two times with ether. Recrystal-

lize by dissolving each 10 g. solid in 15 ml. water and precipitate

with 35 ml. alcohol. Make up the 12% aqueous solution fresh

before use.

POTASSIUM 15

PROCEDURE

1. Subject tissue to freezing-drying (see page 3). ,

2. Transfer to paraffin, infiltrate in vacuo for 10-15 min. at
not more than 60°, and embed. Care should be taken to prevent
condensation of moisture on the dried tissue during the transfer
from the vacuum vessel to the paraffin.

3. Cut 15 fjL sections using no water or ice, mount near edges of
large chemically clean cover slips, press down with finger, melt
paraffin with a tiny flame, and press down again.

4. Remove paraffin by immersing the cover slips with the sections
in dry petroleum ether in a watch glass heated on a warm bar.
(Keep watch glass covered by another at all times except during
actual use. Replace the petroleum ether often.)

5. Remove from petroleum ether and burn off excess quickly.
Allow to cool.

From this point on, the test is carried out entirely in a cold room
the temperature of which should be 0° ± 1° during the manipula-
tions. All instruments and reagents are kept in this room. The
crystals of sodium potassium cobaltinitrite are relatively soluble at
room temperature, hence the precautions to maintain cold.

6. When cover slips with sections are cold, cover each section
with a drop of the sodium cobaltinitrite solution.

7. Drain off the solution and replace with glycerol.

8. Mount on clean slide with section between slide and cover
slip.

9. Examine under microscope with oil immersion lens. (Light
mineral oil is substituted for cedar oil since the latter is too viscous
at 0°.)

Result. Short yellow rods with rounded ends in a diffuse pale
yellow background are the crystals of sodium potassium cobalti-
nitrite.

Carere-Comes Siena Orange Method for Potassium

SPECIAL REAGENTS

Siena Orange Solution. Aqueous sodium p-dipicrylamine (prepared

ready for use by K. Hollborn & Sons) .
10% Hydrochloric acid, p

16 MICROSCOPIC TECHNIQUES

PROCEDURE

1. Fix tissue in neutral formalin and prepare paraffin sections
as usual.

2. Immerse deparaffinized sections in Siena orange soln. for 2
min.

3. Transfer to 10% hydrochloric acid for 3 min.

4. Wash twice in distilled water for 10 min., blot with filter
paper, and dry at 37°.

5. Mount in thickened cedar oil.

Result. Potassium is demonstrated by an orange color on a pale
yellow or colorless background. The author of this method has
failed to consider the effects of potassium diffusion when aqueous
solutions are employed for fixation, etc. Modification of the pro-
cedure to obviate this difficulty would be essential.

CALCIUM

A critical survey of histochemical tests for calcium was pre-
sented by Cameron (1930), who concluded that none of the tests
can be considered wholly specific. In all cases calcium must be
converted to an insoluble salt, if it is not already present as such,
and the insoluble compound is identified directly or it is made more
easily detectable by staining or conversion to a colored compound.
For visualization of calcium in the form of phosphate or carbonate
see page 78. In addition to these tests the Cretin (1924) gallic
acid color test has been extensively used, as has the formation of a
red precipitate by reaction of calcium salts with sodium alizarin
sulfonate (Pollack, 1928). For plant materials it is often sufficient
to produce and identify crystals of the oxalate, carbonate, or sulfate
(Lee's Vade Mecum, pages 293 and 668). The old test of von Kossa
(1901), depending on the reduction of silver salts under bright
light, has been championed by Gomori ( 1945a) . While this method
will demonstrate inorganic deposits in general, it can be considered
specific for calcium in bone or cartilage because the calcium salts
are the only ones present in significant amounts. An adaptation
of the von Kossa test to bone has been described by McLean and
Bloom«( 1940) and Bloom and Bloom ( 1940) .

CALCIUM 17

Calcium Sulfate Test for Calcium in Plant Tissue

SPECIAL REAGENTS

3% Sulfuric Acid.
40% Alcohol.

PROCEDURE

1. Fix tissue in acid-free alcohol or acid-free formalin.

2. Cut sections and bring them down to 40% alcohol.

3. Add 3% sulfuric acid to sections under cover slip.

4. Examine for colorless monoclinic needles of calcium sulfate.

Cretin Color Test for Calcium and Other Minerals

SPECIAL REAGENTS

Gallic Acid Reagent. Grind 0.1 g. trioxymethylene (metaformal-
dehyde) and 0.2 g. gallic acid in a mortar. Dissolve 0.25 g. of
the mixture in 5 ml. boiling distilled water and add 0.5 ml.
ammonium hydroxide (18° Baume, or 14% ammonia) ; stir until
the solution becomes straw colored. When this reagent turns
brown or rose, which it will in a short time, it can no longer be
used.

PROCEDURE

1. Prepare paraffin sections as usual. Remove the paraffin with
xylol and the xylol with chloroform.

2. After excess chloroform has been removed, add gallic acid
reagent. In 10-15 sec, drain off excess reagent and expose slide to
air,

3. Examine when color appears. An eosin counterstain may be
applied (Lison, 1936, page 76).

Result. Calcium gives a blue, barium a bright green, strontium
a water blue-green, cadmium a bluish-gi-een, magnesium a yellowish-
rose, iron a deep violet-brown, zinc and lead a dull yellow, and
silicon a pure yellow color.

Von Kossa Silver Test for Calcium

SPECIAL REAGENTS

5% Silver Nitrate.

18 MICROSCOPIC TECHNIQUES

PROCEDLUE

1. Transfer frozen or parafl5n sections which have been washed
with distilled water to the silver nitrate soln. in the dark for up to 1
hr. ; wash with distilled water in the dark, and expose to bright hght
for 30 min. or longer.

2. Wash well in distilled water, dehydrate, clear, and mount.
Result. Calcium salts are rendered black.

MAGNESIUM

A method for the demonstration of magnesium in plant cells was
developed by Broda ( 1939) . The principle of this method could be
adapted to studies on animal tissue. Most of the tests used for
calcium also give positive results for magnesium.

Broda Method for 3Iagnesiuin

SPECIAL REAGENTS

Quinalizarin Reagent. Triturate 100 mg. quinalizarin and 500 mg.

sodium acetate crystals, and dissolve 500 mg. of the mixture in

100 ml. of 5% sodium hydroxide.
0.2% Titian Yellow.
10% Sodium Hydroxide.
0.1% Azo Blue.

PROCEDURE

1. Prepare paraffin sections as usual.

2. Add 1-2 drops of quinalizarin reagent to a section on the slide,
followed by 1-2 drops of 10% sodium hydroxide.

3. To a different section add 1-2 drops of the Titian yellow
solution followed by 1-2 drops of 10% sodium hydroxide.

4. To another section add a drop or two of the azo blue dye.
Result. In the presence of magnesium the quinalizarin reagent

develops a blue color over several hours, the Titian yellow a brick-
red, and the azo d3'e a violet stain.

ZINC

Very little has been done in regard to the histological localization
of zinc and the procedure of Mendel and Bradley (1905) is still

ZINC AXD IBON 19

the sole method that has been developed. Zinc is precipitated by
nitropmsside and the precipitate is brought out as a deep purple
by treatment with sulfide.

Mendel and Bradley Alethod for Zinc

SPECI.\L REAGENTS

10% Sodium Xitroprusside.

Potassium Sulfide Solution. (Concentration not stated, but 1-5%
should suffice.)

PROCEDLTIE

1. Prepare paraffin sections. (!Mode of fixing tissue not given,
but, as in all other cases, the freezing-drying treatment, see page 3,
would be preferable.)

2. Treat sections with the nitropmsside soln. for 15 min. at 50^.

3. Cool, and wash in a stream of water for about 15 min.

4. Introduce under cover glass placed on section 1 drop of the
sulfide soln.

Result. Zinc elicits an intense purple color.

IRON

The classical histochemical tests for iron are the Prussian and
Turnbtiirs blue reactions and the hematoxylin method of Macallum.
the latter being the least specific (Lee's Vade Mecum, pages 2S9-
292). The Prussian blue test will detect ferric, and Turnbull's blue
ferrous, iron. More recently other methods have been proposed for
which certain advantages have been claimed. The dinitrosore-
sorcinol test of Humphrey (1935) brings out iron as a rich green
of pristine brilliance and the color is much more permanent than
that of Prussian or Tiu-nbull's blue, which fades after a year or two.
Thomas and Lavollay (1935) employed the 8-hydroxj-quinoline
reaction to ^-isualize iron in greenish-black; other metals appearing
in various shades of green and yellow. The strong flourescences of
metallic 8-hydrox^-quinolinates may also be used for identifications
(see page 108).

Iron, like many other metals, occurs in tissues both in the in-
organic or free form, and in the organic or bound form. Before
bound iron can be visualized it must be converted to the free form.

\

c K A R ^ ]

/ >

■^^-rlf

20 MICROSCOPIC TECHNIQUES

Macallum's (1908) technique is still employed for this conversion
and it consists of a treatment of deparaffinized sections with a solu-
tion of either nitric or sulfuric acid in alcohol. The iron liberated
is chiefly in the ferric form. In all tests special care must be taken
to protect tissues and fluids from dust. Iron will also appear in
the tests for lead and copper (see page 22).

Precautions to prevent diffusion of the iron in aqueous solutions
have not been sufficiently exercised in the following procedures.
The investigator should modify them accordingly.

Prussian Blue Test for Ferric Iron and TurnbuU's
Blue for Ferrous Iron

SPECIAL REAGENTS

Prussian Blue Reagent. 2% potassium ferrocyanide (use fresh

soln.).
TurnbuU's Blue Reagent. 2% potassium ferricyanide (use fresh

soln.).
Acid Alcohol. 1% hydrochloric acid in 70% alcohol.
Organic Iron Reagent. Equal vol. of 1.5% potassium ferrocyanide

and 0.5% hydrochloric acid (use fresh soln.).
Organic Iron Conversion Reagent. 3% nitric acid, or 4% sulfuric

acid, in 95% alcohol. (The sulfuric reagent acts more slowly.)

PROCEDURE FOR INORGANIC IRON

1. Fix in 95% alcohol for 24-48 hr.

2. Prepare paraffin sections as usual ( care must be taken to mini-
mize contact with iron — the microtome knife must be free of rust
and not freshly honed and glass needles should be substituted for the
steel ones) .

3. After removing paraffin and passing down to distilled water,
place sections in either the Prussian or TurnbuU's blue reagent for
3-15 min. (If both ferric and ferrous iron are to be visualized, use
a mixture of equal vol. of the two reagents.)

4. Wash in water containing eosin or safranin to counterstain.

5. Dehydrate, clear, and mount in benzol balsam.

PROCEDURE FOR ORGANIC IRON

1-2. Same as inorganic iron.

3. Liberate iron from the bound forms by treating deparaffinized

IRON 21

sections, brought down to water, with the conversion reagent for
24-36 hr. at 35°.

4. Wash in 90% alcohol followed by distilled water.

5. Place in the organic iron reagent for not over 5 min.
6—7. Same as 4-5 for inorganic iron.

Result. The iron appears blue.

Humphrey Dinitrosoresorcinol Test for Iron

SPECIAL REAGENTS

30% Ammonium Sulfide.

Saturated Aqueous Dinitrosoresorcinol or a 3% soln. in 50% alcohol,
(A few days aging improves the reagent; it is stable.)

PROCEDURE

1. Prepare formalin-fixed paraffin sections.

2. Remove paraffin; bring down to water, and place in the am-
monium sulfide solution for 1 min.

3. Rinse in distilled water and place in the dinitrosoresorcinol
reagent for 6-20 hr. depending on the depth of the brown background
desired.

4. Wash in water or dilute alcohol depending on whether an
aqueous or alcoholic reagent was used.

5. Pass through alcohols, carboxylol and xylol, and mount in
balsam.

Result. The iron appears bright dark green against a reddish or
rich brown background. For organic iron introduce steps 3-4 in
Prussian blue procedure ( see page 20) between steps 1 and 2 above.

Thomas and LavoUay Hydroxyquinoline Test for Iron

SPECIAL REAGENTS

Hydroxyquinoline Reagent. Dissolve 2.5 g. 8-hydroxyquinoline in
4 ml. glacial acetic acid with the aid of gentle warming. Quickly
add distilled water to bring volume to 100 ml. and filter the soln.

25% Ammonium Hydroxide.

PROCEDURE

1. Fix tissue in alcohol, neutral formalin, or trichloroacetic acid
soln.

2. Prepare frozen or paraffin sections as usual.

22 MICROSCOPIC TECHNIQUES

3. To the sections washed with neutral distilled water add a
few drops of the hydroxyquinoline reagent, and after 5-15 min. drain
off the liquid.

4. Add 1 drop of 25% ammonium hydroxide to form precipitate.

5. Wash in neutral distilled water. No large crystals should re-
main.

6. A lithium carmine nuclear stain may be applied.

7. Dehydrate with terpinol and mount in petrolatum, or examine
directly in neutral water.

Result. Iron appears as a greenish-black, calcium as a pale
yellow, magnesium as a straw-yellow, aluminum as a yellowish-
green, zinc and manganese as a yellow, and copper as a greenish-
yellow precipitate.

NICKEL

A method has been devised by Cretin and Pouyanne ( 1933) for
the histological demonstration of nickel in bone material by pre-
cipitation of nickel ammonium phosphate.

Cretin and Pouyanne Method for Nickel

SPECIAL REAGENTS

Fixative. Add 30 ml. formalin and 5 drops ammonium hydrosulfide

to 100 ml. physiological saline soln.
10% Ammonium Phosphate.

PROCEDURE

1. Fix tissue in the special fixative soln.

2. Transfer to the ammonium phosphate soln. in order to pre-
cipitate the insoluble nickel ammonium phosphate.

3. Decalcify and section.

4. Stain the nickel compound with alcoholic hematoxylin.

5. Wash, dehydrate, clear, and mount.

Result. Nickel will appear as a lilac deposit, or blue if present
in abundance.

LEAD AND COPPER

For many years the chromate method has been used for the micro-
chemical detection of lead in tissues ( Frankenberger, 1921; Cretin,

LEAD AND COPPER 23

1929) . This procedure depends on the formation of a yellow precipi-
tate of lead chromate when lead-bearing tissue is fixed in Regaud
fluid (20 ml. of 3% potassium dichromate plus 5 ml. formalin) . Lison
( 1936, page 101) has discussed this method as well as the test based
on precipitation of the sulfide, and rather favors the former. Oka-
moto and Utamura (1938) employed 2^-dimethylaminobenzylidene
rhodanine to produce a reddish-violet precipitate with copper in tis-
sues, a reaction given by gold, silver, and other metals (see pages
26, 28, and 29) .

Mallory and Parker (1939) described a method using hema-
toxylin and another employing methylene blue which would visual-
ize both lead and copper. The methylene blue technique was particu-
larly recommended for photomicrography of lead because of the in-
tense blue color developed.

In a study of the histological distribution of copper in the blowfly,
Waterhouse ( 1945) found that the only reagent which could be used,
of those tested, was sodium diethyl dithiocarbamate, which formed a
yellow product with copper. Waterhouse's technique was to drop a
0.1% aqueous solution on the fresh tissue followed by a drop of con-
centrated hydrochloric acid. The acid allowed greater penetration
of the reagent into the cells. Iron can interfere with this test by the
formation of a brown carbamate; however, the reagent can detect 1
part of copper in 100 million and its sensitivity to iron does not
approach this.

Mallory and Parker Hematoxylin Method for
Lead and Copper

SPECIAL REAGENTS * -

Hematoxylin Reagent., Dissolve 5-10 mg. hematoxylin in a few
drops of 95% or absolute alcohol and add 10 ml. freshly filtered
2% dibasic potassium phosphate.

PROCEDURE

1. Fix tissue in 95% or absolute alcohol (formalin may be used
for copper) .

2. Prepare celloidin sections as usual.

3. Stain sections for 2-3 hr. at 54°.

4. Wash in several changes of tap water 10 to 60 min.

24 MICROSCOPIC TECHNIQUES

5. Dehydrate in 95% alcohol, clear in terpinol, and mount in
terpinol balsam.

Result. Lead appears as light to dark grayish-blue and nuclei
as deep blue. Copper or hemofuscin pigment is brought out as an in-
tense blue. Inorganic iron or the pigment, hemosiderin, appears black
provided alcohol was used as the fixative and light to dark brown
if formalin was employed.

Mallory and Parker Methylene Blue
Method for Lead and Copper

SPECIAL REAGENTS

Methylene Blue Reagent. 0.1% of the dye in 20% alcohoL

PROCEDURE

L Fix tissue in Zenker fluid.

2. Prepare paraffin sections as usual, and apply a contrast stain
of phloxine if desired.

3. Treat sections for 10-20 min. with the methylene blue reagent
and decolorize in 95% alcohol for about the same time.

4. Dehydrate, clear, and mount as usual.

Result. Lead is colored intense blue. Copper or hemofuscin ap-
pears pale blue while iron pigment is not colored and hence appears
yellow to light brown. When pigment has both copper and iron it
develops a green color.

MERCURY

Three methods for the visualization of mercury in tissue sections
are given in Lison (1936, page 102). The mercury can be trans-
formed into the black sulfide, reduced by stannous chloride to give
the black free metal, or a violet precipitate can be formed with di-
phenylcarbazide. In addition to these, Okamoto's method for silver
(page 26) using p-dimethylaminobenzylidene rhodanine can be em-
ployed to give a reddish-violet precipitate with mercury.

After trials of the sulfide, diphenylcarbazide, and reduction
methods, Hand et at. ( 1943) favor the latter. They detected mercu-
rous mercury by reducing it to the metal by means of thioglycollic
acid, and the mercuric form was visualized by reducing with stannous
chloride.

MERCURY AND SILVER 25

Method of Hand et al. for Mercurous and Mercuric Mercury

SPECIAL REAGENTS

Mercurous Reagent. Combine 1 ml. thioglycollic acid with 9 ml.

glycerol.
Mercuric Reagent. Combine 5 g. stannous chloride, 5 g. tartaric

acid, and 100 ml. glycerol, and heat until clear. Stabilize by adding

a few grams of metallic tin to the final soln., which should be

stored in a stoppered bottle.
Iodine Reagent. Dissolve 50 g. potassium iodide in 50 ml. distilled

water; add 70 g. iodine and when it has dissolved, add 95% alcohol

to make 1 liter.
1% Chloroauric Acid. Store in a dark bottle.
Control Reagent. Add 5 g. tartaric acid to 100 ml. glycerol. Let
stand overnight to dissolve.

PROCEDURE

1. Prepare fresh frozen sections of tissue 15 /x thick.

2. Place sections on slides and allow to dry.

3. Cover each section with a drop of one of the reagents, depend-
ing on the test to be applied, fit on a cover slip, and blot away excess
reagent.

4. Seal edges of cover slip with commercial gold size (adhesive
used to hold gold foil on glass) .

5. After 10 min. examine under a microscope, comparing sections
with control reagent to those with other reagents. The sections
treated with the mercuric and control reagents remain unchanged
for at least 2 weeks.

Result. The metallic mercury formed in the tissue appears as
minute black spheres which may be dissolved by tincture of iodine,
or made to lose their glossy surface by forming gold amalgam on
treatment with chloroauric acid. In the test for mercurous mercury
characteristic yellowish crystals appear after about 5 min., in addi-
tion to the mercury globules, when mercuric mercury is also present.

SILVER

Particles of reduced silver in tissues may be made more intensely
black by treatment with dilute ammonium sulfide solution.

Okamoto, Utamura, and Akagi (1939) employed the p-dimethyl-

26 MICROSCOPIC TECHNIQUES

aminobenzylidene rhodanine reagent for the precipitation and vis-
ualization of silver in tissue sections. The fact that not only silver,
but also copper, gold, mercury, platinum, and palladium are like-
wise precipitated by this reagent offers little ground for concern
since these elements are not apt to exist in significant amounts in
tissues unless introduced experimentally or perhaps by accidental
poisoning. In these cases only one of the elements at a time is likely
to be present. However, certain differentiations can be made, if it is
assumed that more than one is present, on the basis of the varying
solubility behavior of the precipitated compounds. Thus divalent
copper reacts with the reagent only in neutral solution, whereas
monovalent copper and the other metals will react in either neutral
or acid solution. Furthermore the mercury precipitate is soluble in
dilute hydrochloric acid, the silver compound in potassium bromide
solution, and gold compound in potassium nitrite solution. Neutral
stannous chloride solution reduces the palladium precipitate to the
free metal which can then be converted to the chloride by means of
chlorine gas; this cannot be done with the platinum compound. Based
on these facts, possible separations have been suggested by Okamoto
et al.

Okamoto et al. Procedure for Silver

SPECIAL REAGENTS

Precipitation Reagent I. Add 3-5 ml. of saturated soln. of p-di-
methylaminobenzylidene rhodanine in absolute alcohol to 1-3 ml.
1 N nitric acid and 100 ml. distilled water.

Precipitation Reagent II. Combine 10-20 ml. of the saturated alco-
holic soln. of the rhodanine compound with 1-3 ml. 1 N nitric acid,
5-10 ml. 3% hydrogen peroxide, and 100 ml. distilled water.

Precipitation Reagent III. Add 2-5 ml. of the soln. of the rhoda-
nine derivative to 2 ml. 0.1 N hydrochloric acid, 3-5 ml. 1 N nitric
acid, and 100 ml. distilled water.

PROCEDURE

1. After fixing the tissue in absolute alcohol or neutral formalin,
prepare either celloidin, paraffin, or frozen sections.

2. Place the sections in precipitation reagent I for 24 hr. at 36°,
keeping the vessel closed.

SILVER AND GOLD 27

3. Rinse sections in distilled water and counterstain with hema-
toxylin.

4. Dehydrate, clear, and mount in balsam.

Result. Silver in the tissue is colored reddish-violet. The other
metals will produce shades of the same color.

Monovalent copper may be eliminated from visualization by
substituting precipitation reagent II for I, and mercury may be
similarly eliminated by employing reagent III. The silver precipitate
could be removed from the others by treating the colored sections
with 1% potassium bromide.

GOLD

Several methods have been used for the localization of gold in
tissues and Lison (1936. page 100) has discussed those of Christeller,
of Borchardt, and of Okkels. The first depends on treatment with
stannous chloride; the second employs silver nitrate followed by
nitric acid, and the last merely involves exposure to ultraviolet light
to obtain blackening of the gold granules. The more recent methods
of Roberts (1935), Okamoto, Akagi, and Mikami (1939), and Elft-
man and Elftman ( 1945) follow. The last-named method is probably
the best since it avoids the use of ions that might give rise to arti-
facts, and effects the bleaching of interfering pigments.

Roberts Method for Gold

SPECIAL REAGENTS

Sliver Nitrate Reagent. Just before using dissolve 2 g. pure silver
nitrate in 100 ml. 10% gum arabic soln. in the dark.

Hydroquinone Reagent. The day before using dissolve 1 g. pure
hydroquinone in 100 ml. 10% gum arabic soln.

5% Citric Acid.

5% Sodium Hyposulfite.

PROCEDURE

1. Fix tissue in Bouin fluid or 20% neutral formalin and wash
well with water.

2. Prepare paraffin or frozen sections.

3. Place sections for 5-10 min. in a fresh mixture of 2 ml. silver
nitrate reagent, 2 ml. hydroquinone reagent, and 1-3 drops of 5%

28 MICROSCOPIC TECHNIQUES

citric acid. Shake the mixture for 30 sec. immediately after prepar-
ing it.

4. Transfer sections rapidly to a 5% sodium hyposulfite soln. and
after a few min. wash thoroughly in water.

5. Dehydrate, clear, and mount.

Result. Granules of gold appear black due to a surface deposit
of silver.

Elftman and Elftman Method for Gold

SPECIAL REAGENTS

3 % Hydrogen Peroxide.

PROCEDURE

1. Fix tissue in neutral formalin, prepare paraffin sections, mount
with the aid of egg albumin, and then expose to formaldehyde vapor
for 1 hr. to increase the affixation.

2. After removing the paraffin and running down to water, place
in 3% hydrogen peroxide at 37° for at least 24 hr.; in most cases 3
days or longer gives the best results.

3. Do not stain the sections since staining may mask the gold
deposit. If staining is required, however, the interference is only
slight with light green SF (yellowish) or hematoxylin, and eosin
can be used when the gold is present as a sufficiently dense deposit.

4. Wash the sections in distilled water and dehydrate in the
usual manner.

5. Mount the sections in damar.

Result. Gold is made apparent by its presence in colloidal form.
The color of the deposit depends on the particle size, and accordingly
shades from rose to purple to blue and black are obtained. Usually
rose predominates.

Okamoto et al. Method for Gold

The procedure is the same as that in the Okamoto et al. method
for silver using precipitation reagent II (page 26). The gold ap-
pears as a reddish-violet or brownish-red precipitate. The removal

PLATINUM, PALLADIUM, AND URANIUM 29

of the colored silver precipitate from the sections, if silver was pres-
ent, can be carried out by treating with a saturated soln. of potas-
sium bromide for 1 hr. or more. If the sections are then washed in
distilled water and placed in 1% potassium nitrite for 24 hr. or
longer at 36° (or heated in the nitrite soln. for 1 min.) the gold
precipitate will dissolve leaving those of platinum or palladium,
should either of these be present.

PLATINUM

The Okamoto et al. method for silver may be employed unchanged
for the detection of platinum in tissues (see page 26).

PALLADIUM

The method of Okamoto, Mikami, and Nishida (1939) for the
visualization of palladium in sections of tissue follows the Okamoto
et al. silver method (p. 26) with 1 difference. Between steps 1 and
2 in the procedure, the following is introduced: treat the dry sections
with chlorine gas until the black palladium granules are made
colorless.

URANIUM

Two chemical tests have been presented for the localization of
uranium in tissue sections. Both are founded on the precipitation of
dark brown uranium ferrocyanide. Schneider ( 1903) was the first
to use this technique on the tissues of animals that had been in-
jected with uranium salts. Gerard and Cordier (1932) followed the
Prussian blue method for iron and reported good results. The latter
employed Bouin-Hollande or Carnoy fixatives and their coloring
reagent was 2% potassium ferrocyanide containing 2% hydrochlo-
ride acid. For details of the test, see the Prussian blue procedure for
iron, page 20.

The fluorescent properties of uranium salts subjected to ultra-
violet radiation can be utilized for the detection of these salts in in-
cinerated sections of tissue as indicated by Policard and Okkels
( 1930) ; see page 145.

ft.

30 MICROSCOPIC TECHNIQUES

ARSENIC

Castel ( 1934-1935a, 1936) developed two methods for the histo-
logical localization of arsenic. In one the tissue is fixed in an abso-
lute alcohol-chloroform-hydrochloric acid mixture saturated with
hydrogen sulfide, and the appearance of yellow granules was be-
lieved by Castel to be due to the formation of arsenous sulfide. A
reinvestigation of this technique by Tannenholz and Muir (1933)
led them to conclude that the yellow granules formed were not re-
lated to the presence of arsenic but were more likely composed of a
sulfur-protein complex.

The other method of Castel was based on the precipitation of
either cupric hydrogen arsenite (Scheele's green) or the cupric ace-
tate-cupric arsenite double salt ( Schweinf iirter green), and this pro-
cedure appears to be a reliable one.

Castel Cupric Salt Method for Arsenic

SPECIAL REAGENTS

Formalin-Copper Salt Reagent. Add 2.5 g. cupric sulfate or neutral
cupric acetate to 100 ml. metal-free 10% formalin (hydrogen sul-
fide is used to test for traces of metals in the formalin).

PROCEDURE

1. Place small pieces of tissue in the formalin-copper salt reagent
for 5 days.

2. Wash tissue in running water for 1 day.

3. Prepare paraffin sections as usual and examine after re-
moval of the paraffin.

Result. Green granules are indicative of arsenic.

BISMUTH

The histochemical detection of bismuth is founded on the reaction
of Leger ( 1888) , which is the precipitation of the double iodide of
bismuth and an alkaloid. Komaya (1925) and Christeller (1926)
employed the quinine salt for their tissue studies, and later Castel
( 1936) suggested the use of the brucine salt to avoid the interfer-
ence of iron which plagues the quinine method. He also modified
the earlier procedures by substituting sulfuric for nitric acid in the
reagents, Castel ( 1934-1935b), a change which enables the visual-

BISMUTH 31

ization of the bismuth as a mure reddish, rather than a yellowish-
orange deposit. More recently Wachstein and Zak ( 1946) employed
a modified Castel method in which the black sulfide, the form in
which bismuth appears in tissues, is converted to the white sulfate
by treatment with hydrogen peroxide and the sulfate is then trans-
formed to the brucine iodide salt.

In the procedure of Wachstein and Zak (1946), iron, which may
be present as the black sulfide, is oxidized to golden brown hemosid-
erin, which does not react wath Castel reagent, but which does give
the typical iron reactions. Wachstein and Zak pointed out that lead
sulfide would deposit in tissues in the same fashion as bismuth but
differentiation may be made by the fact that, after the lead sulfide
is converted to the sulfate by the peroxide, it will yield the slightly
yellowish lead iodide on treatment with Castel reagent in contrast
to the brilliant orange-red bismuth product. Similarly silver and
mercury will give yellow, and copper brown, iodides that can be
differentiated from the color of the bismuth precipitate. Melanin in
tissue is not bleached by the short treatment with peroxide and does
not react with Castel reagent. Wachstein and Zak emphasized that
melanin never impregnates capillary walls while bismuth does.

Wachstein and Zak Method for Bismuth

SPECIAL REAGENTS

Modified Castel Reagent. Dissolve 0.25 g. brucine sulfate in 100
ml. distilled water containing 2-3 drops cone, sulfuric acid. After
the brucine salt has dissolved add 2 g. potassium iodide. Store in
a brown bottle and filter before use.

Diluted Castel' s Reagent. Add 3 vol. distilled water to 1 vol.
reagent.

30% Hydrogen Peroxide. (Superoxol, Merck). Store in a refrig-
erator.

Levulose Solution. Dissolve 30 g. levulose in 20 ml. water by warm-
ing at 37'^ for 24 hr. and add a drop of the diluted Castel reagent.

Counterstain Solution. Add 1 ml. 1% aqueous light green SF
{Hartman-Leddon) to 100 ml. undiluted Castel reagent. Filter
before use.

PROCEDURE

1. Prepare either frozen or paraffin sections of formalin-fixed
tissue.

32 MICROSCOPIC TECHNIQUES

2. Treat the tissue on the slide for a few sec. with several drops
of 30% hydrogen peroxide to remove the black sulfide color.

3. Wash well in tap water and place in the Castel reagent for 1 hr.

4. Transfer to the diluted Castel reagent and shake gently to re-
move precipitates.

5. Remove most of the liquid from the slide by careful blotting
and mount in the levulose soln.

Result. Bismuth is indicated by the orange-red deposit. The
color may darken on standing. (If a counterstain is desired, stain
for 4 min. with the counterstain soln.)

Castel Method for Bismuth

SPECIAL REAGENTS

Bismuth Reagent. With the aid of warming, dissolve 1 g. brucine
in 100 ml. distilled water containing 3-4 drops of sulfuric acid
and add 2 g. potassium iodide. As an alternate reagent dissolve
1 g. brucine and 2 g. potassium iodide in 100 ml. of a mixture of
equal vol. alcohol and chloroform,

PROCEDURE

1. Fix the tissue in 10% formalin and prepare paraffin sections.

2. Treat deparaffinized sections for 15 min. with the bismuth
reagent and wash well in distilled water.

3. Mount in syrup of Apathy (heat equal parts of paraffin, m.p.
60°, and Canada balsam).

Result. Red granules are indicative of bismuth.

CHLORIDE AND PHOSPHATE-CARBONATE

Earlier methods for chloride, including Macallum's ( 1908) silver
test, were subject to the difficulty that, in the course of the manipu-
lations, a shift in the topographical distribution of chloride oc-
curred. Distortion due to this cause can be minimized by applying
the freezing-drying process to the tissue before further treatment.
Gersh ( 1938) makes use of this fact in his procedure, which enables
a differentiation between the chloride in the tissue and the phos-
phate and carbonate present. Phosphate and carbonate are visualized
together in this method. Two reagents are used, one permits visuali-
zation of chloride specifically by effecting the maximum precipita-

CHLORIDE AND PHOSPHATE-CARBONATE 33

tion of chloride in the presence of phosphate and carbonate by means
of the phosphoric acid it contains. The acid holds the phosphate in
solution and decomposes the carbonate. The other reagent, with-
out phosphoric acid, precipitates chloride, phosphate, and carbonate.
By comparing sections separately treated with each reagent, chloride
can be differentiated from phosphate and carbonate.

Gersh Method for Chloride and Phosphate-Carbonate

SPECIAL REAGENTS

Anhydrous Petroleum Ether freshly distilled over sodium {b.p.

20-40°).
Dried Paraffin ( GriXhler, m.p. 50-52°). Just before use heat at 100°

or more in vacuo for about 15 min. or until bubbling stops.
Silver Nitrate Reagent 1. To 60% silver nitrate solution add

enough cone, phosphoric acid to prevent precipitation of high

concentrations of phosphate, then saturate with silver chloride.

Filter and add 2-3 drops distilled water to each 10 ml. before

using.
Silver Nitrate Reagent 2. Saturate 60% silver nitrate solution with

silver phosphate and silver chloride. Filter and add water before

using as for reagent 1. Store both reagents in glass-stoppered

brown bottles in the dark.

PROCEDURE

1-5. These steps are identical with those in Gersh method for
potassium (see page 14).

6. Cover sections on one cover slip with reagent 1 and those on
another with reagent 2.

7. Drain off liquid from both cover slips and replace with a
drop of pure glycerol in each case.

8. Mount on clean slides with glycerol-covered sections down.

9. Expose both slides simultaneously to carbon arc radiation at
such distance as to avoid warming .

10. Examine microscopically at once by direct or dark-field il-
lumination. These preparations last only a short time. The highest
power to be used with the dark-field condenser is a 4 mm. high-dry
or 2 mm. oil immersion objective with a numerical aperture of 0.95.

Result. The reduced silver appears yellow to brown with or with-
out black or brown particles when viewed with direct illumination.

34 MICROSCOPIC TECHNIQUES

With the dark field the silver granules appear orange or rust colored.
Reagent 1 gives a test for chloride only. Reagent 2 gives a test for
chloride plus phosphate and carbonate.

IODIDE

A critical study of histochemical methods for the localization of
iodides in tissues was presented by Gersh and Stieglitz ( 1933) . After
a careful examination, these authors conclude that none of the pro-
posed methods is satisfactory. The difficulty is that any precipitat-
ing agent that might be used to fix iodide will also precipitate the
proteins and hence prevent its proper penetration into the tissue.

PHOSPHATE

Microscopic techniques for the detection of phosphorus in tissues
are usually based on reactions designed to visualize the phosphate
ion. Hence phosphate in organic combination must be liberated be-
fore it can be detected. Lison (1936, pages 113-1201 critically re-
viewed the various methods and considered Angeli's procedure to be
highly specific; this is a molybdate method using stannous chloride
as the reducing agent. Serra and Queiroz Lopes (1945) employed a
molybdate reaction with benzidine which they report to give a more
intense color than that developed by stannous chloride, and they
also use a more dilute acid medium, which is less damaging to the
tissue.

Johansen ( 1940, page 198) stated that phosphate may be identi-
fied in plant tissues by treating a section with a drop of solution pre-
pared by adding 25 ml. saturated magnesium sulfate and 2 ml. satu-
rated ammonium chloride to 15 ml. water. Crystals of magnesium
ammonium phosphate should form in the presence of phosphate. This
procedure is doubtlessly less advantageous than those previously
mentioned.

The method for visualization of enzymatically liberated phosphate
(page 78) may also be applied in certain instances.

Serra and Queiroz Lopes Modification of the Molybdate
Method for Phosphate

SPECIAL REAGENTS

Acetic Alcohol-Formalin Fixative. Add a few drops of glacial

<&

PHOSPHATE AND NITRATE 3o

acetic acid to 10 ml. of a mixture of 2 vol. 96% alcohol and 1 vol.

formalin.
Molybdate Solution. Dissolve 0.5 g. ammonium molybdate in 20

ml. distilled water, add 10 ml. cone. (30%) hydrochloric acid, and

dilute to 50 ml. with distilled water.
Acetic Benzidine Solution. Dissolve 25 mg. benzidine in 5 ml,

glacial acetic acid and dilute to 50 ml. with distilled water.
Saturated Sodium Acetate Solution.

PROCEDURE

1. Fix the tissue in the acetic alcohol-formalin mixture and wash
well in water.

2. In order to hydrolyze organic phosphates and precipitate the
free phosphate, treat small pieces of the tissue or frozen sections with
the molybdate reagent at 10-12° for 2-3 weeks, followed by 2-3
days at 20-25°. The temperature is kept low to prevent alteration of
the tissue and the rather long time is required to effect hydrolysis
with the relatively weak acid concentration of the reagent.

3. Cover the tissue with a drop of acetic benzidine soln. for 3
min. and then add 2 drops of the sodium acetate soln.

4. Mount in glycerol which has been stoi-ed with crystals of
sodium acetate in the bottle.

Result. An intense blue coloration characterizes phosphate.

note: Sena and Queiroz Lopez employ a digestion with nuclease to liberate
phosphate from nucleic acid. The visualization of this phosphate then serves
to indicate the nucleic acid.

NITRATE

Cramer ( 1940) developed a method for the histological demon-
stration of nitrate which is based on the doubly refractive property
in polarized light of the insoluble salt formed by the interaction of
nitrate with Nitron (4,5-dihydro-l,4-diphenyl-3,5-phenylimino-l,2,4-
triazole). Busch (1905) originally employed this reaction for the
gravimetric determination of nitric acid.

Cramer Method for Nitrate

SPECIAL REAGENTS

Nitron Reagent. 10% Nitron in 5% acetic acid.

36 MICROSCOPIC TECHNIQUES

PROCEDURE

1. Prepare frozen sections of fresh tissue.

2. Place 1-2 drops of hot Nitron reagent on a cover slip, and
place the cover slip over the section on a glass slide so that the tissue
is bathed in the liquid.

3. Put in a refrigerator for 30 min. to aid in the crystallization of
the nitrate.

4. Examine with polarized light under a microscope immediately
on removal from refrigerator.

Result. The doubly refractive zones are due to the insoluble
nitrate.

SULFHYDRYL AND DISULFIDE GROUPS

The earlier literature dealing with the application to tissue sec-
tions of the nitroprusside reaction for sulfhydryl groups, and the
reduction of disulfide compounds to give this test, has been
reviewed by Lison (1936, pages 133-136). The procedure given by
Pv,apkine and recommended by Lison, as well as the methods of
Bourne (1935) and of Hammett and Chapman (1938-1939), will
be given. The latter investigators critically examined the nitro-
prusside test and concluded that it should not be considered a
quantitative reaction; they established a well-defined procedure
which they believe most likely to yield satisfactory qualitative
results. However, the problem of diffusibility will no doubt limit,
or eliminate, the use of any nitroprusside method.

Nitroprusside Test of Rapkine

SPECIAL REAGENTS

5% Sodium Nitroprusside. For plant tissues.

2% Sodium Nitroprusside. For animal tissues.

Ammonium Sidjate Crystals.

Cone. Ammonium Hydroxide.

10% Potassium Cyanide.

10% Trichloroacetic Acid.

5% Zinc Acetate.

PROCEDURE

A. For free sulfhydryl groups

SULFHYDRYL AND DISULFIDE GROUPS 37

1. Immerse the fresh tissue in the zinc acetate soln. for a few sec.
This will stabilize the red color finally developed, as shown by
Giroud and Bulliard (1933).

2. Add 1 drop of the sodium nitroprusside soln. to a section
on a slide.

3. Add a crystal of ammonium sulfate and a drop of ammonium
hydroxide.

Result. Sulfhydryl compounds such as glutathione produce a
red color.

B. For total sulfhydryl groups

1. Treat sections of fresh tissue with 10% potassium cyanide
for 5-10 min.
2.-4. Proceed with steps 1-3 in A.

C. For protein-bound sulfhydryl groups

1. Treat sections of fresh tissue with 10% trichloroacetic acid
for 15 min. and wash thoroughly in water.
2.-4. Proceed with steps 1-3 in A.

note: The diffusibility of the sulfhydryl compounds formed in B, or liberated
in C, makes for particular unreUability in the localization of the groups in
the sections.

Bourne Nitroprusside Test

SPECIAL REAGENTS

5% Acetic Acid.

5% Sodium Nitroprusside Saturated with Ammonium Sidfate.

Concentrated Ammonium Hydroxide.

PROCEDURE

1. Place fresh frozen sections of tissue in hot 5% acetic acid
for 30-90 sec.

2. Pour off the acid and replace with nitroprusside-ammonium
sulfate soln. for 2 min.

3. Add a few drops of ammonium hydroxide and examine at once.
Result. A purplish-blue color indicates sulfhydryl compounds.

Hammett and Chapman Nitroprusside Test

SPECIAL REAGENTS

27-29% Ammonium Hydroxide.

38 MICROSCOPIC TECHNIQUES

1 % Sodium Nitroprusside.
Ammonium Sulfate Crystals.

PROCEDURE

1. Cover fresh tissue slice with 0.25 niL water.

2. Add 0.05 ml. ammonium hydroxide and then 0.05 ml. of the
nitroprusside soln.

3. Underline the tissue with 0.25 g. ammonium sulfate crystals
and examine at once.

C. ORGANIC SUBSTANCES AND GROUPS

LIPIDS AND CHOLESTEROL*

By means of staining methods, it is impossible to distinguish
with certainty between the various chemical types of the lipids
with the possible exception of cholesterol and its esters. Until
recently, the demonstration of lipids in general was usually carried
out with Sudan dyes which dissolve in the lipids and color them.
However, Jackson (1944) reported an improved method using
acetic-carbol-Sudan III which he claims should supersede all other
Sudan methods since it will bring out lipids that have been con-
sidered refractory to Sudan staining in the past. Jackson's paper
includes an enlightening critical survey of previous work. To cir-
cumvent the loss of small fat globules from the tissue when alcohol
or acetone dye solutions are used, Telford Govan (1944) employed
Sudan dyes suspended in aqueous media. The Kay and Whitehead
( 1935) procedure using Sudan IV, the newer Jackson ( 1944) method
employing acetic-carbol-Sudan III, and the Telford Govan (1944)
technique will be described. The staining of lipids by means of
fluorescent dyes according to Popper (page 105) would appear to
have some advantages, particularly in the use of the water-soluble
dyes such as Phosphine 3R.

Cholesterol and its esters may be visualized by the Liebermann-
Burchardt reaction as adapted for histological use by Schultz ( 1924-
1925) , Romieu ( 1927) , and Yamasaki ( 1931) . The Schultz procedure
has been employed more generally, and hence it will be given in
detail. Lison (1936, page 210) has pointed out that, though the
positive test is specific, a negative result does not necessarily

* See Bibliography Appendix, Ref. 3.

LIPIDS AND CHOLESTEROL 39

exclude the presence of cholesterol or its esters. The Windaus
digitonin test for free cholesterol (Lison, 1936, pages 211-212)
requires further investigation in the opinion of Kay and Whitehead
in Lee's Vade Mecum (1937, page 281). By means of the polarizing
microscope, cholesterol crystals can occasionally be observed in
sections as birefringent rhombic plates. If the temperature is low-
enough, neutral fats and fatty acids can also be observed in some
instances as birefringent crystals.

Kay and Whitehead Procedure for Sudan IV
Stain for Lipids

SPECIAL REAGENTS

Stock Solution of Dye (can be used for at least 6 months). Prepare
a saturated alcoholic solution by boiling 2 g. dye in 1 1. absolute
alcohol; allow to cool.

Staining Solution (good for only about 4 hr. after being mixed).
Add slowly, with stirring, to 7 vol. stock soln., 9 vol. of 45%
alcohol. Filter after standing for 1 hr. The 45% alcohol is prepared
by mixing 4 vol. absolute alcohol with 5 vol. distilled water.

PROCEDURE

1. Place formalin-fixed frozen sections in 50% alcohol for 5 min.
in staining soln. for 30 min. at 37° (turn sections over after 15 min.
for more even staining), in 50% alcohol several sec, and finally in
distilled water a few min.

2. Pass through filtered hemalum and wash in alkaline tap water
for several min.

3. Mount in glycerin jelly.

Result. The lipid will be stained red. -^^

The sections should be stained the day after cutting since they
tend to be sticky for a while just after the cutting. On the other
hand, poor results are often encountered if the staining is delayed
longer than one day after sectioning due, presumably, to crystalli-
zation of lipid material. The stain lasts for only a few months.

Jackson Procedure for Lipids Using Acetic-Carbol-Sudan III

SPECIAL REAGENTS

Sudan III Stock Solution. Cover 2 g. of the finely pulverized dye

40 MICROSCOPIC TECHNIQUES

with 450 ml. of 95% alcohol and heat on a water bath to simmer-
ing. Stir occasionally and then filter while hot. Transfer to a
stoppered bottle and place in a refrigerator overnight. Filter while
cold, and add distilled water dropwise from a burette, while stir-
ring, to reduce the alcohol concentration to 80%. Allow to stand
24 hr. ; filter and keep stoppered.

Acetic-Carbol-Siidan III Reagent. To a given volume of Sudan III
stock soln., slowly add 5% phenol dropwise from a burette, stir-
ring after each addition, to bring the alcohol concentration to 60%
{e.g., add 2 ml. phenol soln. to 6 ml. stock soln.). Let stand for
several hr. keeping the bottle stoppered. Then, in the same drop-
wise manner, add glacial acetic acid in the proportion of 2.5 drops
per ml. carbol-Sudan soln. After standing for 24 hr. in a stoppered
bottle, filter the reagent. Do not use when the soln. is more than
several days old.

5% Glacial Acetic Acid in 50% Alcohol.

PROCEDURE

1. Transfer formalin-fixed frozen sections to 50% alcohol for
1 min.

2. Place in the acetic-carbol-Sudan III reagent for 1.5 hr. or
longer; keep vessel stoppered.

3. Differentiate in the acetic-alcohol soln. for 10-60 sec. and
wash in distilled water for 1 min. In some cases it may be well to
dilute the acetic-alcohol with more 50% alcohol.

4. A counterstain of recently filtered Delafield hematoxylin
diluted with 2 vol. distilled water may be applied for 15 min.,
followed by differentiation in 0.5% hydrochloric acid until reddish
(10-20 sec), and treatment with very dilute ammonium hydroxide
for 5 min. to develop the blue color.

5. Wash in distilled water and mount in glycerin jelly.

Result. The lipid will be stained red.

Telford Govan Technique for Sudan Dye Staining
in Aqueous Media

SPECIAL REAGENTS

Sudan Dye Suspension. Add a saturated soln. of a Sudan dye in
acetone dropwise from a capillary pipette to 1% gelatine soln.

LIPIDS AND CHOLESTEROL 41

containing 1% acetic acid until a deep brick-red color and a
consistency of milk is obtained. Stir well during the addition.
Hold the mixture at 37° for 2 hr. to evaporate the acetone or let
stand in a warm room overnight. Filter off sediment through coarse
paper.
1 % Gelatin Solution.

PROCEDURE

1. Transfer frozen sections from water to 1% gelatin soln. for
2-3 min.

2. Stain for 30 min. at 37° with the suspension.

3. Wash sections in 1 % gelatin soln. for 2-3 min.

4. Wash well in water.

5. Counterstain, if desired, and mount in glycerin jelly or Karo
syrup.

Schultz Cholesterol Test

SPECIAL REAGENTS

2.5% Iron Alum (NH4)2S04.Fe2(S04).-5.24H20.

Concentrated Sulj uric-Glacial Acetic Acid Mixture. Add the sul-
furic acid slowly to an equal volume of glacial acetic acid, stirring
and cooling the while. (Only the purest acids are suitable and
the sulfuric must contain at least 98% sulfuric acid. The reagent
is hydroscopic and must be protected from atmospheric moisture.)

PROCEDURE

1. Place formalin-fixed frozen sections in the iron alum solution
for 3 days at 37°.

2. Rinse in distilled water, mount on slides and blot with filter
paper.

3. Add a few drops of the acid mixture and cover with a cover-
glass.

Result. A positive test for cholesterol, or its esters, is indicated
by the appearance of a blue-green color which reaches its maximum
intensity within a few min. Within 30 min. the sections acquire a
brown discoloration. The appearance of large numbers of bubbles
results from impure acids.

42 MICROSCOPIC TECHNIQUES

CAROTENE, CAROTENOIDS, AND VITAMIN A

The solubility of carotene, carotenoids, and vitamin A in organic
solvents makes it necessary to employ frozen sections of tissue for
histological studies on these substances. Unstained sections show
yellow, orange, or brown regions due to the presence of constituents
of this nature. The blue coloration given by concentrated sulfuric
acid with these compounds has been employed by Steiger (1941)
for the demonstration of carotene in leaves. The deep violet color
developed in the presence of aqueous 1% iodine in 7% potassium
iodide (Lison, 1936, page 245) is also characteristic of these
polyenes, and when treated with oxidizing agents, such as chromic
acid, they are bleached. Bourne (1935) adapted the Carr-Price
reaction to tissue sections by placing frozen sections directly into
a chloroform solution of antimony trichloride. It is well known
that the blue color due to vitamin A fades very rapidly, while that
due to carotene persists. As shown by Raoul and Meunier (1939),
sterols produce a red color in the Carr-Price test. The detection uf
vitamin A in tissue by fluorescence is described on page 104.

Steiger Method for Carotene in Leaves
SPECIAL REAGENTS

Alkali-Alcohol Mixture. Combine 1 vol. saturated potassium hy-
droxide with 2 vol. of 40% alcohol and 3 vol. tap water.
Concentrated Sulfuric Acid.

PROCEDURE

1. Place green leaves in the alkali-alcohol mixture in a wide-
mouth bottle and seal the glass stopper with vaseline.

2. After several days in the dark, when the fluid is green and
the tissue yellow, transfer to distilled water for several hr.

3. Place small pieces of tissue on a slide and dry with filter paper.

4. Add 1 drop cone, sulfuric acid.

Result. Carotene is indicated by the appearance of dark blue
crystals visible under the microscope. Grossly, a green color chang-
ing to blue can be observed.

RIBOFLAVIN AND POLYSACCHARIDES 43

RIBOFLAVIN

The detection of riboflavin in tissue sections, based on reduction
of the vitamin in acid medium to leucoflavin and reoxidation to
bright red granules of rhodoflavin, was employed by Chevremont
and Comhaire ( 1939) . Riboflavin can be recognized in tissue by its
characteristic greenish-yellow fluorescence when irradiated with
ultraviolet (page 104).

Chevremont and Comhaire Method for Rihoflavin

SPECIAL REAGENTS

Fixative. Formol-Nitron or formol-basic lead acetate soln.
Reductant Solution. Add zinc to hydrochloric acid to generate

hydrogen.
Oxidant Solution. Dilute hydrogen peroxide.

PROCEDURE

1. Fix tissue for 5 days and prepare sections.

2. Treat sections for 30 min. with reductant soln.

3. Ptinse in water and treat with oxidant soln.

4. Examine under microscope.

Result. Bright red granules indicate presence of riboflavin. The
localizations are probably unreliable due to the diffusibility of the
riboflavin and its derivatives.

POLYSACCHARIDES IN GENERAL

A general reaction for the microscopic visualization of polysac-
charides has been described by Hotchkiss (1946).* The reaction
involves the oxidation of adjacent hydroxyl groups to aldehydes by
means of periodate, and the coloring of the aldehyde with Feulgen
reagent. The oxidation takes place according to the equation:

— CHOH— CHOH— + HsIOe > — CHOHCO— + HIO3 + 3 H2O

The chief substances in plant tissues that show the stain are
starches, cellulose, hemicelluloses, and pectins and, in animal tissues,
glycogen, mucin, mucoproteins, and presumably hyaluronic acid
and chitin. The pentoses of nucleic acid are so substituted that they

* Subsequent to this writing the author learned of the paper of McManus
(1946) (Bibliography Appendix, Ref. 13) in which the same principal was
independently presented.

44 MICROSCOPIC TECHNIQUES

will not give the reaction and cerebrosides, if present, would be
expected to react.

Method of Hotchkiss for Polysaccharides

SPECIAL REAGENTS

Periodic Acid Solution A. Dissolve 400 mg. periodic acid (H5IO6,
obtainable from G. Frederick Smith Chemical Co.) in 10 ml. dis-
tilled water, add 5 ml. M/5 sodium acetate and 35 ml. alcohol.
^Periodic Acid Solution B. Dissolve 400 mg. periodic acid in 45 ml.
distilled water and add 5 ml. ilf/5 sodium acetate.

lodide-Thiosidfate Solution. Dissolve 1 g. potassium iodide and

1 g. sodium thiosulfate (Na2S203.5H20) in 20 ml. distilled water
and add with stirring 30 ml. alcohol followed by 0.5 ml. 2 N
hydrochloric acid. A sulfur precipitate forms and settles out
slowly although the soln. may be used immediately.

Feulgen Reagent. See page 67.

Sulfite Wash Solution. Add 0.5 ml. cone, hydrochloric acid and

2 ml. 10% potassium metabisulfite to 50 ml. distilled water.

PROCEDURE

1. If water-soluble polysaccharides are to be stained, fix the
tissue in a dehydrating soln. such as Carnoy fluid (page 45),
Otherwise, use any of the usual fixatives. After fixation remove
traces of mercury, if present, with iodine and be sure any formalde-
hyde is completely removed by washing. (70% alcohol may be used
as a washing soln. if it is desired to avoid removal of water-soluble
polysaccharides.)

2. Place section or smear in the periodic acid soln. A or B
(depending on whether an alcoholic or aqueous soln. is desired)
for 5 min.

3. Flood with 70% alcohol (or water) and transfer to the iodide-
thiosulfate soln. for 5 min.

4. Again wash with 70% alcohol (or water) and then transfer
to the Feulgen reagent for 15-45 min.

5. Rinse with the sulfite wash soln., dehydrate, and mount as
usual. ^

^ Result. Polysaccharides are indicated by the violet fuchsin color.

note: If free tissue aldehydes are present it is necessary to remove them
first as on page 93.

ACID POLYSACCHARIDES 45

If the periodate and iodate are not washed out they will give rise to a
brownish coloration in the Feulgen reagent.

Control sections or smears may be made by placing in 70% alcohol (or
water), instead of in the periodic acid soln. A or B, and then carrying
through the remaining steps without change.

Egg albumin adhesive may take a slight stain due to the carbohydrate
content of this material.

If neutral polysaccharides are to be stained, a counterstain with a basic
dye (such as 0.02 mg. malachite green per ml. water) may be used. If
mucin or acid polysaccharides are to be stained, the counterstain should be
an acid dye.

ACID POLYSACCHARIDES — HYALURONIC ACID*

For the demonstration of acid polysaccharides of the hyaluronic
acid type Hale (1946) employed fixation in a dehydrating medium
to prevent solution of the water-soluble acid polysaccharide, com-
bination of the latter with iron, and demonstration of the iron by
means of the Prussian blue reaction. The iron will not combine with
neutral polysaccharides or proteins according to Hale. In order to
differentiate between hyaluronic acid and other substances which
might give the blue stain, Hale suggests that hyaluronidase be
used to digest away the hyaluronic acid in the section on one slide
and a comparison be made to an undigested section.

Hale Method for Acid Polysaccharides

SPECIAL REAGENTS

Acetic-Iron Solution. Mix equal vol. dialyzed ferric hydroxide,
concentration not stated (Hale used the product of the British
Drug Houses Ltd.), and 2 M acetic acid.

0.02 M Potassium Ferrocyanide in 0.14 M Hydrochloric Acid.

PROCEDURE

1. Fix 3-4 mm. pieces of tissue in Carnoy fluid (6 vol. absolute
alcohol, 3 vol. chloroform, and 1 vol. glacial acetic acid) for 0.5 hr.

2. Treat with absolute alcohol; clear, and mount in paraffin.

3. Prepare sections and place on slides without albumin adhesive.

4. Flood the deparaffinized sections with the acetic-iron soln.,
and after 10 min. wash well with distilled water.

5. Treat the sections with the ferrocyanide soln. for 10 min. and

* See page 46 for the staining of mucoproteins.

46 MICROSCOPIC TECHNIQUES

wash with water and counterstain if desired (it is well to use a red
eounterstain such as f uchsin) .

6. Rapidly dehydrate, clear in xylol, and mount in Canada
balsam.

Result. Acid polysaccharide is indicated by the blue color.

MUCOPROTEINS*

Toluidene blue will stain quite a variety of acid substances, but
the metachromatic staining by this dye of mucoid compounds
containing polysaccharide esters of sulfuric acid is specific for
these compovmds, provided that the method of Lison (1935) is
strictly adhered to (Sylven, 1941, 1945). Sylven (1941) has made a
thorough study of the staining and he emphasized that "false"
metachromatic staining can be obviated by the prompt removal, of
water by alcohol after the staining, the alcohol assuring a ''true"
reaction which is the red stain characteristic of, and specific for,
the polysaccharide sulfates in tissue. Holmgren and Wilander ( 1937)
found that basic lead acetate solution was a superior fixative for
tissues to be subjected to the metachromatic toluidene blue stain,
but Sylven now employs a mixture of this fixative with formalin to
reduce the time required for the fixation. The staining time can be
reduced by aging the dye solution, and the greater the alcohol
concentration in the dye solution the paler the resulting stain will
be. In order to bring out mast cell granules properly, the dye is
made up in alcohol of a concentration of 30% or higher (Sylven).

According to the claim of Hempelmann ( 1940) , chondroitin and
mucoitin sulfuric acid proteins can be differentiated from one
another in histological preparations by means of the toluidene blue
stain. In a dilution of 1 : 1,280,000 an aqueous solution of toluidene
blue is supposed to stain the chondroitin material in paraffin sections
a violet-red color, while the mucoitin protein complex remains un-
stained. Differentiation is also claimed when the dye is used in a
1 : 410,000 dilution in a solution of 10 vol. alcohol and 45 vol. water.
The alcohol concentration is stated to be critical, presumably both
mucoproteins will be stained if the proportion of alcohol is less,
and neither if it is greater. No confirmation of these claims has been
made; in fact, to the writer's knowledge several attempts to do so
have failed.

* See Bibliography Appendix, Refs. 8 and 10.

MUCOPROTEINS, GLYCOGEN, AND MUCIN 47

A fundamental study of metachromasy of basic dyes has been
published by Miehaelis and Granick (1945).

See page 45 for another method of staining acid polysaccharides,
and page 50 for the staining of mucin.

Lison Method for Polysaccharide Sulfate Compounds
(after Sylven)

SPECIAL REAGENTS

Fixing Solution. Mix equal vol. of 8% basic lead acetate soln.

and 14-16% formalin.
Toluidene Blue Solution. Prepare separately (a) 0.1% dye in 1%

alcohol and (6) 0.1% dye in 30% alcohol, and let stand for a

number of days to age.

PROCEDURE

1. Fix the tissue for 12-24 hr. in the fixing soln.

2. Prepare paraffin sections in the usual manner.

3. Stain the sections for 30 min. using soln. a on some, and soln.
b on others. Soln. a gives a more intense stain.

4. Wash well in alcohol briefly, immediately after removing from
the dye soln.

5. Mount in natural cedar oil.

GLYCOGEN AND MUCIN*

A critical comparison of the iodine, Best carmine, and Bauer-
Feulgen methods for demonstrating glycogen microscopically was
,made by Bensley (1939), who concluded that the Bauer-Feulgen
method, which depends on the reaction of the aldehyde groups in
the carbohydrate with the reagent, is by far the best if the tissue
is promptly fixed in alcohol-formalin solution. When chrome salts
are present in the fixative, the visualization of intracellular glycogen
was found to require the Best carmine stain since the Bauer-Feulgen
method is not specific in those cases, and the iodine technique is
not suited for high-power studies. A procedure for preparing paraffin
sections for the carmine stain was given by Mullen (1944), who
employed celloidin to hold the deparaffinized sections to the glass
slide.
Mitchell and Wislocki ( 1944) reported that the ammoniacal silver

* See Bibliography Appendix, Refs. 2 and 12.

48 MICROSCOPIC TECHNIQUES

nitrate method which Pap (1929) employed for the staining of
reticulum visualized glycogen more intensely and consistently than
either the Best carmine or Bauer-Feulgen procedures. The admitted
drawback to this method is the fact that since reticulum fibers of
connective tissue are also stained it cannot be applied to this tissue.
However, the authors feel that in other cases the ammoniacal silver
nitrate method has advantages over those previously employed.
Gomori (1946) subsequently modified this procedure and developed
a more selective method which demonstrates glycogen and mucin,
but eliminates possible interference by desoxyribonucleic acid, uric
acid, and granules of enterochromaffin cells, all of which can reduce
silver solutions under certain conditions. However, melanin is stained,
and except for this the method enables the same localizations of the
reducing substances as the Bauer-Feulgen procedure. Should in-
soluble calcium salts be present they too will stain black. A prelim-
inary 10 min. treatment of the sections with citrate buffer of pH
3-4 will remove calcium deposits.

The differentiation of glycogen from other substances that give
positive reactions may be made in some instances by employing
saliva to digest away the glycogen selectively. See page 46 foi* the
staining of mucoproteins.

Bauer-Feulgen Stain for Glycogen (after Bensley)

SPECIAL REAGENTS

Alcohol-Formalin Fixative. 9 vol. absolute alcohol plus 1 vol.
neutral formalin. The alcohol may be first saturated with picric
acid.

Feulgen Reagent* Heat to dissolve 1 g. basic fuchsin in 100 ml.
distilled water. Filter while warm, cool, add 20 ml. 1 A^" hydro-
chloric acid and 1 g. sodium bisulfite. Let stand 24 hr. The soln.
should be straw colored.

1% or 4% Chromic Acid.

Bisulfite Rinsing Solution. 1 vol., 1 M sodium bisulfite plus 19
vol. tap water.

PROCEDURE

1. Fix very small pieces (2-3 mm.) of fresh tissue in the alcohol-

*See pages 65 and 67 for other methods of preparing this reagent.

r

GLYCOGEN AND MUCIN 49

formol solution for 24 hr. (Deane, Nesbett, and Hastings, 1946,
recommend the use of ice-cold alcohol-picric acid-formalin to pre-
serve the glycogen throughout the tissue block.) Wash in absolute
alcohol and embed in paraffin, being careful to prevent overheating.
(It is essential that very fresh tissue be used since glycogen is
rapidly autoiyzed. The Altmann-Gersh freezing and drying tech-
nique for fixation may also be used; in fact it can lead to a truer
picture of the glycogen distribution, as shown by Bensley and
Gersh, 1933a) .

2. Section, mount on slides, and deparaffinize as usual.

3. Place in 4% chromic acid 1 hr. or in the 1% soln. overnight.

4. Wash in running water for 5 min., place in Feulgen reagent
10-15 min., rinse with three changes of bisulfite soln. for 1.5 min. in
each change, and wash in running water for 10 min.

5. Nuclei may be counterstained with hematoxylin.

6. Dehydrate, clear, and mount in balsam.

7. As a negative control, remove the glycogen from some of the
sections, brought down to water, by adding fresh saliva. During a
15-30 min. period, change the saliva several times. Wash with water
at 37° to remove mucus and stain as above beginning with step 3.
Comparison of these sections with those not given the saliva treat-
ment helps to distinguish the glycogen regions.

Result. The glycogen appears deep red-violet, the nuclei laven-
der.

Best Carmine Stain for Glycogen (after Bensley)

SPECIAL REAGENTS

Alcohol-Formalin Fixative. Same as the reagent for Bauer-Feulgen
stain.

Carmine Stain Stock Solution. Gently boil 2 g. carmine, 1 g. potas-
sium carbonate, and 5 g. potassium chloride in 60 ml. distilled
water until color darkens. After cooling, add 20 ml. cone, am-
monia and let stand 24 hr. This solution may deteriorate in a
month in a warm room ; keep well stoppered.

Fresh Carmine Stain. 10 ml. stock soln., 15 ml. cone, ammonia, and
30 ml. methanol (C.P.).

50 ' MICROSCOPIC TECHNIQUES

PROCEDURE

1. Follow steps 1 and 2 for Bauer-Feulgen stain.

2. After bringing down to distilled water, stain nuclei with
hematoxylin.

3. Transfer to fresh carmine stain and after 20 min. wash in
three changes of methanol, dehydrate in acetone, clear in toluol,
and mount in balsam.

4. Run negative control sections as in step 7 for the Bauer-
Feulgen method but apply the carmine stain above.

Result. The glycogen will appear brilliantly red.

Gomori Procedure for Glycogen and Mucin

SPECIAL REAGENTS

Fixative. One of the alcohol fixatives such as alcohol-picric acid-
formalin (page 48) for glycogen. Any routine fixative for mucin.

0.5% Collodion in Alcohol-Ether Solution.

5% Chromic Acid.

1-2% Sodium Bisulfite.

Silver-Methenamine Stock Solution. Add 5 ml. 5% silver nitrate
soln. to 100 ml. 3% methenamine ( hexamethylenetetramine ) soln.
Shake until the initial heavy white precipitate disappears, and
store in refrigerator.

Alkalized Silver-Methenamine Solution. To 25 ml. silver-methena-
mine stock soln. add 25 ml. distilled water and 1-2 ml. 5% borax
(Na2B4O7.10H2O).

0.1 % Gold Chloride.

2—3% Sodium Hyposulflte.

PROCEDURE

1. Fix tissue and prepare paraffin sections as usual.

2. Run sections through xylol, alcohols, and water. (For glyco-
gen, protect sections on slides by dipping into collodion soln. before
transferring to the final alcohol soln.)

3. Place slides in 5% chromic acid for 1 — 1.5 hr.

4. Wash in running tap water for 10 min. and treat with the
bisulfite soln. for 1 min. to remove remaining traces of chromic acid.

5. Wash in running tap water for 5 min., rinse in distilled water,
and incubate at 37-45° in the alkalized silver-methenamine soln.
Examine sections under microscope every 15 min. Staining is com-

GLYCOGEN, MUCIN, AND STARCH 51

pleted when the glycogen 'and mucin appear deep brown or black.
The background will be yellowish. Usually 1-3 hr. is required.

6. Rinse well in repeated changes of distilled water and tone in
the gold chloride soln. for 5 min.

7. Rinse in distilled water and then in the hyposulfite soln. to
remove unreacted silver.

8. Wash in tap water and counterstain if desired.

9. Mount as usual.

Result. Glycogen and mucin will appear in shades from grey-
brown to black on an unstained background. Sometimes the collodion
film becomes stained and it can be removed by acetone or alcohol-
ether.

STARCH

The common practice of employing a dilute iodine solution to
develop a blue color with starch can be applied to sections of plant
material, as can the crystal violet stain followed by washing with
saturated picric acid solution. The use of formaldehyde as a swelling
agent to obtain special effects with safranine and fast green
FCF was described by Bates (1942). Starch granules can also be
recognized by the characteristic black crosses they exhibit due to
their doubly refractive pioperties when viewed under the micro-
scope with polarized light.

The following procedure of Milovidov (1928) is well suited for
the preparation of permanently mounted sections stained for starch.

Milovidov Method for Starch

SPECIAL REAGENTS

Aniline Fuchsin Stain.

5% Alcoholic Aurantia.

2% Tannin.

1 % Toluidine Blue, Gentian Violet, or Methyl Green.

PROCEDURE

1. Fix plant tissue in Regaud fluid and prepare sections as usual.

2. Stain sections with aniline fuchsin for 5 min. and differentiate
in the aurantia soln.

52 MICROSCOPIC TECHNIQUES

3. After washing sections in distilled water, mordant for 20 rain,
in the tannin soln. and again wash.

4. Stain sections in either toluidine blue, gentian violet, or
methyl green for 5-10 min.

5. Differentiate in 95% alcohol, dehydrate in absolute alcohol,
clear in xylol, and mount in balsam.

Result. The starch will appear as either blue, violet, or green
granules depending on which of the stains was used in step 4. The
mitochondria will appear red.

CELLULOSE

Post and Laudermilk (1942) Iodine Stain for Cellulose

SPECIAL REAGENTS

Iodine Solution. 20 ml. 2% iodine in 5% potassium iodide, 180

ml. distilled water, and 0.5 ml. glycerol.
Lithium Chloride Solution. Saturate 15 ml. distilled water at 80°,

cool; use supernatant soln.

PROCEDURE

1. Tease out sections or fibers.

2. Apply 2-3 drops of the iodine soln. and after 10 sec. blot
with filter paper and dry.

3. Add a drop of the lithium chloride soln., cover, and examine.

Result. Cellulose appears in the following colors depending on
its source :

Typical color Fiber

Light blue cotton, soda pulp, bleached sulfite, straw,

esparto

Dark blue pineapple fiber

Greenish blue linen

Green to yellowish green ...sisal, Manila hemp, yucca

Yellow yucca, ground wood, hemp, Manila hemp

Lemon yellow kapok

Brownish yellow \nio

CHITIN*

The horny carbohydrate material, chitin, requires special treat-
ment to soften it sufficiently for the preparation of paraffin sections.

* See Bibliography Appendix, Ref . 16.

CELLULOSE AND CHITIN 53

The most recent method for this treatment is that of Murray ( 1937) ,
but the Diaphanol technique has been widely employed. Once sec-
tions are prepared they may be stained by the procedure of Zander,
Schulze, or Bethe given in Lee ( 1937, page 600j .

Murray Method for Softening Chitin

SPECIAL REAGENTS

Formalin- Saline Fixative. 10% formalin in 0.8% sodium chloride
soln.

Dehydrating Fixative. Equal vol. absolute alcohol, chloroform,
and glacial acetic acid to which mercuric chloride is added to
saturation ( about 4% ) .

Chloral Hydrate-Phenol Reagent. Equal weights of chloral hy-
drate and phenol warmed together until they blend to an oily
liquid that is fluid at room temperature.

PROCEDURE

1. Fix material in the formalin-saline soln.

2. Transfer to the dehydrating fixative.

3. Place specimen in the chloral hydrate-phenol reagent for 12-
24 hr. or longer.

4. Clear with xylol, chloroform, or carbon disufide and imbed
in paraflfin.

Diaphanol Method for Softening Chitin

SPECIAL REAGENTS

Diaphanol Solution. Pass vapors of chlorine dioxide into ice-cold
50% acetic acid. Store in a cool dark place in a glass-stoppered
bottle. (Before the war, Diaphanol was sold by Leitz; and Lee —
1937, page 598 — recommends buying, rather than preparing, the
soln. However, it will probably be impossible to buy for some
time, and there is no reason why the reagent cannot be safely
prepared if the obvious precautions of working in a hood, etc.,
are taken.)

PROCEDURE

1. Fixed material is rinsed in 63% alcohol and placed in Di-
aphanol in a glass-stoppered bottle in diffuse daylight until bleached

54 MICROSCOPIC TECHNIQUES

and softened. The specimen should be pierced or cut to allow escape
of carbon dioxide.

2. If the Diaphanol becomes discolored, transfer to a fresh por-
tion of the soln.

3. Place in 63% alcohol until hardened and then pass through
tetralin into paraffin.

Methods for Staining Chitin

The softened material, or sections of it, may be tested for chitin
by a variety of color reactions. Zander treated for a short time with
a drop of fresh iodine in potassium iodide soln., followed by a drop
of strong zinc chloride soln. Upon removal of the reagents with
water, a violet color is obtained in the presence of chitin. Schulze
divided the material into two portions. One was subjected to the
procedure of Zander and the other was treated with iodine and then
cone, sulfuric acid. The latter test serves to distinguish chitin from
cellulose and tunicin since chitin yields a brown color while the
others give a blue. Bethe employed freshly prepared 10% aniline
hydrochloride containing a drop of cone, hydrochloric acid for each
10 ml. After sections were placed in this soln. for 3-4 min., they
were rinsed with water and the slides were then placed, sections
downward, in a bath of 10% potassium dichromate. Chitin produces
a green coloration which becomes blue in tap water or ammoniacal
alcohol.

ASCORBIC ACID

The stain for ascorbic acid was developed in 1933 by Bourne,
who utilized the fact that reduced silver is deposited when ascorbic
acid in tissue interacts with acid silver nitrate. Bourne (1936)
published a critical survey of this stain and his recommended pro-
cedure is given below with a subsequent modification by Barnett
and Bourne (1941) designed to increase the specificity of the test
by dissolving precipitated silver salts in dilute ammonia. Giroud
and Leblond (1936) also investigated the technique and its appli-
cations, and in reply to criticisms of the specificity of the stain,
these authors ( 1937) point out that the positive test is specific for
ascorbic acid but a negative result does not necessarily mean that

ASCORBIC ACID 55

ascorbic acid is absent. Tonutti (1938) washed the tissue in 5.4%
levulose solution before staining in order to remove blood. The
reliability of the localizations obtained with the silver stain remains
to be proved, according to Danielli ( 1946a) . It would be necessary
to establish that the ascorbic acid is attached to a nondiffusible
body and that the reaction product could not diffuse, or that the
ascorbic acid site has a high affinity for the reaction product.

Bourne Silver Stain for Ascorbic Acid

/. Reduced Ascorbic Acid
SPECIAL REAGENTS

Acid Silver Nitrate. Add 5 ml. glacial acetic acid to 100 ml. 5%

silver nitrate.
5% Ammonium Hydroxide.

PROCEDURE

1. Place frozen sections of fresh tissue in the acid silver nitrate
soln. for a few minutes, and then treat with 5% ammonium
hydroxide.

2. Wash with distilled water. If desired, lipids can be then
stained with a Sudan dye in 90% alcohol.

3. After clearing, mount in glycerin.

Result. Granules containing reduced ascorbic acid appear black.

11. Reduced and Oxidized Ascorbic Acid
PROCEDURE

1. Expose the fresh tissue to the vapor of glacial acetic acid for
several minutes.

2. Cut into thin pieces and subject to an atmosphere of hydrogen
sulfide for 15 min. in order to reduce the oxidized form.

3. Remove hydrogen sulfide by placing in a vacuum for 10-30
min. followed by a good stream of nitrogen gas for 15 min.

4. Treat with acid silver nitrate soln. followed by ammonium
hydroxide as above.

Should glutathione be present in quantities sufficient to inhibit the
test, wash the tissue momentarily after the hydrogen sulfide treat-
ment and immerse at once into a mercuric acetate soln. for a few
minutes. After washing, apply the acid silver nitrate solution and
then the ammonium hydroxide.

56 MICROSCOPIC TECHNIQUES

PROTEIN REACTIONS

Many of the tests for proteins are poorly adapted to histochemical
work because the strong acid or alkaU that they require has too
great a disintegrative eifect on the cellular structure. Tests which
particularly fall into this group are the biuret reaction for com-
ponent peptides, the xanthoproteic reaction for phenolic constitu-
ents, Millon's tyrosine reaction, Romieu's tryptophane test, the
tryptophane reaction of Voisenet-Flirth, and the diazo reaction for
histidine and tyrosine. Serra's arginine test, which is much less
drastic, and Berg's ninhydrin reaction for a-amino acid groups,
which uses no corrosive reagents although heating is required, will
both be described as well as two of the previous group, Millon and
Romieu reactions.

ARGININE AND ARGININE-CONTAINING PROTEINS

In another of those coincidences that occasionally turn up, Serra
at the University of Coimbra, Portugal, and Thomas at the Univer-
sity of Missouri, independently, and without knowledge of the other's
work, adapted the Sakaguchi (1925) reaction for arginine to its
histochemical identification. The reaction is based on the develop-
ment of an orange-red color with arginine when a-naphthol and
hypobromite or hypochlorite react with it in an alkaline medium.

The first description of the method by Serra (1944a,b) was fol-
lowed by a report of Serra and Queiroz Lopes ( 1944) , who empha-
sized the usefulness of the reaction for the visualization of the basic
proteins such as those contained in cell nuclei. Subsequently, Serra
(1946) summarized the work of his group on the arginine reaction
in the course of a more general article dealing with histochemical
tests for proteins and amino acids. Serra pointed out that a positive
reaction is found only with arginine and the rather rare compounds
glycocyamine, gelegine, and agmatine, negative reactions being given
by guanidine, urea, ornithine, creatine, creatinine, asparagine, histi-
dine, and other amino acids. The reaction is specific for guanidine
derivatives in which one hydrogen atom of one or both amino groups
is substituted by an alkyl, fatty acid, or cyano radical. Substitution
of other radicals has not been tested, while guanidine derivatives in
which both hydrogen atoms of one amino group are substituted do
not give the color (Thomas, 1946) .

ARGININE AND ARGININE-CONTAINING PROTEINS 57

The procedures of Serra and Thomas differ in certain details
and, at the date of this writing, no comparison of the two has been
made. An advantage of the method of Thomas is that it does not
employ cooling in an ice bath during the reaction because of the
substitution of hypochlorite for hypobromite. Serra mounts his
sections in glycerol after several transfers through this medium and
he has reported that in this fashion the color, otherwise stable for
only a short time, is stabilized for months.

Serra Method for Arginine and Arginine-Containing Proteins

SPECIAL REAGENTS

Acetic- Alcohol-Formalin Fixative. Add a few drops of glacial

acetic acid to each 10 ml. of a mixture of 2 vol. 96% alcohol and

1 vol. formalin.
1% a-Naphwl in 96% Alcohol. Store in a refrigerator. Dilute

1:10 with 40% alcohol before use.
4% Sodium Hydroxide.
2% Sodium Hypobromite. With stirring and cooling, add 2 g. or

approximately 0.7 ml. bromine to 100 ml. 5% sodium hydroxide.

Store in a refrigerator.
40% Urea.

PROCEDURE

1. Fix the material in the acetic-alcohol-formalin mixture. Wash
well in water.

2. Transfer to a watch glass kept at 0-5° in an ice bath, and
treat for 15 min. at this temperature with a mixture of 0.5 ml. a-
naphthol soln., 0.5 ml. 1 N sodium hydroxide, and 0.2 ml. 40% urea.

3. Add 2 ml. 2% hypobromite, and after 3 mih. stir in 0.2 ml.
urea soln. and then 0.2 ml. of the hypobromite. The maximum color
develops in 3-5 min.; intensify it by a subsequent treatment with
the hypobromite for 3 min.

4. Pass through four glycerol baths, leaving for 2-3 min. in
each. In glycerol the color is stable for months even at room tem-
perature. The fading is inhibited by storage in the cold.

Result. An orange-red color characterizes a positive reaction.

58 MICROSCOPIC TECHNIQUES

Thomas Method for Arginine and Arginine-Containing
Proteins

SPECIAL REAGENTS

0.1% a-Naphthol in 10% (by vol.) Ethyl Alcohol.

0.15 N Sodium Hyp/ochlorite in 0.05 N Sodium Hydroxide. For
preparation of the hypochlorite see page 239; or prepare from
Clorox which is standardized and then stored at 3-5° in a dark
bottle. Dilute the hypochlorite to the proper strength each time
before use. Clorox is approximately 1.6 A^; standardize by adding
1 ml. Clorox to 5 ml. 1 A^ potassium iodide, 8 ml. cone, hydro-
chloric acid (sp. gr. 1.19), and 45 ml. water. Titrate with 0.1 N
sodium thiosulfate using starch indicator.

20% Urea in 0.05 N Sodium Hydroxide.

Tertiary Butyl Alcohol Solution. Add 1 ml. 5 N sodium hydroxide
and 19 ml. water to 80 ml. of the tertiary butyl alcohol. Shake
well and let stand; an aqueous layer collects on the bottom of the
vessel.

Pure Tertiary Butyl Alcohol.

Aniline.

Toluene.

PROCEDURE

1. Fix animal tissues in Bouin fluid (75 ml. saturated picric
acid soln., 25 ml. formalin, 5 ml. glacial acetic acid) . Onion root
tips were treated with medium chrome-acetic fixative.

2. Prepare paraffin sections in the usual manner. Do not remove
paraffin from sections until test is to be applied.

3. Place slide with sections in the a-naphthol soln. for at least
3 min.

4. Transfer to each of the following solns. in succession for the
periods indicated: hypochlorite, 20 sec; urea, 5 sec; 80% tertiary
butyl alcohol, 30 sec; pure tertiary butyl alcohol, 2 min.; aniline,
2 min.; toluene, 5 sec; and finally mount in Clarite. Use a stop watch
to time the immersions in each fluid.

TRYPTOPHANE IN PROTEINS

The red or violet color formed with proteins in the presence of
phosphoric acid is the basis of the Romieu reaction. Blanchetiere

TRYPTOPHANE AND TYROSINE 59

and Romieu (1931) presented evidence that the effect was the
result of tryptophane groups in the protein. As in the other protein
tests, the drastic nature of the reaction seriously interferes with its
use in most instances, as does the diffusibility of the color formed.

Romieu Reaction for Tryptophane in Proteins

SPECIAL REAGENTS

Syrupy Phosphoric Acid.

PROCEDURE

1. Fix tissue in alcohol, formalin, or Bouin fluid.

2. Prepare fairly thick paraffin or celloidin sections and remove
the infiltrating agent.

3. Place a drop of the phosphoric acid on a section and set in
an oven at 56° for a few min.

4. Examine on removal from oven.

Result. A positive test is manifest by the formation of a red or
violet color.

TYROSINE IN PROTEINS

Bensley's histochemical adaptation of well-known Millon reac-
tion for proteins containing tyrosine has been employed in studies
by Bensley and Gersh ( 1933b) .

Millon Reaction for Tyrosine in Proteins
(after Bensley and Gersh)

SPECIAL REAGENTS

Millon Reagent. Add 1 vol. 40% nitric acid (add 600 ml. distilled
water to 400 ml. cone, nitric acid, sp. gr. 1.42; let stand for 48
hr.) to 9 vol. distilled water and saturate with mercuric nitrate
crystals by frequent shaking over several days. Filter, and, to 400
ml. of filtrate, add 3 ml. 40% nitric acid and 1.4 g. sodium nitrite.

1 % Nitric Acid.

PROCEDURE

1. Mount sections on slides without using water. The freezing-
drying technique is preferable.

2. Place in cold Millon reagent. Since the maximum color is
developed in about 3 hr., remove each slide at a different time, dip

_(Ll$ii^ARY

60 MICROSCOPIC TECHNIQUES

immediately in 1% nitric acid and dehydrate rapidly in absolute
alcohol.

3. Clear in xylol and mount in balsam.

Result. A brick-red or rose color develops in the presence of
tyrosine or proteins containing tyrosine.

a-AMINO ACID GROUPS IN PROTEINS

Less soluble peptides and proteins containing a-amino acids may
be demonstrated at their loci in tissue sections by either the alloxan
or ninhydrin reactions. A tendency for the color to diffuse in the
alloxan reaction indicates that caution should be applied in inter-
preting the test, as Giroud (1929) has warned; furthermore the
specificity is not great enough to exclude the need for confirmatory
tests (Romieu, 1925). Hence only the ninhydrin reaction of Berg
( 1926) will be described. A positive reaction is obtained with many
amines, aldehydes, and ammonium compounds as well as with the
amino acids, but the solubility of these compounds enables their
easy removal as a rule.

Berg Ninhydrin Test for a-Aniino Acid Groups

SPECIAL REAGENTS

0.2% Ninhijdrin.

PROCEDURE

1. Fix tissue in 10% formalin.

2. Wash in water and prepare frozen sections.

3. Boil sections in 2 ml. of the ninhydrin soln. for 1 min.

4. Wash in water and mount in glycerin or glycerin jelly.

Result. a-Amino acid groups give rise to an intense blue or violet
color; this should be observed the same day as it fades rapidly.

note: Sena and Quieroz Lopes (1945) emploj^ed a mixture of equal vol.
of 0.4% ninlij'drin in distilled water and phosphate buffer of pH 6.98 (6 ml.
M/15 secondary sodium phosphate — 1L1876 g. Na2HP04.2Hi.O per liter —
and 4 ml. M lib primary potassium phosphate — 9.078 g. KHsPO* per liter).
They heat the sections in the liquid in a watch glass placed over a boiling
water bath. For cementing of the preparations they employ the mixture of
Romeis, which is 80 g. colophonium and 20 g. carefully heated lanolin.

MELANIN 61

MELANIN

Perhaps the most characteristic microchemical test for melanin
is its ability to reduce ammoniacal silver nitrate. Of course many-
other tissue constituents have this property so that the test is of
value only when possible interferences (page 48) are considered.
Dublin (1943) applied the Bodian silver method to the demonstra-
tion of melanin; his procedure follows.

Dublin Application of the Bodian Method to Demonstration
of Melanin

SPECIAL REAGENTS

Protargol Solution. Prepare fresh each time by adding one or more
drops of Protargol (Winthrop) — other brands do not appear to
be satisfactory — to water in a staining jar. The color should be
light amber. Do not stir or mix the solution since this results in
gumming.

1.0% Hydroquinone.

0.5% Auric Chloride.

5.0% Oxalic Acid.

10% Sodium Thiosulfate.

PROCEDURE

1. Fix tissue in 10% formalin.

2. Prepare paraffin sections 8 fx thick.

3. Treat the deparaffinized sections, after passing through graded
alcohols to water, with the Protargol soln. overnight.

4. Rinse with water and place in the hydroquinone soln. for 10
min.

5. Rinse with water and place in the auric chloride soln. for 5
min.

6. Rinse with water and place in the oxalic acid soln. for 5 min.

7. Rinse with water and place in the thiosulfate soln. for 5 min.

8. Wash in running tap water for 10 min.

9. Dehydrate, clear, and mount as usual.

Result. The melanin will appear black and the background a
purplish brown or gray.

62 MICROSCOPIC TECHNIQUES

HEMOGLOBIN

Of the many tests for hemoglobin in tissue, smears and blood cells
the more recent procedures of Ralph ( 1941) , Goulliart ( 1939, 1941) ,
and Dunn ( 1946) will be given. Previously Dunn and Thompson

(1945) had modified the Van Gieson stain, and later these authors

( 1946) adapted the patent blue method of Lison ( 1938) for the
staining of hemoglobin. The cyanol method of Dunn given below is
a simplification of the technique of Fautrez and Lambert ( 1937) .

Ralph Method for Hemoglobin

SPECIAL REAGENTS

Benzidine Reagent. 1% benzidine in absolute methanol.
Peroxide Reagent. 25% Superoxol in 70% ethanol.

PROCEDURE

1. Flood the dried blood or tissue smear on a glass slide with the
benzidine reagent for 1 min.

2. Drain off and flood the smear with the peroxide reagent for 1.5
min.

3. Wash in distilled water for 15 sec.

4. Dry and mount in neutral balsam.

Result. Hemoglobin will appear dark brown.

Goulliart Method for Hemoglobin

SPECIAL REAGENTS

Glacial Acetic Acid Containing a Few Crystals of Potassium Iodide.
Do not use after a week.

PROCEDURE

1. Treat a dried smear or frozen section on a slide with a drop of
reagent.

2. Examine after 30 min. with a polarizing microscope for groups
of very small boat-shaped birefringent crystals of protoiodoheme.
These crystals slowly change into square tabular Teichmann crys-
tals. The reaction may be speeded by warming.

HEMOGLOBIN, BILE PIGMENTS AND ACIDS 63

Dunn Method for Hemoglobin

SPECIAL REAGENTS

Cyanol Stock Solution. Dissolve 1 g. cyanol {National Aniline
Division, Allied Chemical and Dye Corp.) in 100 ml. distilled
water, add 10 g. pure zinc powder, and 2 ml. glacial acetic acid.
Bring mixtm^e to a boil and the blue color will soon fade out. The
soln. is stable for several weeks.

Cyanol Working Solution. Just before use filter 10 ml. of the stock
soln., add 2 ml. glacial acetic acid and 1 ml. commercial 3%
hydrogen peroxide.

PROCEDURE

1. Prepare frozen or paraffin sections of tissue fixed in 4%
formaldehyde buffered to pH 7.0.

2. Bring sections to water and stain in cyanol working solution
3-5 min.

3. Rinse in water and counterstain in safranin (1:1000 in 1%
acetic acid) 1 min.

4. Wash in water, dehydrate, clear, and mount in Clarite.

Result. Hemoglobin stains dark blue to bluish-gray; nuclei,
red; and cytoplasm, light pink.

BILE PIGMENTS AND ACIDS

The well-known Gmelin test has been adapted to the microscopic
detection of bile pigments by simply adding a drop of nitric acid
containing some nitrous acid to the sample on a slide. A positive
test is indicated by the appearance of a green color changing to red
and finally to blue. Stein's test ( 1935) , given below, is probably more
satisfactory. Bile salts and acids may be precipitated by barium and
the precipitate stained with acid fuchsin according to the technique
of Forsgren ( 1928) .

Stein Test for Bile Pigments

SPECIAL REAGENTS

Iodine Reagent. 2 or 3 vol. Lugol solution (6 g. potassium iodide
and 4 g. iodine dissolved in 100 ml. distilled water) plus 1 vol.
tincture of iodine.

5% Sodium Hyposidfite.

64 MICROSCOPIC TECHNIQUES

PROCEDURE

1. Fix for a short period in alcohol or 10% formalin.

2. Prepare paraffin sections and employ egg albumin to hold to
slides.

3. After removal of paraffin and bringing down to water, subject
sections to the iodine reagent for 6-12 hr.

4. Wash in distilled water and decolorize with the sodium hypo-
sulfite for 15-30 sec.

5. Wash in distilled water and stain with alum carmine for 1-3
hr.

6. Wash in distilled water, dehydrate in acetone, clear in xylol,
and mount in balsam.

Result. Bile pigments appear emerald green. Localizations can-
not be considered reliable due to the diffusibility of the reactants and
the final color.

Forsgren Test for Bile Acids

SPECIAL REAGENTS

3% Barium Chloride.
0.1 % Acid Fuchsin.
1 % Phosphomolybdic Acid.

Aniline Blue-Orange G Stain. Dissolve 0.5 g. aniline blue, 2 g.
orange G, and 2 g. oxalic acid in 100 ml. distilled water.

PROCEDURE

1. Treat small pieces of tissue for 6-12 hr. with the barium
chloride soln.

2. Fix in 10% formalin for 12-18 hr.

3. Prepare paraffin sections.

4. Stain sections for 1-3 min. in the acid fuchsin soln., and
wash in distilled water.

5. Place sections in the phosphomolybdic acid soln. for 0.5-1.0
min. and wash in distilled water.

6. Treat sections for 3-5 min. with the aniline blue-orange G
staip and wash in distilled water.

7. Dehydrate, clear, and mount in balsam.

Result. Bile secretory granules appear reddish.

ALDEHYDES, NUCLEIC ACIDS, AND PLASMAL 65

ALDEHYDES, NUCLEIC ACIDS, AND "PLASMAL"

The research of Feulgen and co-workers (1924,1938,1939) and
Imhiiuser ( 1927) led to the demonstration of a loosely bound alde-
hyde, "plasmal," in animal tissues. The bound form, "plasmalogen,"
liberates "plasmal" when treated with mercuric chloride or sub-
jected to prolonged acid hydrolysis. The Feulgen reaction, which
depends on the formation of a purple-colored compound when alde-
hydes react with fuchsin-sulfurous acid, is also given by desoxyri-
bonucleic acid (thymonucleic acid) after its purine bases are re-
moved by acid hydrolysis, but ribonucleic acid does not give the re-
action. The application of the Feulgen reaction to histochemical
studies on animal tissues was elaborated by Cowdry ( 1928) and
Verne (1928). Milovidov (1938) published a complete bibliography
of the 450 papers dealing with the Feulgen reaction up to 1938. Since
then Whitaker (1938) described a technique for plant tissues,
Stowell (1945a) studied the Feulgen reaction for the photometric
measurement of desoxyribonucleic acid (page 126), and Oster and
associates (1942,1944) employed the histochemical approach to
study tissue aldehydes in sections of fresh frozen material. An im-
proved preparation of the Feulgen reagent was reported by Cole-
man (1938). Rafalko (1946) claimed that small and diffuse chro-
matin elements could be detected with greater delicacy when the re-
agent was made by decolorizing a 0.5% solution of the dye by
bubbling sulfur dioxide through it.

The specificity of the Feulgen reaction for aldehydes has been
brought up repeatedly. It has been variously claimed that oleic and
cinnamic acids give a positive reaction, and that ketosteroids can-
not be differentiated from aldehydes by the reaction. Oster and
Oster (1946) have examined the question of specificity and have
found that the "true" reaction is indeed specific for aldehydes, other
carbonyl compounds giving a "pseudo" reaction in certain instances.
The differentiation between the "true" and "pseudo" reactions may
be made according to Oster and Mulinos ( 1944) x)n the basis that
the purple color developed in the former can be decolorized with di-
lute sodium hydroxide and restored to its original intensity with
hydrochloric acid, while the reddish color of the latter cannot be re-
stored by acid after the decolorization.

66 MICROSCOPIC TECHNIQUES

A means for the microscopic demonstration of ribonucleic acid
was developed by Diibos ( 1937) and Brachet ( 1940) , who employed
ribonuclease to break down the compound and thus destroy its
basophilic staining properties. The crystalline ribonuclease prepared
by Kunitz ( 1940) provided a more satisfactory reagent for carrying
out the procedure. Opie and Lavin ( 1946) demonstrated that ribo-
nucleic acid can be protected againct ribonuclease by precipitation
of the acid with lanthanum acetate. The basophilia of the precipitate
was retained even after treatment with ribonuclease.

The danger of an uncritical acceptance of the localizations ob-
tained by the Feulgen reaction has been emphasized by Danielli
(1946a). He pointed out that it remains to be proved whether the
experimental treatment of the nucleic acid has rendered it diffusible
enough for this factor to become significant in the interpretation. In
addition, he stressed the point that the use of an enzyme to digest
away a particular substance is open to some question with reference
to the specificity of the enzyme and the degree to which a clear-cut
removal of the substrate is possible. On the other hand, Stowell
( 1946) reviewed the evidence for and against the specificity of the
Feulgen technique for thymonucleic acid, and he concluded that
with the proper precautions it is one of the most specific histochem-
ical reactions. This does not mean that Stowell considers the tech-
nique beyond all criticism. No doubt he would agree with Danielli
that the interpretation of the results should be tempered with a
healthy awareness of the limitations involved, particularly the
diffusibility factor.*

Turchini and co-workers (1943, 1944, 1945) reported the use of
9-phenyl (or methyl) -2,6,7-trihydroxy-3-fluorone for the differential
staining of ribo- and desoxyribonucleic acids, the former giving rise
to a yellow-pink color, and the latter to a blue-violet. It is necessary
to hydrolyze the nucleic acid, as it is the pentose, thus liberated,
which yields the color. The hexoses formed by the hydrolysis of
tannins produce an orange-yellow color in the staining reaction when
it is applied to plant tissues (Turchini and Gosselin de Beaumont,
1945). -^i

* Other publications which have appeared subsequent to this writing are
given in the Bibliography Appendix, Refs. 15, 18 and 31.

ALDEHYDES, NUCLEIC ACIDS, AND PLASMAL 67

The use of ultraviolet absorption for the localization of nucleic
acids is discussed on page 113.

Coleman Preparation of Feulgen Reagent

Dissolve 1 g. basic fuchsin in 200 ml. boiling water; filter, cool, and
add 2 g. potassium metabisulfite (K2S2O5) and 10 ml. 1 N hydro-
chloric aci^. Let bleach for 24 hr., and then add 0.5 g. activated car-
bon (Norit), shake for about 1 min., and filter through coarse paper.
The filtrate should be colorless.

Whitaker Feulgen Technique for Plant Tissues

SPECIAL REAGENTS

Modified Brenda Fixative. Combine 30 ml. 1% chromic acid with

10 cc. 2% osmic acid.
1 N Hydrochloric Acid.
Feulgen Reagent. See above.
45% Acetic Acid.

PROCEDURE

1. Fix tissue in the modified Brenda fluid for a period depending
on the specimen, e.g., 15-20 min. for root tips, 30-45 min. for whole
anthers.

2. Hydrolyze in 1 .V hydrochloric acid at 50-60° for the same
time used in fixation.

3. Place in stain for 15-20 min. and then transfer to 45% acetic
acid for 10-15 min. or longer.

4. Put specimen in a drop of 45% acetic acid on a glass slide and
perform any dissections at this stage.

5. Place cover slip over the material and heat the slide nearly to
boiling at least three times. Apply pressure to cover slip with each
heating to make the tissue adhere to the slide.

6. Float off the cover slip in a mixture of equal vol. absolute
alcohol and glacial acetic acid.

7. Transfer to 95% alcohol for at least 15 min. and mount in
euperal. The mounting must be done in low humidity and care must
be taken to avoid breathing on the slide since moisture results in
cloudiness. The mounted material keeps well permanently.

Result. A positive reaction is indicated by a purple color.

68 MICROSCOPIC TECHNIQUES

Cowdry Modification of Feulgen Reaction for Paraffin
Sections of Animal Tissues

SPECIAL REAGENTS

Sublimate-Alcohol Fixative. Combine equal vol. saturated mercuric

chloride soln. and absolute alcohol.
1 N Hydrochloric Acid.
Feulgen Reagent. See page 67.
Sodium Bisulfite Solution. Add 30 ml. 1 M sodium bisulfite soln.

to 600 ml. tap water.

PROCEDURE

1. Prepare paraffin sections of tissue fixed in the sublimate-
alcohol fluid.

2. Pass through graded alcohols to water and place in the hydro-
chloric acid for 1 min.

3. Place in another portion of the acid at 60° for 4 min.

4. Treat with the Feulgen reagent for about 1.5 hr. The time may
have to be varied to suit the particular sections used.

5. Pass through three separate portions of the sodium bisulfite
soln. leaving in each for 1.5 min. and agitating frequently.

6. Wash for 5 min. in tap water.

7. Dehydrate, clear, and mount in balsam.

Oster Modification of Feulgen Reaction for Fresh-Frozen
Sections of Animal Tissues

SPECIAL REAGENTS

1 % Mercuric Chloride.

Feulgen Reagent. See page 67.

0.01 N Hydrochloric Acid Containing 1 % Sodium Bisulfite.

PROCEDURE

1. Cut 50 /x sections of fresh tissue on a freezing microtome. (The
sectioning should be carried out within 2-3 hr. after the death of the
animal and removal of the tissue. Until ready for use, keep the tissue
before cutting, and the sections after cutting, in physiological salt
solution.)

2. Place the sections in 1 % mercuric chloride for 5 min. in order
to liberate free aldehyde from "plasmalogen." Wash with water.

WATER-INSOLUBLE CARBONYL COMPOUNDS 69

3. Transfer to the Feulgen reagent for 15 min. and hold the
stained sections in the hydrochloric acid-sodium bisulfite solution.

4. Examine sections immediately after washing in distilled water.
The stain will last for a few days if the sections are kept in sulfurous
acid solution.

Method of Turchini et al. for Nucleic Acids

SPECIAL REAGENTS

Nucleic Acid Reagent. Dissolve 80 mg. of 9-phenyl (or methyl) -
2,6,7-trihydroxy-3-fluorone in 100 ml. 95% alcohol containing 15
drops cone, sulfuric acid.

1 N Hydrochloric Acid. (Or 25% cone, hydrochloric acid in 90%
alcohol.)

1 % Sodium Carbonate.

PROCEDURE

1. Fix the tissue (either plant or animal) in Bouin fluid.

2. Prepare paraffin sections in the usual manner.

3. If the methyltrihydroxyfluorone reagent is used: Hydrolyze
the deparaffinized sections in 1 A^" hydrochloric acid at 60° for 5 min.
wash with water, then alcohol, and treat for 5-10 min. with the re-
agent. Wash with several drops of 90% alcohol, then with 1%
sodium carbonate, rinse with water, and finally mount in balsam.

3a. With the phenyltrihydroxyfluorone reagent: Use the same
procedure as in step 3 but carry out the hydrolysis in the cold in
alcoholic 25% hydrochloric acid for 3-5 min.

WATER-INSOLUBLE CARBONYL COMPOUNDS

While the fuchsin-sulfurous acid test can be used for the localiza-
tion of aldehydes in tissue, other histochemical tests employed by
Bennett (1939, 1940) will react with either aldehydes or ketones.
Bennett concluded that his tests for the carbonyl group were indica-
tive of ketosteroids in the outer layer of the fascicular region of the
adrenal cortex. These carbonyl reactions can only indicate lipid
aldehj'-de or ketone and are in no way specific for ketosteroids as
Gomori (1942) pointed out; however, if other supporting evidence is
at hand, it may be reasonable to ascribe a positive reaction to the
ketosteroids present in a particular tissue. Subsequent work of
Albert and Leblond ( 1946) indicated that '"plasmalogen" rather
than ketosteroids is revealed by the phenylhydrazine reaction.

70 MICROSCOPIC TECHNIQUES

Bennett ( 1940) first removed ascorbic acid from the tissue to pre-
vent its interference with the tests. Albert and Leblond (1946)
substituted 2,4-dinitrophenylhydrazine for the phenylhydrazine of
Bennett. This enabled a more intense staining in thinner sections.

Bennett Use of Phenylhydrazine Reaction for Water-Insoluble
Aldehydes and Ketones

SPECIAL REAGENTS

M/10 Acetate Buffer, pH 6.0 to 6.5.

1 % Iodine in Alcohol.

1% Sodium Thiosulfate Solution.

1% Buffered Phenylhydrazine. Prepare just before use by mixing
equal vol. of 2% phenylhydrazine hydrochloride and the acetate
buffer. Gently bubble carbon dioxide through the solution for 15
min. to remove oxygen.

Control Reagent. Same as the 1 % buffered phenylhydrazine with-
out the phenylhydrazine.

PROCEDURE

1. Transfer frozen sections of fresh tissue from the microtome
directly into acetate buffer. If fixed tissue is employed, transfer to
water.

2. Add the iodine solution dropwise until a faint straw color
persists, and let stand 15 min.

3. Add sodium thiosulfate solution dropwise until the color is
discharged and a little more has been added ; let stand 5 min.

4. Wash the sections several times in distilled water.

5. Place the sections in glass-stoppered bottles containing
buffered phenylhydrazine solution. Fill the bottles to the top so that
no air bubbles are present under the stopper.

6. Run control sections as in previous steps, only use the control
reagent in place of phenylhydrazine.

7. After standing several hr. or overnight, wash all sections
with distilled water a few times.

8. Mount in glycerol or glycerol-gelatin and examine by means
of incident illumination from above.

Result. A yellow color appears in areas giving the positive test.
It is essential that care be taken in conducting this test since the
appearance of a yellow deposit on the walls of the bottle or on top
of the liquid indicates decomposition of the reagent, and when this

WATER-INSOLUBLE CARBONYL COMPOUNDS 71

occurs the yellow color in the sections cannot be relied upon to be
specific for the groups tested.

Albert and Leblond Use of 2,4-Dinitrophenylliydrazine Reaction
for Water-Insoluble Aldehydes and Ketones

SPECIAL REAGENTS

2,4-Dinitrophenylhydrazine Reagent. To a saturated soln. of 2,4-
dinitrophenylhydrazine (No. 1866 Eastman Kodak Co.) in 30%
alcohol, add sufficient 0.2 A^ sodium acetate to bring the pH to
neutrality.

PROCEDURE

1. Fix tissue for 48 hr. in formalin (neutralized with magnesium
carbonate) and wash in nmning water for 24 hr.

2. Prepare frozen sections 10-15 fx and place in 17% alcohol
for 4 hr.

3. Place sections in the 2,4-dinitrophenylhydrazine reagent over-
night and wash in 17% alcohol for 20 min.

4. Carry to distilled water and mount in glycerol gelatin.
Result. A positive reaction is shown by a yellow color.

Bennett Use of Seniicarl>azide Reaction for Water-Insoluble
Aldehydes and Ketones

SPECIAL REAGENTS

Acetate Buffer, Iodine, and Thiosulfate Solutions. Same as in the

preceding phenylhydrazine test.
Semicarbazide Reagent. Grind 10 g. semicarbazide hydrochloride

with 15 g. crystalline sodium acetate, take up the mixture in 100

ml. absolute methanol, and filter.
Control Reagent. Same as the semicarbazide reagent without the

semicarbazide.

PROCEDURE

1.-4. Follow the first four steps in the procedure for the phenyl-
hydrazine test,

5. Place sections in the semicarbazide reagent.

6. Place control sections in the control reagent.

7. After overnight standing, wash all sections several times with
distilled water.

8. Examine at unce with incident light from above.

72 MICROSCOPIC TECHNIQUES

Result. A yellowish deposit of semicarbazones appears in the
areas where the aldehydes and ketones are present.

PURINES

Tests that have been found to give positive results with all of the
purines have the unhappy characteristic of being highly unspecific.
Thus the reduction of silver salts is a reaction much too unspecific
to merit consideration; Saint-Hilaire's method involving precipita-
tion of insoluble copper salts of purines, and the transformation of
the copper into its red ferrocyanide is a reaction also given by
protamines, histones, and other protein products (Lison 1936, pages
183-186) .

The murexide test, which is positive with uric acid, xanthine and
its methyl derivatives, and guanine, is not given by adenine or
hypoxanthine (Lison 1936, pages 186-187). This reaction has the
disadvantage of being too drastic to permit its use for fine structures
and its disintegrating effect on tissue sections presents technical
difficulties. However, it may prove useful in some cases and for this
reason it will be described. Since it is no different from the xantho-
proteic reaction, a yellow-orange color would be indicative of pro-
teins, but it should be kept in mind that the xanthoproteic test is
not specific for proteins since other compounds, such as alkaloids,
benzene derivatives, etc., can also be nitrated in this manner to yield
products having the same color. •

Cowdry (1943, page 196) suggested that the modification of the
Courmont-Andre method by Hollande (1931) be used. It enables a
more reliable localization of urates in tissue.

Murexide Test for Certain Purines

SPECIAL REAGENTS

Concentrated Nitric Acid.
Concentrated Avfimonium Hydroxide.

PROCEDURE

1. Prepare sections by any of the usual methods.

2. Place a drop of nitric acid on a section and warm gently for
30 sec.

PURINES, INDOLE AND RELATED COMPOUNDS 73

3. Drain off the acid by means of blotting paper and add a drop
of water, which is also removed in the same way.

4. Expose the section to ammonia vapors.

Result. A purple-violet color is a positive test for uric acid,
guanine, and xanthine and its methyl derivatives. A yellow-orange
color is usually indicative of protein material. The effect of diffus-
ibility should be considered in the interpretation of localizations.

HoUande Modification of Courmont-Andre Method
for Uric Acid and Urates

SPECIAL REAGENTS

Silver Nitrate-Neutral Formalin Fixative. Equal vol. of 1% silver
nitrate and 4.4% formalin (neutralized with calcium carbonate)
mixed just before using.

0.5% Phosphomolyb die Acid.

PROCEDURE

1. Fix tissue in silver-formalin mixture for 12-24 hr. in the dark.

2. Wash for 24 hr. in several changes of distilled water.

3. Prepare paraffin sections.

4. Stain sections with hemalum for 10 min. and wash in running
tap water for 30-60 min.

5. Place in 1 % aqueous orange G or eosin 30-60 min. and wash
rapidly in distilled water.

6. Treat with the phosphomolybdic acid soln. and wash in dis-
tilled water.

7. Stain with 0.12% aqueous light green for 1-10 min.

8. Differentiate quickly in 95% alcohol, dehydrate in isoamyl
alcohol, clear in xylol, and mount in balsam.

Result. Urates will appear black, chromatin blue, protoplasmic
inclusions red to orange, and collagenic fibers green.

INDOLE AND RELATED COMPOUNDS

Lison (1936, pages 160-162) lists five reactions for the histo-
chemical detection of compounds containing the indole structure.
All of the tests leave much to be desired; their specificity is rather

74 MICROSCOPIC TECHNIQUES

poor. The Ehrlich reaction employing p-dimethylaminobenzal-
dehyde will give a violet color with phenols, aryl amines, and
heterocyclic compounds. Diazotization may indicate the same classes
of substances. The indophenin reaction utilizing isatin and sulfuric
acid gives a reddish-violet color in the presence of five-membered
heterocyclic compounds including indole. The nitrosamino reaction
of Lison converts the imino group in pyrrole or indole to a nitros-
amine by means of nitrous acid, and the nitrosamine is then made
to produce a green color through the Liebermann reagent (5%
phenol and concentrated sulfuric acid). This test is given by imino
groups, phenols, and primary aryl amines. The nitro reaction
enables differentiation between pyrroles and indoles; the sections
are treated with a mixture of equal parts of sulfuric and nitric acids,
and benzene ring compounds including indoles develop a canary
yellow color while pyrroles are not colored.

PHENOLS

Four main staining reactions have been employed for the detection
of phenols in tissue preparations. The azo reaction is based on diaz-
otization to form colored compounds; the indo reaction depends
on the formation of a green or blue indamine when an aromatic
para diamine is oxidized in the presence of tissue phenol; the
"argentaffin" reaction makes use of the reduction of ammoniacal
silver hydroxide and applies to ortho and para polyphenols, poly-
amines, and aminophenols ; and the "chromaffin" reaction, which is
used particularly to indicate adrenaline, gives rise to a brown color
when tissue is fixed with dichromate salts. A discussion of these tests
was given by Lison ( 1936, page 139-160) . The argentaffin test is
quite unspecific since many reducing substances can likewise give
a positive reaction. The chromaffin test is not entirely specific for
adrenaline, but has proved useful for the histochemical localization
of this biologically important substance.

Lison Modification of Chromaffin Reaction

SPECIAL REAGENTS

Formol-Milller Fixative or 5% potassium iodate in 10% formalin.
3% Potassium Dichromate or Potassium Iodate.

PHENOLS, UREA, AND SULFONAMIDES 75

PROCEDURE

1. Fix tissue in one of the solns. indicated. The iodate gives a
less intense reaction but is less prone to pseudoreactions.

2. Prepare sections and treat them with the 3% reagent for a
few hours.

Result. A brownish color indicates a positive reaction.

UREA

Two methods have been proposed for the localization of urea in
tissue sections. The Leschke procedure is based on fixation of the
tissue in a half-saturated mercuric nitrate solution in 1% nitric
acid and subsequent treatment of the sections with a saturated
hydrogen sulfide solution. The mercury urea compound is converted
to black mercuric sulfide, which is easily visualized. The xanthydrol
method depends on fixation of the tissue in a solution of xanthydrol
in acetic acid in order to precipitate dixanthylurea which can be
recognized in sections by its double refraction when examined under
a polarizing microscope. Lison ( 1936, pages 165-170) critically dis-
cussed these methods. As he pointed out, the usefulness of the mer-
cury reaction is entirely vitiated by its extreme lack of specificity,
far too many tissue constituents being capable of precipitation by
mercury salts. The xanthydrol reaction is chemically specific, but
its serious fault lies in a combination of unfortunate factors including
the great diffusibility of urea, the poor penetrability of xanthydrol,
and the slowness of the reaction between the two. The result is that
the position of the crystals formed bears little or no relation to the
regions originally containing urea. A suitable method for the his-
tological localization of urea is not available at present.

SULFONAMIDES

^lacKee et al. ( 1943) described a test for sulfonamides in frozen
tissue sections depending on the formation of a yellow to orange
precipitate of the p-dimethylaminobenzylidene derivative when sul-
fonamides react with p-dimethylaminobenzaldehyde. It should be
borne in mind that procaine, phenacetin, acetanilid, and aromatic
amino compounds in general will also give the reaction.

76 MICROSCOPIC TECHNIQUES

Another method was published by Hackmann (1942), who em-
ployed the freezing-drying technique for the fixation of the tissue,
prior to the preparation of paraffin sections. Colored sulfonamides
were observed directly in the sections, and colorless ones were
visualized by forming a red azo dye in the following manner: The
sections were exposed to nitrous acid vapor for 30 sec. by placing
the slide over a measuring cylinder 20 cm. high containing several
milliliters of 0.1 A^" hydrochloric acid and a few milligrams of sodium
nitrite. The diazotized sulfonamide was coupled with a-naphthyl-
amine by immersing the slide in a 5% solution of the amine in xylol.

The detection of sulfonamides by fluorescence microscopy is
discussed on page 108.

Method of MacKee et ah for Sulfonamides

SPECIAL REAGENTS

Sulfa Reagent. Dissolve 1 g. pure p-dimethylaminobenzaldehyde
in a soln. of 95 ml. absolute alcohol and 5 ml. cone, hydrochloric
acid. Store in a glass-stoppered amber bottle, and do not use after
2-3 weeks or when the soln. becomes yellow.

5% Concentrated Hydrochloric Acid in Absolute Alcohol.

PROCEDURE

1. Fix the tissue for 2-24 hr. with formaldehyde gas by covering
the bottom of a beaker with paraformaldehyde and the top with a
piece of gauze, placing the tissue on the gauze, setting the whole in
a glass jar whose floor has also been covered with paraformaldehyde,
and closing the jar tightly with a glass lid.

2. Cut frozen sections of the fixed tissue 10-20 /^ thick and place
directly on glass slides.

3. Cover each section with a drop or two of the sulfa reagent,
and after 3-5 min. add a drop or two of the alcoholic hydrochloric
acid soln.

4. Dry quickly without heat by absorbing excess fluid on filter
paper and holding in a current of air.

5. Cover at once with a drop of damar resin in xylol (10 g.
resin dissolved in 10 g. xylol) and fit cover slip, taking care to re-
move air bubbles. Seal edges with melted paraffin.

SULFONAMIDES AND UREASE 77

6. Run controls by repeating the above steps but omitting the
treatment with the sulfa reagent, or repeat the complete procedure
on a portion of the same kind of tissue known to be free of sulfona-
mides.

Result. Sulfonamides are indicated by the presence of a pre-
cipitate that ranges in color from lemon-yellow to orange. However,
the color fades rapidly, particularly in the presence of air, making
it necessary to examine the sections as early as possible. Colored
photomicrographs should be taken not later than 3-4 hr. after the
reaction has occurred.

D. ENZYMES

UREASE

Sen ( 1930) elaborated a method for the localization of urease in
tissue sections which he employed for a study on the jack bean. The
carbonic acid formed on decomposition of urea is precipitated as
calcium carbonate, which may be visualized by conversion to silver
carbonate and reduction of the latter to a black deposit of metallic
silver; or the carbonic acid may be converted to cobalt carbonate and
the latter changed to a brown or black precipitate of cobalt sulfide.
The latter method is to be preferred. This principle was later employed
by Gomori for the localization of the phosphoric acid hberated by
phosphatases, pages 78 and 80. However, Sen digested the tissue
in the substrate medium before paraffin infiltration and sectioning,
and only treated the deparaffinized sections with sulfide to convert
the colorless cobalt salt to the black sulfide. This procedure has many
disadvantages; the schedule of Gomori, in which the sections are
prepared prior to digestion, should be used instead, if the enzyme
can stand the dehydration, paraffin embedding, and deparaffinization.

For jack bean tissue. Sen employed a preliminary treatment for
1 hr. with 1% cobalt nitrate in 80% alcohol followed by a 48-60
hr. digestion with a substrate medium consisting of 0.5% urea and
0.5% cobalt nitrate in 80% alcohol. The cobalt carbonate was con-
verted to sulfide by the action of either dilute sodium sulfide or a
saturated solution of hydrogen sulfide. For animal tissues, Sen used
cobalt-urea solns. in graded alcohols from 60 to 80%.

78 MICROSCOPIC TECHNIQUES

ALKALINE PHOSPHATASE*

The same staining technique for the visualization of alkaline phos-
phatase activity was developed independently and simultaneously,
oddly enough, by Gomori ( 1939) in Chicago and Takamatsu ( 1939)
in Japan. Their method was based on the finding that, when sections
of tissue were placed in an alkaline medium containing sodium
glycerophosphate, the sites of the enzymatic liberation of phosphate
could be determined if calcium ions were present to precipitate the
phosphate as it was formed. The deposit of calcium phosphate then
could be converted to a more easily visualized black precipitate of
cobalt sulfide or metallic silver. Gomori ( 1939) , Hepler et al. ( 1940) ,
Takamatsu (1939), and Kabat and Furth (1941) have employed
the von Kossa silver stain; but, as Bourne (1943) has indicated, it
is probably inferior to the cobalt stain used extensively in the latter
work of Gomori ( 1941a, 1943) .

The specificity of the stain for phosphatase has been demonstrated
by Gomori (1939, 1941a) and Kabat and Furth (1941), and in a
critical study later Danielli (1946b) claimed reliability for the
localizations obtained. Preformed insoluble calcium salts will give a
positive test and therefore these should either be removed by treat-
ing the sections, before incubation with substrate, with citrate buffer
of pH 4.5 to 5.0 for 15 min. (Gomori, 1946c), or control sections
stained to demonstrate the preformed salts should be compared to
the sections treated to visualize the enzyme reaction. Of course the
former is preferable.

Tissues too hard to be sectioned without decalcification present
a particular problem since phosphatase is destroyed by the usual
processes of decalcification. Kabat and Furth (1941) circumvented
this difficulty to some degree by the use of 10% diammonium citrate,
which they found could effect certain decalcifications without
damaging phosphatase. Bourne (1943) has proposed that small
pieces of bone tissue be fixed in 80% alcohol, treated with the sub-
strate medium and then with cobalt solution and sulfide to convert
the calcium phosphate precipitate to one of cobalt sulfide, and finally
subjected to decalcification with trichloroacetic acid. Cobalt sulfide
is insoluble in trichloroacetic acid and hence the decalcification can
be performed as a final step. Controls can be made by following the

*See Bibliography Appendix, Ref. 7.

ALKALINE PHOSPHATASE 79

same procedure but oniitting the treatment with substrate. Another
procedure for use with bone has been given by Bourne (1943) in-
volving the addition of 0.01% sodium alizarin sulfonate to Gomori's
substrate medium. The calcium phosphate produced is automatically-
stained red by the alizarin dye.

The use of magnesium ions to activate the phosphatase was intro-
duced by Kabat and Furth (1941) and is employed in the revised
method of Gomori ( 1946c) .

It is of interest to call attention to the different approach to the
staining technique for alkaline phosphatase that was brought for-
ward by Menten, Junge, and Green (1944). These investigators em-
ployed a reaction of the organic, rather than the phosphate, moiety
of the substrate to precipitate a reddish-purple dye at loci of phos-
phatase action. Employing calcium /j-naphthol phosphate as the sub-
strate, |3-naphthol liberated by the enzyme was made to react at
once with diazotized a-naphthylamine present in the substrate solu-
tion. While this procedure can undoubtedly be applied in many in-
stances, it would appear to offer no advantage over the Gomori
method, and, as Menten et al. readily admit, in its present form the
test is intricate and probably not well suited to routine use. Never-
theless, Yin (1945) employed this method for plant tissues in order
to avoid interference by preformed phosphates. Since the preformed
phosphates can be removed with citrate buffer (page 78) it would
appear that interference from this source need not be made a deter-
mining factor in the choice of a method.

Gomori Revised Method for Alkaline Phosphatase

SPECIAL REAGENTS

0% Acetylcellulose (Eastman's No. 4644) in acetone. Optional.

1-2% Cobalt Acetate, Chloride, or Nitrate.

Ammonium Sulfide Solution. A few drops of yellow ammonium

sulfide soln. to a Coplin jar of distilled water.
Substrate Medium, pH 9.4. (Will keep in refrigerator for months.)

Combine 25 ml. 2% sodium glycerophosphate, 25 ml. 2% sodium

barbital, 50 ml. distilled water, 5 ml. 2% calcium chloride, 2 ml.

2% magnesium sulfate, and a few drops of chloroform.

PROCEDURE

1. Place slices of fresh tissue, under 2 mm. thick, in chilled abso-

80 MICROSCOPIC TECHNIQUES

lute acetone and fix for 12-24 hr. in a refrigerator. Dehydrate at
room temperature in two changes of absolute acetone for 6-12 hr.
each time.

2. Optional step. To strengthen sections which tend to break up
when floated on the lukewarm water after sectioning, impregnate the
tissue with the acetone-acetylcellulose soln. 24 hr.

3. Drain off the fluid rapidly and place in two changes of benzol
for 30 min. each.

4. Embed in paraffin not over 56° up to 2 hr. To hasten the proc-
ess use 3 changes of paraffin, each for 20 min., and carry out the
second change in vacuo in a wide-mouth bottle with a one-hole
rubber stopper fitted with a glass tube. Connect the glass tube by
rubber tubing, passed through an air hole in the paraffin oven, to a
water aspirator via a safety bottle.

5. Cut sections 4-8 ju, thick, fioat them on lukewarm water
(30-35°), and mount on slides.

6. Let slides dry, place in the paraffin oven for 10 min. to melt the
paraffin, and run through xylol and alcohols to distilled water. Re-
move preformed mineral deposits as described on page 78.

7. Incubate the sections for 1-2 hr. at 37° in the substrate me-
dium.

8. Rinse with water, immerse in the cobalt soln. for 5 min., and
rinse well with several changes of distilled water.

9. Place in the diluted ammonium sulfide soln. for 1-2 min.

10. Wash well in water, counterstain if desired, dehydrate, and
mount.

Result. Sites of the phosphatase activity appear brown or black.

ACID PHOSPHATASE*

The staining method for alkaline phosphatase cannot be used for
acid phosphatase since calcium phosphate is soluble at a pH around
5, which is optimum for the action of the latter enzyme. Hence,
Gomori (1941b) employed lead ions in the substrate medium at pH
4.7 so that insoluble lead phosphate would be formed at the sites of
enzymatic activity. The lead phosphate was then converted either
to brown or black lead sulfide, or was stained a purplish-red with
acridine red. Wolf, Kabat, and Newman (1943) introduced several

*See Bibliography Appendix, Refs. 1 and 9.

ACID PHOSPHATASE 81

improvements in the procedure, and subsequently Gomori (1946c)
revised his original method. His experience with the technique led
Gomori (1946c) to state: "For some unknown reason, the staining
for acid phosphatase sometimes turns out patchy, occasionally even
negative, when it should be positive. This seems to happen especially
in cases when the pieces have been exposed to the temperature of
the paraffin oven for more than an hour, or when the temperature of
the oven is over 56°C."

The intensification of the acid phosphatase test by manganese ions
has been demonstrated by Moog (1943a), who found that a con-
centration of 0.01 M manganese sulfate gave the most satisfactory
results. This investigator recommends that the activator be added to
a clear portion of substrate medium just before use, and points out
that the incubation period may be approximately halved when the
reaction is accelerated by the manganese. Moog found that 4-5 hr.
was a satisfactory incubation period for tissues of the 6-day chick
embryo. No doubt a certain amount of trial and error must be ap-
plied to determine the proper incubation time for the particular
tissue under investigation. In a later study Moog (1944) found that
0.01 M ascorbic acid activated acid phosphatase and in some respects
appeared to have advantages over the action of manganese sulfate.

The application of the Gomori method to grains and sprouts was
made by Glick and Fischer (1945b). The modification in technique
necessitated for these tissues will be presented.

Gomori Revised Method for Acid Phosphatase in Animal Tissues

SPECIAL REAGENTS

5% Acetylcellulose (Eastman's No. 4644) in acetone. Optional.

£% Acetic Acid.

Ammonium Sulfide Solution. A few drops of yellow ammonium sul-
fide soln. to a Coplin jar of distilled water.

Substrate Medium, pH 5. Combine 30 ml. of 1 M acetate buffer
(100 ml. 13.6% sodium acetate, CHsCOONa.SHsO, -f- 50 ml. 6%
acetic acid), 10 ml. 5% lead nitrate, and 60 ml. distilled water,
and add slowly while stirring 30 ml. 2% sodium glycerophosphate.
Shake the mixture, let stand for a few hr., and store in a refrig-
erator. Before use, filter a small amount and dilute it with 2-3
parts distilled water.

82 MICROSCOPIC TECHNIQUES

PROCEDURE

1-6. Same as for alkaline phosphatase (pages 79 and 80).

7. Incubate the sections for 1-24 hr. at 37° in the substrate
medium.

8. Rinse well in distilled water, followed by 2% acetic acid, and
then in distilled water again.

9-10. Same as for alkaline phosphatase, page 80.

Result. Sites of the phosphatase activity appear brown or black.

Glick and Fischer Adaptation to Grains and Sprouts of Gomori
Method for Acid Phosphatase

SPECIAL REAGENTS

Substrate Medium. Combine the following, shake thoroughly,
centrifuge, and use the clear liquid: 4 ml. 0.1 ilf acetate buffer of
pH 5.1, 1 ml. 0.1 M lead nitrate, 0.6 ml. distilled water, and 0.4 ml.
3.2% sodium-a-glycerophosphate. Booth (1944) showed that the a
compound is hydrolyzed more rapidly than the (3 by wheat phos-
phatase, and Gomori (1941) found that the a compound has the
additional advantage that its lead salt is more soluble than that of
the /? at this pH value. Because it was easily available, the mix-
ture, containing 52% a and 48% fi of Eastman Kodak Co., was
used by Glick and Fischer ( 1945) .

2% Acetic Acid Solution.

Ammonium Sulfide Solution. 1 ml. to a Coplin jar of water.

PROCEDURE

/, Preparation of paraffin sections. A, Kernel sections:

1. Soak kernels in water about 7 hr.

2. For longitudinal sections, cut off a layer from both the crease
side and the opposite side of the kernel. For cross sections, cut o&
kernel just behind germ. This enables more efficient penetration of
liquids.

3. Let kernels stand overnight in absolute alcohol. In the morn-
ing change to a mixture of 1 vol. absolute alcohol -f- 3 vol. n-butyl
alcohol.

4. In the evening transfer to n-butyl alcohol.

ACID PHOSPHATASE 83

5. The following morning place in xylol, and let stand until
evening.

6. Transfer to a xylol-paraffin mixture containing just enough
xylol to keep the paraffin in soln. at room temperature and let stand
overnight. Tissuemat (Fisher Scientific Co.) gives better results than
paraffin. In a warm room the variety melting at 60-62° gives better
sections than the material having a lower melting point.

7. Place in a soln. of 1 vol. xylol + 2 vol. melted paraffin in
a 60° oven for 1-2 hr.

8. Infiltrate with melted paraffin for 2 hr, in the oven, then
change to fresh paraffin for 4 hr., and finally embed.

9. Cut sections 10 /x thick and mount on slides with aid of Mayer
albumin. (Combine 1 vol. filtered fresh egg white with 1 vol. glyc-
erol, and add a bit of camphor as a preservative.) Smear the liquid
in a thin film on a slide and rub with the finger, cover with water,
transfer section to the slide, place in oven for 5 min. at about 55° to
soften the paraffin and allow wrinkles to straighten out, drain off
water with a towel, and allow to dry for 2 hr. in the 55° oven. Store
mounted sections in refrigerator until ready for use.

10. Remove paraffin from sections with two changes of xylol fol-
lowed by two changes of absolute alcohol.

11. Dip slides into 0.5-1.0% collodion in alcohol-ether to cover
section with a protective film; harden film by dipping into 80%
alcohol, and wash with distilled water.

B. Rootlet and leaf section of the sprout:

1. Place in the following solns., for 1 hr. in each case, in the order
given :

a. 70% alcohol

b. 80% alcohol

c. 65 ml. 80% alcohol -^ 35 ml. n-butyl alcohol

d. 45 ml. 95% alcohol ~ 55 ml. n-butyl alcohol

e. 25 ml. absolute alcohol ^ 75 ml. n-butyl alcohol

f. n-butyl alcohol

g. xylol

2. Follow step 6 under A in the preceding part, allowing the
material to stand in the mixture for only i/^ hr.

3. Subject the material to three changes of melted paraffin in the

84 MICROSCOPIC TECHNIQUES

60° oven during the course of 1 hr. If air bubbles are present in the
leaves, apply suction to remove them.

4. Embed in paraffin colored red by stirring a few grains of Sudan
IV in the molten material. In uncolored paraffin it is difficult to see
the tissue in the sections.

5. Cut sections, mount on slides, remove paraffin, and protect
with collodion film just as in steps 9, 10, and 11 under A.

IT, Preparation of frozen sections. A, Kernel sections (rootlet and
leaf sections of the sprout are too fragile to permit satisfactory
frozen-section technique) :

1. Soak kernels 4-6 hr. in water.

2. Mount in a drop of water on freezing head of microtone.

3. Cut sections 15 fi thick, keeping knife cold with Dry Ice, and
transfer, with a needle cooled by Dry Ice, into 80% alcohol. (The
80% alcohol is used rather than water since, in the latter medium,
the starch endosperm disintegrates, and separates from the rest of
the section.)

4. Float the section onto a glass slide immediately. If wrinkled,
straighten section in a drop of 70% alcohol.

5. Dehydrate by covering section with five successive drops of
absolute alcohol, draining off after each drop is added.

6. Cover section with a small drop of 0.5-1.0% collodion soln.
and harden film by dipping slide in 80% alcohol. Wash in distilled
water.

Ill, Demonstration of enzyme activity:

1. Remove preformed mineral deposits by placing sections in dis-
tilled water for 24 hr. at room temperature. Citrate buffer could
probably be used too (page 000) .

2. Place sections in substrate medium at 37°, for the following
digestion periods:

Kernel, paraffin sections — Embryo, 1 hr.; non-embryonic part 30
min.

Kernel, frozen sections — Embryo, 15-30 min.; non-embryonic part,
5-10 min.

Rootlets, paraffin sections — 3 hr.

Epicotyl, paraffin sections — 24 hr.

3. Wash sections with three changes of distilled water, dip into
2% acetic acid, and wash well with distilled water.

ACID AND OTHER PHOSPHATASES 85

4. Place in ammonium sulfide soln. for 2-3 min.

5. Wash with several changes of distilled water, dehydrate in
95% alcohol for 2-3 min., and follow by 5 min. in absolute alcohol.

6. Clear in oil of thyme for 3-4 min., treat with three changes of
xylol. (Treatment with xylol should be brief since the black
precipitate indicating enzyme action is soluble to some degree in
xylol.) Mount in balsam.

Result. The phosphatase activity is visualized as a brown or
black precipitate.

OTHER PHOSPHATASES

Various investigators have studied the histological distribution
of phosphatases capable of hydrolyzing substrates other than those
commonly employed for the acid and alkaline phosphatases. This
work has been accomplished by simply substituting the new
substrates for the glycerophosphate usually used, and employing
the standard phosphatase procedures.

Wolf, Kabat, and Newman ( 1943) used ribonucleic acid and
glucose- 1-phosphate as additional substrates in their work on acid
phosphatase distributions, particularly in the human and guinea
pig nervous systems. Glick and. Fischer (1945b, 1946a) employed
adenosine triphosphate, thiamine pyrophosphate, and glucose- 1-
phosphate in a study of the enzyme distributions in wheat and parts
of the germinated grain. In investigations on the mouse duodenum,
Dempsey and Deane ( 1946) , and in work on the thyroids of various
species, Dempsey and Singer (1946), utilized adenylic acid, ribo-
nucleic acid, glucose- 1-phosphate, fructose diphosphate, and lecithin
as their additional substrates. Krugelis (1946) studied the phos-
phatases in the larval salivary glands of Drosophila and in various
organs of the mouse using adenylic acid, guanylic acid, cytidylic
acid, ribonucleic acid, desoxyribonucleic acid, and a depolymerized
form of the latter, as substrates.

With the staining technique it is difficult at times to define the
specificities of the various phosphatases which act on the different
substrates. For instance, both alkaline phosphatase and adenyl-
pyrophosphatase (adenosinetriphosphatase), which are known to
be two distinct enzymes, can act on adenosine triphosphate, as

86 MICROSCOPIC TECHNIQUES

Moog and Steinbach (1946) have emphasized. Accordingly, differ-
ences in their localizations or in their properties must be exploited
to enable their separate identification in tissue sections (Glick,
1946). From the work of Dempsey and Deane (1946) it would
appear that several phosphatases may coexist in the same cellular
location, and that their differentiation must depend on differences
in pH optima or other properties. When glucose- 1 -phosphate is
employed as the substrate, enzymatic liberation of phosphate might
occur either by phosphatase action or by the phosphorylase action
which converts the substrate to glycogen or starch; accessory
evidence would be required to determine which of these two enzymes
was being visualized by the staining reaction.*

While it does not offer rigorous proof, in some cases differentiation
between enzymes may be based on differences in their localization
as seen in stained sections. The presence in the nucleoli of the cells
of the wheat epicotyl of an enzyme capable of hydrolyzing adenosine
triphosphate, but not thiamine pyrophosphate (Glick and Fisher,
1946a), would suggest that these substrates are acted upon by
different enzymes. Likewise, the fact that a strong enzymatic
activity is found in the cytoplasm of cells in mouse tissues when
ribonucleic acid is used as the substrate, while only a slight reaction
is observed in the nuclei, and the reverse distribution is seen when
a depolymerized form of desoxyribonucleic acid is used, indicates
that separate enzymes are involved in the hydrolysis of these
substrates (Krugelis, 1946). Furthermore, the approximately equal
activities in both cytoplasm and nucleus toward glycerophosphate
as substrate might be taken as an indication of the presence of a
third enzyme in these cells, as Krugelis has pointed out. Another
example is to be found in the differences in the localizations of the
enzymes acting on glucose-1-phosphate and fructose diphosphate
when the mucosa of the mouse duodenum is studied (Dempsey and
Deane, 1946). Other cases might be cited to illustrate the same
general point.

ZYMOHEXASE (ALDOLASE plus ISOMERASE)

Aldolase converts hexose diphosphate to both dihydroxyacetone
phosphate and phosphoglyceraldehyde; isomerase catalyzes equi-

* See Bibliography Appendix, Ref. 19.

ZYMOHEXASE 87

librium between the two. Together these two enzymes are referred
to as zymohexase.

Allen and Bourne ( 1943) adapted the microscopic technique for
phosphatase (page 78) to this enzyme system, whose distribution
they studied in skeletal, heart, and smooth muscle tissue. By
incorporating iodoacetic acid into their substrate media, they
prevented further enzymatic breakdown of the triose phosphates.
The distinct difference in localization of zymohexase and alkaline
phosphatase precluded the possibility of confusing the two; however,
the phosphatase activity could be selectively blocked by fluoride.
Allen and Bourne utilized the fact that the triose phosphates formed
by the zymohexase action will spontaneously liberate inorganic
phosphate at room temperature in alkaline solution. The phosphate
could then be precipitated, and finally visualized in the manner of
Gomori (page 78). It was observed that sections which had been
infiltrated with paraffin lost their enzyme activity and, accordingly,
frozen sections were employed.

Allen and Bourne Method for Zymohexase

SPECIAL REAGENTS

0.1 M Sodium lodoacetate. Neutralize 1.86 g. iodoacetic acid with
1 N sodium hydroxide to bromothymol blue and dilute to 100 ml.

0.1 M Sodium Fluoride.

2% Cobalt Chloride (C0CI2.6H0O).

Ammonium Sulfide Solution. Dilute 1 ml. yellow ammonium sulfide
to 50 ml. with distilled water. Prepare fresh before use.

Magnesia Mixture. Dissolve 5.5 g. magnesium chloride
(MgCl2.6H20) and 7.0 g. ammonium chloride in 35 ml. 5 N am-
monium hydroxide. Filter after 1 hr. and add 60 ml. 4 A^" am-
monium hydroxide to the filtrate.

Purified Sodium Hexose Diphosphate. Formula I: mix 40 ml. 4%
sodium hexose diphosphate ( prepared by treating the calcium salt
with sodium oxalate) and 20 ml. magnesia mixture, and filter off
precipitated phosphate after 30 min. Formula II: mix 20 ml. of
4% sodium hexose diphosphate with 20 ml. of magnesia mixture
and 20 ml. of distilled water; filter after 30 min. as above.

Substrate A. Combine 10 ml. of the purified salt soln. (Formula I)
with 1.7 ml. of 0.1 M sodium iodoacetate and 5 ml. distilled water.

88 MICROSCOPIC TECHNIQUES

Substrate B. Same as A, but use 3.3 ml. water and 1.7 ml. of 0.1 M
sodium fluoride in place of the 5 ml. water. (Substrate B was used
to prevent hydrolysis by phosphatase through the action of the
fluoride. Actually Allen and Bourne found that it was not needed
in their work since no phosphatase action was observed in their
particular experiments.)

Substrate C. Combine 15 ml. of the purified salt soln. (Formula II)
with 2.5 ml. of the sodium iodoacetate and 7.5 ml. water.

PROCEDURE

1. Fix tissue in 80% alcohol for 24 hr.

2. Wash in water for 5-10 min.

3. Prepare frozen sections and place them in either substrate
A or C for 1-2 hr. at 37°. (An extraneous dark precipitate forms
on the surface of some of the sections with substrate A, but not
with C, presumably because of the higher dilution of the substrate
in the latter.)

4. Treat the sections with the cobalt chloride soln. for several
hr. and then with the ammonium sulfide soln. for 10 min.

5. Dehydrate in alcohols, clear in xylol, and mount in balsam.

6. Prepare control sections to demonstrate preformed phosphate
by omitting the treatment with substrate in step 3 and proceeding
with steps 4 and 5.

Result. The formation of a brownish-black precipitate indicates
zymohexase activity.

LIPASE

Gomori (1945b) adapted the principle of his method for the
demonstration of phosphatases in tissue sections to the localization
of lipase. The great difficulty encountered in previous attempts has
been to find a substrate that is water soluble and whose acid split-
product could be precipitated by some ion having no adverse effect
on the enzyme. The ordinary esters of mono- and dicarboxylic acids
do not meet both of these requirements. Gomori was able to circum-
vent the difficulty by the use of some new long-chain fatty acid
esters of hexitans and hexides in which most of the hydroxyl groups
are etherified. These compounds were developed by Atlas Powder

LIPASE 89

Co., and are known as "Tweens." Tween 40 and Tween 60, employed
by Gomori, are described by Atlas Powder Co. as "sorbitan mono-
palmitate polyoxyalkylene derivative" and "sorbitan monostearate
polyoxyalkylene derivative," respectively. Gomori states that these
substrates were found to be hydrolyzed by pancreatic lipase at a
rate about half that of olive oil. In the presence of 0.2% sodium
taurocholate an intensification of the reaction in pancreatic tissue
was noted, while in all other tissues, enzyme inhibition was observed.
In a later publication Gomori (1946a) reported that "Product 81"
of Otiyx Oil and Chemical Co. could also serve as a lipase substrate
for histochemical purposes. This compound is a stearic acid ester
of "comparatively short-chained polyglycols." The only shortcoming
observed by Gomori to the use of these substrates is an occasional
failure to obtain proper counterstaining with hematoxylin, especially
after use of the "Tweens" and more rarely with "Product 81." The
method as finally given by Gomori (1946c) will be described.

Gomori Revised Method for Lipase

SPECIAL REAGENTS

5% Acetylcellulose (Eastman's No. 4644) in acetone.

1-2% Lead Nitrate.

Ammonium Sulfide- Solution. A few drops of yellow ammonium
sulfide soln. to a Coplin jar of distilled water.

Substrate Medium. Stock solution I: combine 150 ml. glycerol,
60 ml. of 10% calcium chloride, 50 ml. M/2 maleate buffer pH
7 to 7.4 (dissolve 5.8 g. maleic acid in 94 ml. of 4% sodium
hydroxide and 6 ml. water), and distilled water to make 1000 ml.
Stock solution II: 5% Tween 40 or 60 or "Product 81." Add
merthiolate to 0.02% in each stock soln. and store in refrigerator;
the solns. may be used for many months. Before use, add 2 ml.
stock soln. II to 50 ml. stock soln. I.

PROCEDURE

1-6. Same as for alkaline phosphatase (pages 79 and 80).

7. Incubate the sections in the substrate medium for 6-12 hr.
at 37°.

8. Rinse with distilled water and transfer to the lead nitrate
soln. for 10 min.

90 MICROSCOPIC TECHNIQUES

9. Rinse in repeated changes of distilled water and immerse in
the diluted ammonium sulfide soln. for 1-2 min.

10. Wash well in water, counterstain lightly with hematoxylin
and eosin, dehydrate in alcohols, clear in gasoline or tetrachloro-
ethylene, and mount in Clarite dissolved in the same liquid. (Toluol
or xylol causes fading of the stain.)

Result. Sites of lipase activity appear golden brown.

PEROXIDASE

Most of the various microchemical methods for the histological
localization of peroxidase activity are based on the oxidation of
benzidine. The methods of McJunkin ( 1922) , designed for use with
human tissues, and Armitage ( 1939), developed for examining blood
and bone marrow smears, have been chosen for presentation here
since they are the most recent and seem to be the best. Peroxidase
actually occurs most abundantly in plants, but the methods appear
to have been worked out for animal tissues or cells exclusively.
However, there appears to be no reason why these methods cannot
be adapted to plant material as well.

McJunkin Method for Peroxidase in Tissue Sections

SPECIAL REAGENTS

Benzidine Reagent. Dissolve 100 mg. benzidine in 25 ml. 80%
methanol and add 2 drops 3% hydrogen peroxide. Dilute with
1-2 vol. distilled water before using. Store in the dark.

PROCEDURE

1. Place formalin-fixed tissue, in pieces 1 mm. thick, in 70%
acetone for 1 hr. followed by pure acetone for 30 min., benzol for
20 min., and melted paraffin 20 min.

2. Cut sections 3.5 to 5.0 [i, fix to slides with albumin, and dry
overnight at room temperature.

3. Remove paraffin by placing in benzol for 20 sec. and acetone
for 10 sec.

4. Plunge in water for a few sec. and remove excess water. Apply
the benzidine reagent for 5 min. and transfer to water for 5 min.

5. The sections may be stained with Harris hematoxylin for 2
min., rinsed in water 1 min., and stained with 0.1% eosin for 20 sec.

PEROXIDASE AND DOPA OXIDASE 91

6. Dehydrate in 95% alcohol for 30 sec, and absolute alcohol
for 5 sec.

7. Clear in xylol and mount in balsam.

8. Run controls in which the benzidine is omitted.

Result. Peroxidase manifests itself by an initial blue color which
changes to brown. Diffusibility, particularly of the color produced,
can be expected to interfere with proper localization of the enzyme.

Armitage Method for Peroxidase in Blood or Bone
Marrow Smears

SPECIAL REAGENTS

Fixing Solution. 10% Formalin in 96% alcohol. Prepare the soln.

just before using.
Benzidine Reagent. Dissolve 750 mg. benzidine in 500 ml. 40%

alcohol, filter, add 0.7 ml. 3% hydrogen peroxide, and shake

before using. If stored in the dark, the reagent will be good for

months.

PROCEDURE

1. Fix the smear in the alcoholic formalin.

2. Cover the material with the benzidine reagent for about 2 min.
if the smear is fresh, and up to 20 min. if it is old.

3. Wash in 40% alcohol until yellow granules appear in the
leucocytes.

4. Dehydrate in absolute alcohol and dry at about 37°.

5. A counterstain of dilute Giemsa or dilute Leishman stain may
be applied for 30 min. followed by washing in water, blotting, and
drying.

Result. The appearance of yellow granules is a positive test for
peroxidase.

DOPA OXIDASE

The enzymatic oxidation of 3,4-dihydroxyphenylalanine — or dopa
— has been applied, histologically, to the identification of melano-
blasts, since these appear to be the seat of the oxidase activity and
the conversion of dopa to melanin results in their becoming black-
ened. Bloch's earlier work has been adapted by Laidlaw (1932) and
Laidlaw and Blackberg ( 1932 ) to the demonstration of dopa oxidase

92 MICROSCOPIC TFX'HNIQUES

activity in histological preparations. Sharlit et al. (1942) have
pointed out that the method for demonstrating dopa oxidase may
of itself, in the absence of substrate, cause an increase in melanin.
This makes it necessary to run suitable control experiments. The
reaction may be hastened by employing a buffer of a little higher
pH, or retarded by shifting toward the acid side.

Laidlaw Melliod for Dopa Oxidase

SPECIAL REAGENTS

Dopa Stock Solution. Dissolve 0.3 g. dopa (Hoffmann-La Roche,
labeled "for Bloch's dopa reaction") in 300 ml. cold distilled
water. Store in a refrigerator and discard when a distinct red color
has developed.

Buffered Dopa Solution (pH 7.4) . Add 2 ml. potassium dihydrogen
phosphate (9 g. KH2PO4/I.) and 6 ml. disodium hydrogen phos-
phate (11 g. Na2HP04.2H20/l.) to 25 ml. dopa stock soln. Filter
through fine paper.

Buffer Solution for Control Experiment. Replace the 25 ml. dopa
soln. with an equal vol. distilled water in the buffered dopa soln.
above.

PROCEDURE

1. Prepare frozen sections of fresh tissue. (It is stated that tissue
hardened in 5% formalin for 2-3 hr. may be used.)

2. Rinse in distilled water for a few sec. and transfer at once to
the buffered dopa soln. At 30-37° the soln. becomes red in about
2 hr. and sepia brown in 3-4 hr. Do not let the sections remain in
the soln. once it becomes sepia colored since overstaining may result.
Examine from time to time under a microscope to determine the
proper intensity of staining. It is good practice to change to a fresh
dopa soln. after the first 30 min.

3. Wash sections in distilled water, dehydrate, and counterstain
w^ith alcoholic cresyl violet or methyl green-pyronine.

4. Clear and mount in balsam.

5. Run controls by treating tissue as above with the substitution
of buffer soln. for the buffered dopa soln. in step 2.

DOPA AND AMINE OXIDASE 93

Result. Dopa oxidase is indicated by blackening in the sections.
Leucocytes and melanoblasts appear grey or black due to their
dopa oxidase content. Melanin maintains its natural yellow-brown
color, and collagen appears colorless or pale grey.

AMINE OXIDASE

Oster and Schlossman ( 1942) developed a histochemical method
for the demonstration of amine oxidase based on the detection of
the aldehyde formed as the product of amine oxidation. The fuchsin-
sulfurous acid reagent of Feulgen was used for the visualization oi
the aldehyde (page 65). Naturally occurring aldehydes and "plas-
mal" are prevented from interfering with the test by binding them
with bisulfite prior to the application of the tyramine substrate
solution. The diffusibility of the color produced subjects the locali-
zations which may be observed to criticism.

Oster and Schlossman Method for Amine Oxidase

SPECIAL REAGENTS

2% Sodium Bisulfite Solution.

Substrate Solution. 0.5% tyramine hydrochloride in M/15 phos-
phate buffer of pH 7.2.
Control Solution. Omit the tyramine in the substrate soln.
Feulgen Reagent. See page 67.

PROCEDURE

1. Place frozen sections of fresh tissue in 2% bisulfite solution
at 37° for 24 hr. Wash thoroughly and test some of the sections
with the fuchsin-sulfurous acid reagent — the test should be negative
(no color) indicating all free aldehyde has been bound,

2. Incubate sections in the substrate solution for 24 hr. at 37°.
Run parallel controls with the control solution.

3. Immerse in fuchsin-sulfurous acid reagent. -

4. Examine sections when the rapidly formed blue color seems
to be maximum.

Result. Regions of enzymatic activity appear blue, offering a
distinct contrast to the reddish-purple given by "plasmal" (see
page 65).

94 MICROSCOPIC TECHNIQUES

CYTOCHROME OXIDASE

Tests for cytochrome oxidases have been adapted to histochemical
work and a discussion of them has been given by Lison ( 1936, pages
269-290). The enzyme has been referred to as "nadi oxidase" and
"indophenol oxidase/' but Keilin and Hartree ( 1938) have made it
clear that it should be called "cytochrome oxidase" since its catalytic
effect applies to the oxidation of reduced cytochrome. In the
presence of cytochrome c, cytochrome oxidase effects the oxidation of
a mixture of p-aminodimethylaniline and a-naphthol (nadi reagent)
to indophenol, or of p-phenylenediamine to the diimine. The diffusi-
bility of the colored compounds produced must be considered with
reference to localizations of the enzyme in tissue.

In order to check whether a nadi reaction is being given by cyto-
chrome oxidase or some other factor, Moog (1943b) exposed fresh
tissue (chick embryo) to a 0.005 M azide solution in acidified
physiological saline (pH 5.8) for 3 min., and then transferred it to
freshly prepared nadi reagent containing 0.005 M azide. Azide
specifically inhibits cytochrome oxidase. As a control of the possi-
bility that indophenol blue might be reduced to the leuco form as
fast as formed, Moog also placed the tissue in 0.003 M phenylurethan
in saline for 3 min. to saturate the reducing systems, and then trans-
ferred to the nadi reagent containing 0.003 M phenylurethan. The
reagent used by Moog was prepared by combining, just before use,
equal parts of 0.01 M /^-aminodimethylaniline in 1 % sodium chloride,
0.01 M a-naphthol in 1% sodium chloride, and 0.066 M phosphate
buffer. The reaction was carried out at 38° for the interval required
to attain a standard coloration (5-14 min.). Under the conditions
employed, identical results were obtained at pH 5.8 and 7.2.

Graff Method for Cytochrome Oxidase in Fixed Tissue
("M. Nadi Oxidase")

SPECIAL REAGENTS

a-Naphthol Solution. Boil 1 g. a-naphthol in 100 ml. distilled water
and add 25% potassium hydroxide dropwise until the melted
a-naphthol is dissolved. Store in the dark; keeps for at least 1
month.

1% p-Aminodimethylaniline or Its Hydrochloride. Boil to dissolve
solid in the water. Store in the dark; may be used for 2-3 weeks.

CYTOCHROME OXIDASE 95

The hydrochloride is favored because it is more stable.
Nadi Reagent. Prepare just before using by combining equal vol.

of the a-naphthol and p-aminodimethylaniline solns. and filtering.
Strong Ammonium Molybdate Solution or Dilute Lugol Solution.

Concentration not stated.
Dilute Lithium Carbonate Solution. Concentration not stated.

PROCEDURE

1. Fix tissue for two hr. in formalin vapor or in a mixture of
10 ml. formalin and 40 ml. 96% alcohol.

2. Prepare frozen sections and place them on slides which are
then laid in a thin layer of nadi reagent in a petri dish. Oxygenation
of the fluid is effected by careful agitation. After 1-5 min., rinse in
water and examine.

3. Make the color more permanent by treating for 2-3 min. with
dilute Lugol soln. The Lugol soln. converts the blue granules to
brown. Washing sections in dilute lithium carbonate restores the
blue. Strong ammonium molybtlate soln. has been used instead of
Lugol soln.

4. Counterstain with Bismark brown, safranine or alum carmine
and mount in glycerin or glycerin jelly.

Result. Cytochrome oxidase is supposed to be indicated by the
blue coloration.

Graff Method for Cytochrome Oxidase in Fresh Tissue
("G. Nadi Oxidase")

The pH of the nadi reagent must be adapted to the requirements
of the particular material under investigation. Lison (1936, page
274) states that the pH range 7.8 to 8.2 is most suitable for animal
tissues and 3.4 to 5.9 for plant material.

SPECIAL REAGENTS

a-Naphthol Solution. Prepare a 10% alcoholic soln. and just before

use dilute 100 times with distilled water.
0.12% p-Aminodimethylaniline Hydrochloride. Store in the dark.
Nadi Reagent. Prepare just before using by combining equal vol.

of the diluted a-naphthol soln. and the p-aminodimethylaniline

soln.
Buffered Nadi Reagent. Mix the reagent with the suitable buffer

96 MICROSCOPIC TECHNIQUES

(acetate, phosphate, glycine, und carbonate buffers have been
employed) in the respective proportion of 50:10 or 5:20.
5% Potassium Acetate Solution.

PROCEDURE

1 . Prepare frozen sections directly from very fresh tissue.

2. Repeat step 2 in the preceding cytochrome oxidase method,
but wash sections with physiological saline instead of water.

3. Nuclei may be stained with lithium carmine.

4. Examine under a microscope with the section covered with
potassium acetate soln. Permanent preparations cannot be made.

Result. The blue or blue-violet color is also produced in this case.

?9

Loele Method for "a-Naphthol Oxidase

SPECIAL REAGENTS

Naphthol Reagent. Add 10% potassium hydroxide dropwise to a
little a-naphthol in a test tube until the a-naphthol is dissolved.
Add 200 ml. distilled water, and after 24 hr. the reagent may be
used for about 3 weeks.

PROCEDURE

1. Prepare frozen sections of formalin-fixed tissue.

2. Treat sections with the a-naphthol reagent and observe effect
under the microscope within a few minutes.

Result. Violet or black granules which soon disappear are sup-
posed to be indicative of a-naphthol oxidase.

SUCCINIC DEHYDROGENASE

Semenoff ( 1935) gave a method for the localization of succinic
dehydrogenase in tissue sections which depends on the reduction of
methylene blue. The diffusibility of the dye should obviate the
possibility of good localizations by this method.

Semenoff Method for Succinic Dehydrogenase

SPECIAL REAGENTS

Substrate Medium. To 2 ml. 0.05% methylene blue add 2 ml.

SUCCINIC DEHYDROGENASE 97

10% sodium succinate and make up to 10 ml. with A//15 phos-
phate buffer, pH 7.6 to 8.0.
Control Medium. Omit the succinate in the substrate medium.

PROCEDURE

1. Prepare fresh frozen sections.

2. Treat sections 10-15 min. with the substrate medium under
a cover slip, taking care to avoid air bubbles. Seal edges of cover
slip with paraffin to exclude air.

3. Observe under microscope and compare with section in control
medium treated in the same fashion.

Result. Fading of dye characterizes the enzyme activity.

///. PHYSICAL METHODS

A. FLUORESCENCE MICROSCOPY

The detection and localization in tissues and cells of certain
substances by virtue of their flourescent properties when subjected
to ultraviolet irradiation is finding increasing application. Sub-
stances investigated in this manner include naturally occurring
compounds, such as vitamin A, and others introduced into organisms
for experimental purposes, such as 20-methylcholanthrene. The
fluorescence exhibited in tissues or cells may be "primary," i.e.,
produced directly by certain compounds, or "secondary," i.e., result-
ing from treatment with so-called "fluorochromes," fluorescent sub-
stances taken up selectively by particular cellular structures having
no fluorescence of their own. For the most part, the use of fluoro-
chromes is limited to purely morphological studies without regard
to chemical nature and hence need not be discussed here, except as
applied to lipids (page 105). However, for those who may be in-
terested, lists of fluorochromes and their properties may be found in
Haitinger (1938), Jenkins (1937), and Metcalf and Patton (1944).
General discussions of fluorescence microscopy are to be found in
Sutro ( 1936) , Jenkins ( 1937) , Ellinger ( 1940) , Simpson (in Cowdry,
1943, pages 76-78) , and Metcalf and Patton ( 1944) . The books of
Haitinger (1938), Radley and Grant (1939), and Pringsheim and
Vogel (1946) are useful for reference.

1. Apparatus

The set-up of the fluorescence microscope is shown in Figure 3.
Ultraviolet radiation having a high intensity in the range 300-400
m/j. is produced by means of a carbon arc or one of the mercury vapor

99

100

MICROSCOPIC TECHNIQUES

lamps (.4) such as those manufactured by Hanovia Chemical Co.
or the H3 or H4 lamps of General Electric Co. or Westinghouse
Electric Co.

The ultraviolet radiation is passed through a filter (F) to screen
out visible light. A variety of filters may be used for this purpose.

The Corning color glass filter No. 584
( new No. 5840) may be used in combina-
tion with a 5-10% copper sulfate solu-
tion, to which a drop or two of sulfuric
acid has been added, contained in a
quartz or other cell transmitting ultra-
violet rays. The copper sulfate solution
may be replaced by a Corning glass filter
No. 428 (4308), but the latter does not
remove the heat rays as well as does the
solution, and cannot be adjusted to com-
pletely absorb red light. Other filters
that may be employed are a combination
of the Shott glass filters UG2 and BG14,
the Corex filter of nickel oxide glass, or
the Uvet glass filters with a copper sul-
fate solution.
The filtered ultraviolet radiation practically free of visible light
is directed into a substage condenser (QC), made of quartz or ultra-
violet-transmitting glass, by means of a reflector (R) consisting of
either a quartz prism, a polished mirror of aluminum-magnesium
alloy, or, if the ultraviolet intensity is great, the usual plane micro-
scope mirror. Of course, w^hen the apparatus is aligned either
vertically or horizontally on a single optical axis, the reflector is
omitted. In those instances in which the fluorescence is generated
by the longer wavelengths of ultraviolet radiation, as is the case for
vitamin A, it is often possible to employ the ordinary substage
condenser, rather than one made of quartz or special glass, since the
usual grade of optical glass does not absorb much of the radiation
in this range. The condenser may be eliminated entirely if radiation
of lesser intensity can be used. Of the three most common forms of
condensers, the aplanatic gives the best results, although the Abbe
is generally quite satisfactory; achromatic condensers reduce the
intensity of the radiation due to the absorption of their many lenses.

Fig. 3. Diagram of apparatus
for fluorescence microscopy.

FLUORESCENCE MICROSCOPY 101

Metcalf and Patton (1944) suggest the use of a drop of water or
petrolatum on the top of the condenser to serve as a connecting
fluid between condenser and slide in order to obtain illumination of
high intensity when objectives of twenty times magnification, or
higher, are used. With low-power objectives, they point out that it
is necessary to remove the top lens or lenses of the condenser so
that the field can be properly illuminated. Other investigators have
employed sandalwood or Shillaber oil between the slide and con-
denser.

The specimen is mounted on a slide (S) made of an ultraviolet-
transparent glass such as the Corex D slide of Corning Glass Co.
However, Metcalf and Patton ( 1944) have found that ordinary
glass slides of 1.2 to 1.5 mm. thickness may be used when the inten-
sity of the ultraviolet radiation is great. A nonfluorescing medium
must be used for moimting the specimen; glycerol or mineral oil
was recommended by Simpson ( in Cowdry, 1943, pages 76-78) ; but
Popper (1944) reported a disturbing fluorescence from glycerol
(although others beside Simpson have found no difficulty with it)
and the use of mineral oil is limited to substances that will not
dissolve in it, e.g., for vitamin A studies Popper ( 1944) used water
as the mounting medium. In some instances, petrolatum serves as
a good temporary mount, and, for permanent mounts, isobutyl
methacrylate (du Pont), suggested by O'Brian and Hance (1940),
is probably the best. With immersion objectives, sandalwood or
Shillaber oil may be employed as the immersion medium. No special
objectives or oculars are required; however, Jenkins (1937) has
pointed out that some of the older objectives contain balsam that
gives rise to its own fluorescence, and in these cases a darkfield
stop must be used in the condenser to prevent the entrance of direct
ultraviolet rays into the objective.

A filter (EF) that excludes ultraviolet, and passes visible rays,
is placed on the ocular. Either the Corning filters No. 3389 or 3060,
the Leitz No. 8547A, the Bausch and Lomb or the Zeiss Euphos
filter, or a circle of Wratten 2A gelatin filter cut to fit inside the
eyepiece may be used. INIetcalf and Patton (1944) recommend a 5%
solution of sodium nitrite contained in a plane-sided glass cell 5-10
mm. in optical depth which may be conveniently placed on the dia-
phragm of the ocular or, better still, on the diaphragm of the
microscope tube.

102 MICROSCOPIC TECHNIQUES

2. Preparation of Tissues

The preparation of sections from frozen dried material has the
advantage that soluble or diffusible constituents will have no chance
of being lost or displaced from their original sites. However, while
this is the method of choice, paraffin sections of formalin-fixed
tissue have been employed with success in certain instances, although
neither celloidin nor gelatin sections can be used since these media
give rise to fluorescence.

When paraffin sections are employed, fixation is usually carried
out in 5-10% formalin for not longer than 24 hr. The sections are
cut 7-8 /x thick; egg albumin has been used to make them adhere to
the slides. The paraffin is removed by immersion in xylol for 30
min. and the sections are dried at room temperature. In this form
they have been kept for months, and may be examined without
further treatment. All reagents and materials should be the purest
obtainable to avoid adventitious fluorescence of contaminants.

3. Photomicrography

Photomicrographs of fluorescing preparations can be made if
certain precautions are taken. Popper and Elsasser (1941) found
that, in the photomicrography of vitamin A fluorescence, it is pref-
erable to use film of maximal daylight sensitivity. The Fluorapid
film of Agfa-Ansco Corp. is particularly well suited for the purpose
(Fig. 4C,D) . Kodachrome film can be used when the tissue is rich in
vitamin A, but the lower sensitivity of this film limits its value. In
general black-white film is preferable to the color variety.

With substances of fading fluorescence, such as vitamin A, the
time employed for focusing must be kept to a minimum even at
the expense of the sharpness of the picture. The exposure itself must
be short (a maximum of 30 sec, regardless of magnification, in the
case of vitamin A) in order to obtain greater contrast between the
fluorescence and the background. Only one exposure can be made
as shown in Figure 4A,B.

When the fluorescence does not fade, Kodachrome film can give
excellent results. The exposure time has to be determined by trial
in each case, usually falling in the range of 1-15 min. with an
average of about 2 min., according to Metcalf and Patton (1944).
The loss in intensity with greater magnifications makes it impracti-

FLUORESCENCE MICROSCOPY

103

cal to employ magnifications exceeding 500 times. Metcalf and
Patton ( 1944 > also report that with black-white film having a
Weston rating of 50, exposures of from 2 sec. to 5 min., with an
average time of about 10 sec, are required with a 35 mm. camera.
In all fluorescence photomicrography the dark-field stop on the

Fig. 4. A, Human liver showing collagenous fibers in the periportal field
and vitamin A fluorescence in the liver cells. B, Second exposure of the same
field; the vitamin A fluorescence has faded. C, Rat liver photographed with
sensitized film (Fluorapid) ; a large amount of vitamin A fluorescence in the
Kupffer and the liver cells is evident. D, Picture of a liver taken with normal
ultraspeed film. From Popper and Elsasser (1941)

condenser and the filter placed on, or in, the eyepiece or microscope
tube to screen out ultraviolet rays must be used. Otherwise fogging
may occur from the stray radiation. Metcalf and Patton ( 1944)
especially recommend the sodium nitrite filter for color photog-
raphy. In order to switch from ultraviolet to visible illumination
a piece of opal or ground glass is substituted for the filter placed in
front of the light source.

104 MICROSCOPIC TECHNIQUES

4. Characterization of Substances

Direct Observation of Fluorescence

Vitamin A. Particularly intensive work has been done on vita-
min A, starting with the work of von Querner (1935), Hirt and
Wimmer (1940), and others, and continuing in greatly expanded
scope and development with the research of Popper (1944) and
associates. The fading green fluorescence of vitamin A in tissue
sections has been found to run parallel with the results of chemical
determination (Popper and Elsasser, 1941). The fading green fluo-
rescence is characteristic of vitamin A], found in salt water fish, and
a slowly fading pale yellow-brown fluorescence characterizes vitamin
A2, found in fresh-water fish. An admirable review of the studies
made on vitamin A distribution in the tissues of animals and man,
both in normal and pathological states, has been presented by Popper
(1944). He pointed out that carotene may easily be differentiated
from the A vitamins by its very slowly fading green fluorescence
which is apparent only in higher concentrations, and that the bio-
logically inactive anhydro ("cyclized") vitamin A may be recog-
nized by its dark brown fluorescence which gradually becomes a dull
green and finally fades out entirely. Volk and Popper (1944a) re-
ported the existence of a factor in biological fluids, particularly
plasma, and in organ emulsions that delays the disappearance of
vitamin A fluorescence in tissue sections.

Riboflavin. EUinger and Koschara (1933), von Euler et al.
( 1935) , Hirt and Wimmer ( 1939a) , and Metcalf and Patton ( 1942)
utilized the yellowish-green fluorescence of riboflavin for its identi-
fication in tissues. According to Ellinger (1938), and confirmed by
Metcalf and Patton (1942), another form of riboflavin exists (prob-
ably bound to another compound) which gives a yellow-orange
fluorescence. INIetcalf (1943> subsequently concluded that in the
American roach, Periplaneta americana L., the bound riboflavin is
converted to the free form in vivo by the injection of pantothenic
acid or thiamine. With the latter compound the conversion proceeds
more slowly.

Other Vitamins.* Attempts have been made to characterize, and
to determine the distribution of, other vitamins by their fluorescent
properties. Hirt and Wimmer ( 1939b) investigated nicotinic acid
and its amide which they claimed gave a stable yellow fluorescence.

* See Bibliography Appendix, Ref . 29.

FLUORESCENCE MICROSCOPY 105

They reported that in the dry state the amide has a weaker fluores-
cence than the acid, but that in a 1% aqueous solution the reverse
is found. EHinger (1940) was not able to observe fluorescence with
purified solutions of these compounds, and he also disagrees with
Hirt and Wimmer (1939b) that ascorbic acid can be detected
microscopically by fluorescence. The histochemical opportunities of
studying vitamin K by means of its well-known fluorescence are
obvious. One may expect that studies of this nature will be made in
the future.

Lipids. During the course of his work on vitamin A, Popper
(1941) also studied the histological detection of lipids by means
of the fluorchromes: methylene blue, thioflavin S, rose bengal
magdala red, and phosphine 3R. The last stain appeared to be the
best. Further examination revealed that fatty acids, soaps, and
cholesterol are not made apparent bj^ phosphine 3R which, however,
does visualize neutral fat as a silver-white fluorescence on a brown
background (Volk and Popper, 1944b; Popper, 1944). The advan-
tage claimed for this method is that, because of the water solubility
of the dye, more and finer droplets of lipid can be detected than
would be possible by the usual stains. Popper (1944) recommends
the use of a 0.1% aqueous solution of phosphine 3R {Pfaltz and
Bauer) for 3 min. on frozen sections of tissue.

Pigments. The fluorescence technique has been applied to studies
of certain biological pigments. Thus the red fluorescence of porphy-
rins has been employed in histological studies of these compounds
(Lison, 1936, page 256; Ellinger, 1940; Dobriner and Rhoads, 1940;
Grafflin, 1942) . Chlorophyll has been localized microscopically in
plant tissues by Tswett (1911) and Wilschke (1914) by means of
its red fluorescence. The fluorescent properties of bile pigments in
the presence of zinc acetate, and of uropterin, might be adapted to
microscopic studies of these pigments.

Carcinogenic Hydrocarbons. Investigations of carcinogenic
hydrocarbons in tissues have made use of the fluorescence of certain
of these compounds. Graffi ( 1939, 1940) investigated the distribu-
tions in normal and tumor cells of pyrene, benzpyrene, anthracene,
dibenzanthracene, and methylcholanthrene, while Gunther (1941)
and Doniach et. al. ( 1943) confined their studies to benzpyrene, and
Simpson and Cramer (1943-1945) explored the histological localiza-
tion of 20-methylcholanthrene in skin.

106

MICROSCOPIC TECHNIQUES

PQ
<

o

d

o

'3

o

CO

OOO

ooooo

oo

oo oo

oo

oo

OOO

!-<' ^'

c

^

Xi

o

^•bi^

'T ^- bb bhi^

1 1

a 44

bb

"? bb

bbi

i; bb C.

lO

O

>> d tJD

bijD-d-d >i

>> >>

'O-^ >.>>

XiO

-3-D

bC^'O

"^

05
1-H

04

fl

&'tjb>;

ti . . . bb

1 t-l t< tH 1

^'^

C u . .

t-T

i

P*^ 1 1 1 ^

1 1 II

1

bb>:,

f'f-'f

4)

t3

n

ti dbb

-d >) bb bb-d

bb >i

t^. bb bb bb

bbo

>> bb

bb bb >>

'3

a

1 1 r>^

. t," C !m" .
bC^nD,^ bC

>>^

bb

^ bb >,>>

>>

bb >>

>>>.^

ffi

o

S

^ — ; '

1 1 1 1 1

1 l_

"^ '. '. '.

l_

1. I_

'. '. t:

lO

-CiX> bb

>>>>>.>>>>

bb >-.

■d >i bb bb

bbo

>^, bb

bO bC42

d

.

,

bb

,

■§

is?:

bbj2 >^ bb
III 1

t-?'

^ bb bb &
P>-1 1 II

bb

i_ 1

ti ''T

o3

CO

-o d >>

>,>,Ui>,>,

& >.

-d >> >. >.

>.o

>i bb

>> bb^

d

. M

g

3

^.^i

^^■z f ^

1 t

bb . •. bb3
1 bC 1 1

^

bb3

_3-?

H

o
O

O

U5

d-Q >,

>i >^ dbb bb

bb bb

>.!X >■, bb

bbo

>i bb

bb ci. bb

1

__;

<v

d

,

3_.

•

1

Q
o

■§

^^^

bb bC k! . •

1 1

^ bb bb^

3

f-3

^3-?

M

dS >-.

& >)3 bb bb

bb bb

>> d d bb

do

bb d

bb d bb

bb

d
1

1

^

r^

03

lO

32 ^

"?|^ '5«>>^

bb .

& • '-3

1 bC bC-T

3

3

m3-?

q;>

ffi

o

d-d >)

rO t>>X5 bC bC

a^

>> &. a bc

do

bb d

d dbb

^

bb

rH

1

bb

03

d

.3 =!^

-3

Ih*

0)

3

C3

bb bb ;^'?
1 1 ?*-> 1^

bb3

T bC ^^

3-?

3

iiirO bb

M
M

2-d >.

3 >^2 bb3

d-d

>. d 3-3

d>>

ti) d

dd-d

H

bb

CO

d

-3

3

o

O

•g

_.2f

^ f'fi

_;

^bb J3

bb^-

^

tH

lO

bb^

^-T

-!;-°

bb3

o

2-6 ti

3 >.3 >^2

d-d

bb-d 3-d

d >»

3 d

d-do

O M

3

bb

i^

S a

<»

1— »

O

0-J3

.3^

CO

f bb '?!' v^l

bb

33

f bD .X3

_;3

33

3-63

^ >,^^i-i

d-d

hC-3 ^-3

^ >-.

3-d

d-do

c

en

i2

CO '3

;3

33 o

o

CO

73

13 a SB

rr, 2

a

CO

-7! S O

s

s

iS »3 ■»

en

?^ M m

1=1 ^

.3.sJ|

u a 3

i2^

•^03 3

g'd;S&

.2§)J

So-o

cSs

S-Sg

o

-t^ 03 +^

bC >> 3

a >^ =* c >>^

J^ >--^

►3

m >- 3 "

rxi

J3^

o

h5 Pi

107

05

-a

CI
o3

c

'-D
o

CO

I— I °
M

c

O

r3
C
o3

tc
bC

3

CO

O

o3
>

O
S

CO

<u

o

3

c 1 .
o^-S in

O ^ '
O

-J o

o o ooooo ooo

ooooooooooooooooooooo
ooooooooooocooooooooo

+ +
+ + +++
++++++++

^^

>1 >-, Crb >! I

-1 CI, >-l >;^^

bC

t.1 ^H fc. .

s

U5

o
o

+
+ ++ + +++
++++++ +++
+++++++++■

+ +4- +

++++++++++

■ r I

hC

^^

bC

bC

'-" o o ^-

>> >^, i; >..Q >; bb >-. >^ J >. >^, >-. L." ^ ^ g.^ >. >; d

u

.»^

d

a
o

d

U

Ui

h

o

to

o

I— 1

U

+++ ++ +++++ +++++++
+++++++++++++++++++++
+++++++++++++++++++++

-. !; he
.-2 •

i'^ bC

bC

u
^

+ + + + + +++++

+ + + ++ +++++++
+++ I ++++++++ I +++++++

hC

bC

b£

,C >, bCJ2 >> C5- >i >>

hr

t-'

a

fl

■*s

o

O

J3

«

O

S

a)

o

O

+
+

+
+
+

bit
I

bC

.3 o

"» 2 !;

0)0.0

2o"

hi S

+

+ + + +

o
O

X! Cr^ bJD

So

2 >' bbS

^H bC bC"Ci X) >i >i

+

+ + + +
+ + + +
+ + + + +

t^ bJC bC-3 .Q

+ +
+ + + +

+ + + +

++++++++

bH

■^ >, C c nxiX^j^

'^ a

IM COCCI 00

I I ■ ■

I 1

F— I CD
■ GO "*

c
I

mo 00(M C5 "-I
(M t^ iC »^ Tf CO
■^ (M ■— I Cj ^ (N

o
oo

1

CO

CD

CO

I

0(M

C: O
-- (N

3

Q

o

•n

'S
o

O

0.3 3 -f^

■73

o

o

<D

rn t_ m M • '-^

L. (A

o -^ o o

O O fc< w

-<<<<:QQ(ilWKPwPMPH

^J . -1 . -H 3

5 P-^^ o

ii S ^ O ^ w

3.g aj o3 (U

Si-

■§1

PhCli

1

^ 1

^

._.

u.

^ — '

ca

d

0)

>^

«

o

o

_o

^

a

i_,

S3

03

H

fl O

O O
o3 03

03

s

3

'3

£S

o3 o3 O 3

5J

c

&.— .
!~ e

a

T3 N
o3 03

0:7:

«
>>

^ o

O M
. o

II

A

-M

ro

bC

bO

tc

bC

+

II

C4-t

b<

hr

Xi

-^j

f-l

rr

III

3

C

S

o-c

-M

^

II

.^

,

cS

it

■^

C

bC

c

H

■s

>

X!

+

II

>v

^

tH

-0

^

^

^

0;

J-

03

OJ

s.-^

CO

II

-t-'

,

r/i

a o3

-4—

0)

C

3

n

^

«

OJ

II

-t^

J2

r/1

^ ,

en

rn

0)

r!

e

.^^

r!

f^

>

7J

^

aj

XiA

-<H

108 MICROSCOPIC TECHNIQUES

Chemotherapeutics. Helander ( 1945a) has reported the detec-
tion of chemotherapeutics in paraffin sections prepared from frozen
dried tissue. The blue fluorescence of tissue itself, when illuminated
by ultraviolet, may obscure the fluorescence of compounds present
if they too have a blue fluorescence. This difficulty can be circum-
vented in some cases by heating the sections for a short period in
order to change the color of the fluorescent light emitted by the
drug. In a work of a high order of excellence Helander (1945b)
presented a table (Table I) of the fluorescences of various tissues,
and after intramuscular injection of drugs into mice, he found that
sections of muscle tissue showed the fluorescences indicated in
Table II.

Fischl and Singer (1935) made use of the fluorescences of atebrin
and trypaflavin to show that these compounds are taken up in vivo
by trypanosomes. A table of the fluorescences of compounds, and of
certain parasites after treatment with them, has been given in a
paper by Bock and Oesterlin ( 1939) .

Metals. The detection of uranium salts in incinerated sections
of tissue has been made possible by the fluorescence of these salts,
(page 145).

The colors of the insoluble 8-hydroxyquinoline derivatives of iron,
calcium, magnesium, aluminum, manganese, zinc, and copper have
been used for identifications of these metals in tissue sections (see
page 21). However, no one seems to have exploited the possibility
of utilizing the fluorescences of these derivatives. DeMent (1942)
lists the fluorescent colors of 27 metallic hydroxyquinolinates, and
it would appear that, in some cases at least, a more clear-cut
differentiation between the metals would be possible on this basis.

Spectroscopic Analysis of Fluorescence

The identiflcation of substances in tissues by means of the spec-
troscopic analysis of the fluorescent light they emit has been utilized
to some degree. Dhere ( 1939) presented a thorough and able review
of this technique and gave data relative to various carbohydrates,
lipids, proteins and amino acids, alkaloids, chlorophylls, bile pig-
ments, porphyrins, carotenoids, flavins, thiochrome, and a few other
special compounds. Either direct vision or grating microspectro-
scopes may be employed, and the spectral lines of mercury can be

EMISSION HISTOSPECTROSCOPY 109

used conveniently for calibration after the filters are removed, since
a mercury arc is the source of the radiation.

More recently Sjostrand (1946a,b) has employed this technique
for the localization of riboflavin and thiamine in cells. Sjostrand
used the freezing-drying treatment on this tissue (frog eyes, etc.)
followed by paraffin infiltration. The 3-5 [x sections were subjected
to ultraviolet illumination on the stage of a microscope and the
fluorescent light which resulted was passed from the ocular into a
spectroscope for analysis.

B. EMISSION HISTOSPECTROSCOPY

The technique for the identification of certain elements in selected
histologically defined portions of tissue by means of emission spectra
was developed independently and at about the same time by Gerlach
and Gerlach, and by Policard. Modifications were made later chiefly
by Scott and Williams. The emission spectra are obtained from the
radiation produced by consuming, in high-frequency spark, a chosen
cellular region of a tissue section or a selected bit of biological
material. The amount of tissue destroyed in sections is a disc of
about 1 mm. diameter and 0.1 mm. depth; this would have a weight
of approximately 0.08 mg. and hence elements giving good spectral
lines can be detected in this quantity of fresh or fixed tissue. Ele-
ments that have been identified by this technique include potassium,
sodium, calcium, magnesium, manganese, aluminum, iron, zinc, lead,
copper, mercury, silver, gold, carbon, silicon, phosphorus, and boron.
Attempts at quantitative work have been made, but the technique
is primarily qualitative at present.

There is little essential difference between the apparatus developed
by Gerlach and Gerlach (1933) and Policard (1931-1932, 1933a).
The former employed a high-frequency generator that was superior
to the one used by Policard, and the latter used a metal slide to hold
the tissue section instead of the glass slide used by the Gerlachs.

1. Policard Technique

A frozen section, 50-200 fi thick, cut preferably from fresh tissue
or else from alcohol or formalin-fixed material, is placed on a slide
of pure platinum or gold which is fastened to the microscope stage

no

MICROSCOPIC TECHNIQUES

(Fig. 5). A platinum or gold needle embedded in a cylinder of
insulating material, which is fixed in one of the objective holes of
the microscope's revolving nosepiece, serves as a movable electrode.
The circuit is completed through the frame of the microscope which
is grounded. The cellular region to be examined is chosen with an
objective in position, and then the objective is swung out and the
electrode needle brought into place. The needle should be in line with
the optical axis of the microscope so that it is centered over the

Fig. 5. Arrangement of Policard's
apparatus for emission histospectros-
copy: 1, Microscope stage under re-
volving nosepiece bearing tlie upper
electrode; 2, high-frequency appara-
tus — dermatological type; 3, Hilger
quartz spectrograph; 4, cylindrical
quartz condensing lens; 5, optical
bench; 6, ground lead; 7, current
source; 8, upper electrode lead. From
Policard (1931-1932)

selected spot. The distance between the point of the needle and the
section is fixed between 1-2 mm. When the circuit is closed, the
radiation from the spark produced is condensed by the quartz lens
and directed through a 0.1-0.2 mm. slit into the quartz spectro-
graph. For a strong spark, a period of not more than 1 sec. seems
to be sufficient for the exposure of the photographic plate on the
spectrograph, but for weak sparks 20-30 sec. may be required
according to Policard. The best spectral lines were observed in the
ultraviolet range, 100-300 m/x.

2. Scott and Williams Technique

Scott and Williams (1935) modified the apparatus chiefly by
screening the electrodes, placed 1.4 cm. apart, in order to practically
eliminate electrode lines from the spectra and to obviate the need
of cleaning the electrodes often. Since these workers did not employ
needle electrodes mounted on a microscope, they could not make
as fine a histological selection of the sample as the European investi-
gators. Scott and Williams (1934) minimized this point and claimed
that removal by fine dissection of the bit of material to be studied
is sufficient. In most instances this would be true, but there are

EMISSION HISTOSPECTROSCOPY

111

conditions conceivable in which the needle electrode assembly
might have advantages. In any case the Scott and Williams appara-
tus has a nmnber of significant improvements, as will be apparent
from the following description of it.

The spark generator construction is shown diagrammatically in
Figure 6. Oscillations are produced in the primary circuit by a
spark passing between two thick brass discs rotated by a small
motor to prevent erosion and local heating. The space between the
discs is supplied with photoelectrons by focusing ultraviolet radia-
tion on it from a quartz mercury lamp. The analyzer spark in the
secondary circuit has an improved continuity and steadiness as a

o "

—I i£>

g >

, o
. o

: o
:7

O rsi OJ

o > j: -o

g ™ m e

o fu (/I y^
1^ ^

Quartz

mercury arc

Fig. 6. Wiring diagram of spark
generator assembly. The ionizing beam
of ultraviolet radiation is shown
focused between the rotating discs of
the spark gap. From Scott and Wil-
liams (1935)

Fig. 7. View of box enclosing ana-
lyzer spark gap and electrode screen.
Portion of spectrograph seen at right.
From Scott and Williams (1935)

result. A large open type oscillation transformer is employed having
two heavy primary, and twenty smaller secondary, turns wound
on a wooden frame as close to the edge of a 36 in. square as possible.
A clearance of 0.5 in. allowed between the secondary turns. The
condenser is composed of a stack of glass plates (8 X 8 X % in.)
separating eight sheets of aluminum foil (6 X 6 in.) connected alter-
nately. The current passing over the analyzer spark gap is about
0.45 ampere.

The electrodes of the analyzer spark gap are steel balls (0.25 in.
diameter) placed 1.4 cm. apart which are supported through the
bakelite back of a metal box housing fitted with quartz and glass

112

MICROSCOPIC TECHNIQUES

windows, removable for cleaning. An aspirator tube is fitted into
the top of the box to remove vapors and the box itself is placed on
alignment pegs to maintain fixed optical relations. The disposition
of the spark gap relative to the spectrograph is shown in Figure 8.
As mentioned previously, the horizontal slit in the side of the box
facing the spectrograph serves to screen out electrode lines from
the tissue spectra and to obviate the need of cleaning the electrodes
oftener than once a month. The end of a 2 in. length of Pyrex tubing
(2-3 mm. inside diameter) or rod (2-3 mm. diameter) is centered
in the spark gap. It has been shown that the glass does not affect
the spectra. Bits of cellular material (3-6 [A.) are placed on the
end of the glass so that they will be in the center of the spark. A

To spark generator

Small Pyrex tube carrying bit of tissue
l"

1.4 cm. between faces

Spectrograph collimator

14.0 cm. *^

\ *.> \ < — 5.3 cm. — »4<

1.3 cm.-^

Fig. 8. Diagram of disposition of the spark gap

relative to the spectrograph.

From Williams and Scott (1935)

film of purified Eastman gelatin may be helpful in effecting the
adherence of dry material. A strip of tissue, 5 mm. or more long,
may be placed in the glass tube and fed into the spark with a push
rod; and in a similar manner liquid taken up on a small strip of
ashless filter pulp can be subjected to test. In the latter case, con-
trol spectra for the pulp alone must be obtained.

A Gaertner L250W quartz prism spectrograph taking 3i/4 X 4l^
in. photographic plates was used. The spectrograph slit was fixed at
0.05 mm., and Eastman "50" plates were employed for high sensi-
tivity and Eastman Process plates for high contrast. Thirty ex-
posures may be made on each plate. Two microscopes with a com-
parison ocular were utilized for comparing the positions and inten-
sities of lines on different plates.

Faint iron lines from the electrodes are apparent in the spectra

VISIBLE AND U.V. ABSORPTION HISTOSPECTROSCOPY 113

SO that, if iron is to be investigated, another electrode metal should
be used. For an adequate exposure (15-30 sec), 2-4 /xl. of tissue
are usually sufficient. When tissue is subjected to fixation, a control
experiment is necessary to test the fixing fluid spectrographically;
the filter pulp method may be employed for this purpose.

The Williams and Scott ( 1935) photoelectric apparatus for dark-
field photometry and densitometry has been used to determine the
intensity of the spectral lines in order to obtain a more quantitative
estimation of the elements. A description of this apparatus is given in
the section dealing with microincineration (page 146) . The blackness
of the photographed spectral lines is measured by placing the photo-
graphic plate on the mechanical stage, adjusting the reflecting prism
so that the light emerging from the ocular is reflected directly down-
ward on the slit of the photocell box, and observing the galvanometer
deflection after the proper focusing has been made. The deflection for
an unexposed portion of the photographic plate is then taken.

C. VISIBLE AND ULTRAVIOLET ABSORPTION
HISTOSPECTROSCOPY*

The application of the quartz microscope to measurements of the
absorption spectra of cellular components in situ, particularly as
developed and applied by Caspersson and co-workers at Karolinska
Institutet, Stockholm, offers a new and promising approach to the
solution of many histo- and cytochemical problems. The ingenious
apparatus of Caspersson (1940), subsequently modified by Gersh
and Baker (1943), has already yielded valuable information con-
cerning the nucleic acids of chromosomes (Mirsky, 1943), the nature
of thyroid colloid (Gersh and Baker, 1943), and chemical character-
istics of the Nissl bodies in nerve material (Gersh and Bodian, 1943-
a,b). (It should be pointed out that the absorption method cannot
differentiate between ribonucleic acids and desoxyribonucleic acids
since their absorption spectra are almost identical, having maxima
at about 260 m/i. However, the differentiation can be made qualita-
tively by staining reactions, (pages 65, 66).

The great advantage of studying cell structures m situ by this
technique is made particularly impressive by the fact that the
spectral measurements can be carried out on quantities of material

*See Bibliography Appendix, Ref. 31.

114 MICROSCOPIC TECHNIQUES

down to 10"^ /tg. Investigations may be made on selected structures
in microtome sections or on mechanically separated cellular com-
ponents. When sections are to be employed, they are best prepared
from tissue embedded in paraffin or celloidin after freezing-drying
treatment. The technical requirements in absorption spectra meas-
urements by means of the ultraviolet microscope have been examined
critically by Cole and Brackett (1940). Laviri' (1943) simplified
the focusing of the ultraviolet microscope by using a willemite
screen which produces a visible image with ultraviolet illumination.

The absorption technique applied in situ has certain drawbacks
that should be considered, elegant though the technique is. Thus,
Danielli ( 1946a) has sounded a warning that hazards exist in as-
cribing to particular substances the absorptions found in different
parts of a cell. Effects of molecular interactions and interferences
by other substances are possibilities that are not to be ignored. Hence
the method will be of greatest value when the results are interpreted
with appropriate regard to these limitations.

A most ingenious microscope arrangement for the colorimetry of
0.5-1.0 ix\. drops of liquid was developed by Norberg (1942), also
at Karolinska Institutet, and applied by him to the measurement of
phosphorus in quantities down to 0.5 n\[xg. This technique requires
the removal of the specimen from the rest of the tissue and its
chemical treatment to yield a solution which can be subjected to
absorption analysis.

Stowell (1942) designed an apparatus for the measurement of the
amounts of stain or pigment in tissue sections. For the measurement
of stained constituents the quantitative significance of the method
depends on the degree of correlation between the amounts of the
stain and the substance for which the stain is specific, a correlation
often poorly defined. Stowell applied his technique to the estimation
of desoxyribonucleic acid by means of the Feulgen stain.

1. Casper sson in Situ Technique

A diagrammatic representation of the apparatus is given in Figure
9. The source of radiation may be either the Philips water-cooled
super-high-pressure mercury lamp (A), a tungsten lamp (B) sup-
plied with current from storage batteries, or Kohler's rotating elec-
trode spark gap (P) . The mercury lamp may be used for radiation in
the visible portion of the spectrum and in the ultraviolet range down

VISIBLE AND U.V. ABSORPTION HISTOSPECTROSCOPY

115

to 230 mix. However, since variations both in voltage and water pres-
sure affect the mercury lamp, the tungsten is employed for the
longerwave ultraviolet range. The spark source is used for wave-
lengths shorter than 235 m/i,.

The radiation, after passing through a monochromator (C) is
concentrated on the object (/) on a quartz slide by the condenser
(//). (The second monochromator slit is at D, lens at E, and a 90°
ciuartz prism at F.) For ultraviolet work a fused quartz condenser
is used, and in the visible range a good achromatic type is employed.
Zeiss objectives designed for the longer ultraviolet (down to about
340 mju) or the fused quartz objectives of Kohler and von Rohr for
the shorter wavelengths are used (K) . For routine work down to

V^

i^ o

Fig. 9. Arrangement of Caspersson's apparatus

for photoelectric absorption histospectroscopy.

From Caspersson (1940)

240 m/A, Caspersson ( 1940) found it convenient to have a glycerin
immersion lens corrected for 257 m/x with an aperture of 1.25, one
corrected for 275 m^a, and a long-wave ultraviolet recorrected
apochromat. Oculars with iris diaphragms (L) were employed. The
lenses in them were of quartz for the ultraviolet work.

A 90° quartz prism (M), adjustable by means of micrometer
screws, is used to deflect the radiation through the opening of an
electrically driven rotating sector (A^) on to a photoelectric cell
(R). The prism can be replaced either by a Kohler focuser (Y) or

116 MICROSCOPIC TECHNIQUES

a photomicrographic camera (X). Every object measured must be
photographed in order to estabhsh its exact position and dimensions.
The photocell is connected to a string electrometer and both of these
instruments are well shielded and also protected from moisture by-
means of phosphorus pentoxide. Various photocells are used for
different wavelengths; gas cells are usually employed. For the
shortest ultraviolet, cadmium; for medium ultraviolet (260-350 m/x),
sodium; for long ultraviolet and visible (350-550 nifi) , potassium;
and for wavelengths over 550 m/x, potassium-cesium cells are used.
The telescope (0) is placed in front of the photocell to control the
optical centering of the system ; this centering must be very exact.

It is necessary to compensate for variations in the source of
intensity of the radiation, and for this purpose a quartz plate (G)
is interposed in the optical path in order to reflect a small percentage
of the radiation on a photocell ( V) . Readings of the changes in the
photocell current can be used to correct the readings of the electrom-
eter (S) . (T and U are leak resistance and four-step potentiometer,
respectively.)

Measurements are made by taking the deflection of the electrome-
ter with the object in position and in focus, and then moving the
object away so that a clear space on the slide lies in the optical
axis. The opening in the rotating sector is reduced until the amount
of radiation striking the photocell is the same as before, i.e., the
same electrometer deflection is produced. The absorption in the
object will then be equal to the decrease effected by the sector, and
extinction coefficients may be calculated.

Gersh and Baker Modification. A somewhat simplified set-up,
with American-made instruments, is employed by these investiga-
tors, as may be seen in the diagram of their apparatus (Fig. 10) .

The source of radiation is a Daniels and Heidt (1932) type of
medium pressure mercury arc in a quartz capillary tube which is
mounted about 1 cm. from a quartz window in a large copper box.
The lamp is water cooled, and since the rate of cooling affects the
radiation output, the water line is equipped with a pressure regulator.
The lamp consumes 500-700 watts from a 220 volt D.C. line; a
ballast resistance is placed in series with the lamp.

The entrance opening of the monochromator is a circular hole of
about 0.8 mm. diameter in a thin sheet of copper fixed 0.5-1.0 mm.
in front of the arc, in the water bath. The two equilateral quartz

VISIBLE AND U.V. ABSORPTION HISTOSPECTROSCOPY

117

prisms of the monochromator are 5 cm. long and 4 cm. high; the
table on which they are mounted can be rotated by means of a
slow-motion screw. The collimating lens has a focal length of about
8 cm. and the telescope lens about 80 cm.; both are held in adjustable
brass mountings and these with the prism table are fastened to an
iron plate on leveling screws. The whole apparatus is enclosed in a
wooden box with the required apertures. A spectrum from the first

Photoelectric cell

Amplifier Lii
Galvanometer ( o

Ultraviolet
light source

^

Quariz
monochromator

CENTRAL

BEAM

k

tf 37

Quartz obiective

Quartz slide
and cover slip

Quartz condenser

^>^ — - — y^

^Quartz ^

reflector

Fig. 10. General arrangement of light source,
monochromator, microscope, and measuring device
for determining the absorption of ultraviolet Hght
by minute volumes of tissue. From Gersh and Baker
(1943)

prism, without entering the second, passes out of one of these
apertures to fall on a calibrated wall scale 2 meters distant. By this
means the prisms can yield any chosen wavelength when manually
adjusted.

A Zeiss quartz microscope of Kohler design is mounted on a
leveling table so that it can receive the radiation reflected by a
quartz right-angle prism. In order to fill the objective field, the
substage condenser is focused on the field of the telescope lens rather

118 MICROSCOPIC TECHNIQUES

than on the image of the entrance opening of the monochromator.
The diameter of the condenser diaphragm is set at 6 mm. to insure
sufficient spectral purity.

Both photoelectric and photographic recording may be employed.
The quartz photoelectric cell is a No. FJ-405 of General Electric Co.,
and it is mounted 56.5 cm. above the ocular in a large brass cylinder.
This cylinder has a quartz window close to which is a frame that
holds a series of circular slits. The cylinder is horizontally adjustable
so that a slit can be brought into the optical axis of the microscope.
The brass cylinder also contains a type D96475 electrometer tube of
Western Electric Co., and a 10^" ohm SS white grid resistor. The
output from the photocell is connected to the grid of the tube which
is included in a Penick amplifier circuit maintained on three storage
batteries of large capacity. The amplified current is measured by a
Leeds and Northrup type R galvanometer with a scale 1.5 m. from
the galvanometer mirror.

In order to bring the photocell into adjustment in the optical
axis of the microscope, the shadow of an ocular cross hair is pro-
jected by means of "white" light on the photocell aperture, which
is then adjusted until its center and the center of the image conin-
cide. The cross hair in a fluorescent finder placed above the ocular
is adjusted similarly with "white" light and ultraviolet radiation.

For the measurement of absorption curves of larger uniform
objects, the object is centered in the cross hair of the finder, the con-
denser is focused on the plane of the telescope lens, the objective is
adjusted to give a sharp image, and the current generated in the
photocell is measured. Then the object is moved away so that only
the clear slide is in the optical path and another measurement is
made. From these data the percentage transmission and the extinc-
tion coefficient can be calculated. The measurements are then re-
peated at each wavelength chosen. For studies of thyroid colloid,
Gersh and Baker (1943) used a 6 mm. objective, lOX ocular, and a
photocell aperture of 11.9 mm. With these optics the light transmis-
sion was measured through a tissue area of 143 /x- and a volume of
2861 /i,^. For measurements on cytoplasm and nucleoli the respective
systems were 6 and 1.7 mm. objectives, 14 and lOX oculars, and
8.73 mm. photocell aperture in both cases.

When the measurements are to be made on smaller and less
homogeneous objects such as Nissl bodies, a different and more

VISIBLE AND U.V. ABSORPTION HISTOSPECTROSCOPY

119

accurate technique is used. A quartz plate 1 mm. thick and 2 cm.
square is placed in the path of the monochromatic beam at an angle
of 45° and at a distance from the telescope lens of 46 cm. (Fig. 10 L
By this means 6-7 % of the incident radiation is reflected to another
photocell mounted at a distance from the quartz plate equal to that
of the aperture of the microscope condenser. The photocell output is
amplified by a Huntoon (1935) direct-current amplifier and passed
through a Leeds and Northrup type P wall galvanometer. Simul-
taneous readings are taken on the same scale, from both this gal-
vanometer and the one used for recording transmission, at each
wavelength without moving the object, but with careful focusing of
condenser and objective at each wavelength. After these readings
are obtained the object is moved away and the readings for the
clear quartz slide are taken. The data are used to calculate the
transmission and extinction coefficients as usual.

l.U

_ 1

' ' L

0.5

I A

/

0.4

-

=j^

0.3

1

/

-

0.2

^/'^^'^

/
/

/

-

0.1

X 2^

— •

- •

/ 1
^^ 0.5

0.1

^ B

-^^'

-

U.Ub
0.04
0.03

nn?

-;^Z^'

-'' 2

1

400

300

! 1

1 !

in

ro

o CM

n CO

rt o

ro m

'J- ^

(T> o

00 00

C\J CM

cvj i^ m o

ID on CO o

^ Ln ^ ^

CM CM CM CM

WAVE LENGTH, A

Fig. 11. Absorption curve of colloid in a single
follicle of a thyroid gland in alkaline (^1) and acid
(.42) medium, as compared with the ultraviolet ab-
sorption curves of extracted sheep thyroglobulin
made by Ginsel in alkaline (fil) and acid medium
(B2). From Gersh and Baker (1943)

The reliability of the technique is shown by Figure 11, taken
from Gersh and Baker (1943). The curves in the inset (B) were
obtained by Ginsel (1939) for extracted sheep thyroglobulin in both
acid and alkaline media, while those in (A) were established by
Gersh and Baker by their histospectrographic technique on the
colloid in a single thyroid follicle.

120

MICROSCOPIC TECHNIQUES

2. Norberg Technique

Apparatus

A diagram of the optical system is given in Figure 12. For work in
the visible range, the light source employed is a 100 watt tungsten
band lamp (A) supplied with current from a large capacity (150-
200 amp. hr.) storage battery. Monochromatic light is obtained from
a Winkel-Zeiss monochromator (B) . (C is second monochromator
slit.) A filter (F) may be used between the condenser (E) of the
microscope and the sample slide (G) . The microscope objective is
indicated by H and the ocular by 0. During measurement the ocular
is removed and the light is reflected by the prism (/) , which is mov-

Af L .

/

h!

H

,.G

^7

KyE

Fig. 12. Microphotometer.
From Norberg (1942)

Fig. 13. Photocell amplifying circuit.
From Norberg (1942)

able about both a horizontal and vertical axis, to the photocell (L).
Potassium cells are employed for wavelengths 450-550 m^u, and
potassium-cesium cells for longer wavelength. The current generated
in the photocells is amplified by circuits in M and then conducted to
the galvanometer (A^). Details of the amplifying circuit will be con-
sidered subsequently.

For measurements in the ultraviolet region a high-pressure mer-
cury lamp (P) with movable prism (S) and 90° prism T, of the
Philips Philora H P 300 type, is used with a spectral filter {R) .
Variations in the intensity of the radiation from the mercury lamp

VISIBLE AND U.V. ABSORPTION HISTOSPECTROSCOPY 121

are coaipensated by casting a portion of the radiation on photocell
Lc by means of the semireflecting glass (D). The current generated
in this photocell is amplified and made to oppose that from photo-
rell L.

The galvanometer is employed as a null-point instrument and the
amount of radiation striking the photocell L is controlled by the
rotating sector (K) . The rotating sector with its motor is mounted on
a slide, adjustable by a rack and pinion arrangement. The radiation
passing the sector can be controlled over a greater range (0-50%)
by the use of two discs, each with a 90° segment removed, which are
made to rotate in opposite directions. Kortiim (1934) has described
a simple sector wdiich operates on a similar principle. Norberg finally
employed a sector patterned after the Askania-Werke (Berlin)
model, which enables adjustment and reading while the sector is in
action, and the accuracy obtained in the absorption is 0.025%.

The amplifying circuit for the photocell current is shown in Figure
13. It is a modification of that described by Custers (1933) and it
utilizes two Philips 4060 electrometer tubes. The apparatus can be
used with either the single photocell (L, Fig. 12), employing com-
pensation with the potentiometer, or with both photocells (L and
Lc) . The galvanometer used by Norberg was a Zernike C (Kipp and
Zoonen) instrument having a tension-sensitivity of 10,000 scale divi-
sions per volt and a stability level of 3 X lO^^-'' amp. By means of the
Ayrton shunt ( V) (Fig. 13) the sensitivity can be reduced by 0.1 and
0.01. The galvanometer is shown at iV; the potentiometer for com-
pensation where only one photocell is used at U; Re resistance
= 1.2 X 10^^ ohms; R© leakage resistance = 1.1 X 10^*^ ohms.

Other resistances are marked in ohms. The apparatus must be
mounted within a Faraday cage to avoid electrostatic disturbances
and unsoldered contacts must have large frictional contact surfaces.

Manipulations

The filament current in the amplifier is turned on 30 min. before
measurements are made, and the tungsten band lamp is also turned
on long enough in advance to attain steady illumination. The sample
slide on the microscope stage is adjusted so that the image of the
exit slit of the monchromator ajipears in the middle of the sample

122 MICROSCOPIC TECHNIQUES

drop. The size of the image is 0.12 X 0.12 mm. The slit image is pro-
jected over the opening of the measm-ing photocell (L, Fig. 12) by
the objective and the prism (7). The sector is started; the shutter in
front of the photocell is opened; and the photocurrent is compen-
sated by a potential applied by means of the potentiometer ( U, Fig.
13) to the grid connected to the other photocell so that the zero read-
ing on the galvanometer may be obtained. The galvanometer is
usually constant within 0.5 mm. after 1-2 min. The sample slide is
now shifted so that the light will pass through solvent alone or a
suitable blank. The sector is adjusted until the galvanometer zero
reading is again obtained. It is well to repeat the measurements
several times.

When the mercury lamp is used as the source of radiation, the
photocurrent from the measuring photocell is compensated by the
photocurrent from the other cell which is illuminated by the semi-
reflecting glass as previously mentioned. Thus, with the mercury
lamp the compensation current from the potentiometer is replaced by
the compensating photocurrent.

The Sample Slide for Absorption Measurements

Very clean microscope slides and cover glasses are coated with
hydrophobic films of nitrocellulose in order to prevent the aqueous
drops from spreading. This is accomplished by pouring over the glass
a solution of 1 g. of highly nitrated cellulose (about 13% nitrogen),
0.1 g. diethyl phthalate, and 0.01 g. butyl stearate in 100 ml. butyl
acetate. The glasses are set aside in a tilted position to drain and
dry for at least 3 days in a dust-free place. A drying drum with an
electric fan may be used to reduce the drying time. If plastic plates
are substituted for the glass, no film is required, but care must be
taken that the plastic is optically homogeneous. Two narrow strips
of polished glass having a thickness of about 0.35 mm. ( from a hemo-
cytometer cover glass) are placed on the hydrophobic film parallel
to one another. Between them, sample and blank drops of the order
of 0.5-1.0 ju,l. are pipetted, and paraffin oil is added to fill the area
between the glass strips. A cover glass is set on the glass strips to
complete the cuvette. To determine the layer thickness in the
cuvette:

VISIBLE AND U.V. ABSORPTION HISTOSPECTROSCOPY

123

1. Draw two thin parallel lines 1 cm. apart on the upper surface
of the slide with India ink.

2. In a similar manner, draw two lines 3-5 mm. apart on the
under surface of the cover glass.

3. Place the cover glass so that the lines on the two glass surfaces
intersect.

4. Measure the distance between the upper and lower lines at the
four points of intersection by focusing on the lower line with the
microscope and then using the micrometer screw to focus on the
upper line.

5. Multiply the distance obtained with the micrometer screw by
the refractive index of the paraflBn oil to get the thickness of the
layer.

Accessories for Norberg Technique

Quartz or Supremax Glass (Schott, Jena) Needles. For the

isolation and incineration of the sample, fine needles with slightly
thickened points are used. The needles are cleaned by boiling in 2 A''
nitric acid and rinsing with distilled water.

Fig. 14. Apparatus
for microhydrolysis : A,
Steam mantle ; B, cham-
ber of hydrolysis; C,
holder with quartz
needles; D, water.
From Norberg (194£)

Hydrolysis Chamber. The arrangement shown in Figure 14 is
used for the hydrolysis of certain constituents in the ash after the
sample has been incinerated on the tip of a quartz needle. For the
hydrolysis of pyro- and metaphosphate to orthophosphate, Norberg
placed 1-2 /xl. oi 1 N hydrochloric acid on the tip of each needle and
heated for 1 hr. at 100° in the chamber. Should the acid evaporate
in the chamber, the hydrolysis must be repeated. After hydrolysis
the drops are allowed to evaporate to dryness at room temperature.

Muffle Furnace. An ordinary muffle furnace may be used for
the ashing of the sample on the tip of a quartz needle.

124 MICROSCOPIC TECHNIQUES

Methods
Phosphorus

By means of his microscopic photometric technique (page 120)
Norberg ( 1942 ) developed a method for the estimation of phos-
phorus in quantities down to 0.5 nijug. with an error not greater than
about 20% for single analyses. Naturally the error is less with larger
samples and greater accuracy is obtained by averaging the results
of several determinations. For amounts of phosphorus under 1 m/^g.,
Deniges' stannous chloride method is recommended, while for 1 m/^g.
and more it is preferable to employ Fiske and Subbarow's amino-
naphtholsulfonic acid method, which is technically easier.

Norberg Method for Phosphorus

SPECIAL REAGENTS

0.005 N Calcium Acetate.

1 N Hydrochloric Acid.

Deniges Reagents, (a) 0.01 M sodium molybdate in 0.6 .V sulfuric
acid; (b) 0.2 N stannous chloride in concentrated hydrochloric
acid, approximately 2.5% SnC^-HoO.

Fiske and Subbaroiv Reagents, (a) 0.022 M sodium or ammonium
molybdate in 0.75 N sulfuric acid; (6) dissolve 12 g. sodium meta-
bisulfite in 80 ml. water, stir in 0.2 g. 1,2,4-aminonaphtholsulfonic
acid (some commercial preparations of this compound are not
suitable, that of British Drug Houses Ltd. proved to be good) and
add 2 ml. 20% crystallized sodium sulfite. Let stand overnight
and filter off the undissolved aminonaptholsulfonic acid. Store in
a dark bottle.

PROCEDURE

1. Obtain the sample on the tip of a quartz needle (page 123). If
the sample is liquid, pipette 1 /il. 0.005 N calcium acetate onto the
needle tip with the sample and allow to dry in the air. If the sample
is solid, the calcium acetate is placed on the tip and allowed to dry
before taking on the sample. The excess calcium is required to pre-
vent loss of phosphorus during the incineration.

VISIBLE AND U.V. ABSORPTION HISTOSPECTROSCOPY 125

2. Place the needle in a cold muffle furnace and turn on the heat.
\Mien the temperature reaches 500° turn off the heat and remove the
needle when the furnace is cool. As an alternative, place the needle in
the furnace at 500° and remove it after 20-30 min. at this temper-
ature.

3. Hydrolyze the pyro- and metaphosphate to orthophosphate
by pipetting 1-2 fA. 1 X hydrochloric acid onto the needle tip and
heating for 1 hr. at 100° in the hydrolysis chamber (page 123) . Allow
to dry at room temperature.

4. For the Deniges method, add 0.05 ml. of the stannous chloride
soln. to 10 ml. of the sodium molybdate soln. This mixture nmst be
used within 3 min. after its preparation. Pipette a 0.6 to 1.0 fxl. drop
of the reagent mixture on a prepared slide near one of the inked
hnes (page 123) as a blank. Then pipette a suitable vol. (0.6-1.0 ^,1.)
of the reagent mixture onto the needle tip bearing the ashed sample.
Use the end of the pipette to mix the drop on the needle tip for 10-15
sec. in order to dissolve the sample and obtain a homogeneous soln.
During the next 15 sec, transfer the drop from the needle to the
slide, placing it beside the other iliked line a few mm. from the blank
drop. Surround the drops with paraffin oil and place cover glass as
described on page 122. The entire process from the addition of the
reagent to placing the cover glass can be performed in 35-45 sec.
Standardize the manipulations to maintain a constant evaporation
effect. Carry out the photometry after 5 min. from the beginning of
the color development, and finish within 45 min.

5. For the Fiske and Subbarow method, mix 24 ml. of the molyb-
date soln. with 1 ml. of the aminonaphtholsulfonic acid soln. This
mixture may be used for at least 1 hr. after its preparation, and it
is applied in the same manner as the Deniges reagent. Let the drop
stand for 30 min. before starting the photometry, and finish the
measurement within 2 hr. from the beginning of the color develop-
ment.

3. Stowell Technique

The apparatus consists of a lamp, a microscope, a photocell, and
amplification and recording equipment. A 50 c.p. automobile head-
lamp operated by a .storage battery is used as the source of light.

126 MICROSCOPIC TECHNIQUES

The lamp is housed in a Spencer (No. 367) lamp case which has a
filter holder. A microscope fitted with a mechanical stage, a 44 X
achromatic objective, and a 15X compensating ocular are employed.
The field of observation is limited to an area 50 X 35 /a by inserting
a rectangular diaphragm in the ocular. The light from the microscope
is thrown on a vacuum photocell (RCA 929) enclosed in a light-tight
box having a side tube to extend over the microscope tube. This
side tube contains a movable mirror so mounted that it can be
interposed in the light path to enable inspection of the field, and
then tm'ned out of the path to permit photoelectric measurement.
The photocell current is amplified by a General Electric FP54 tube
in a Barth circuit with a 10^^ ohm grid resistor (Penick, 1935). The
amplified current is measured with a Leeds and Northrup student
type potentiometer and a Leeds and Northrup type R galvanometer.

Measurements are made of the light transmittance through both
stained and unstained sections in order to obtain the absorption
due to the stain itself. In both cases, measurements are also made
of the transmittance through blank portions of the glass slides
adjacent to the sections to correct' for variations in the intensity
of the light source, changes in amplification or potentiometer bat-
teries, and alterations in thickness of cover glasses, slides, or mount-
ing media. When it is possible, fifty adjacent areas on each section
are measured and the mean percent absorption calculated.

Stowell and Albers ( 1943) employed a Coleman Model 10 S double
monochromator spectrophotometer by means of which they meas-
ured absorptions of light bands, having a 5 rap. spectral width, by
stained sections of tissue. Since no microscope is employed in this
apparatus, the absorption of the section as a whole, rather than a
chosen cellular region, is measured.

The photometric procedure was employed by Stowell ( 1942) for
the estimation of desoxyribonucleic acid in tissue sections by means
of the Feulgen stain (page 65). Light from the source was passed
through a heat-absorbing filter (Corning Aklo No. 396) and a green
gelatin filter (Wratten No. 58) before entering the microscope. The
extension of the method to absorption studies with a variety of
stains commonly used in histological examination was included in
the reports of Stowell and Albers (1943) and Stowell (1945b).
Subsequently Stowell (1945a) subjected the Feulgen reaction to
detailed study and described each step in his method of using it.

ROENTGEN ABSORPTION HISTOSPECTROSCOPY 127

D. ROENTGEN ABSORPTION
HISTOSPECTROSCOPY*

One of the most significant advances in histo- and cytochemical
technique has come from the work of Engstrom ( 1946) at Karohnska
Institutet, Stockhohn, who, by employing the roentgen absorption
of tissue sections or of very small volumes of liquid, developed a
procedure whereby quantitative elementary analyses can be directly
]ierformed with an accuracy of about 5-10% on 1 X 10"^ to 1 X
10"^- gram of material, i.e., quantities of the order found in single
mammalian cells. Thus, in specific instances, phosphorus and calcium
can be determined in a 10 /x section of bony tissue within an area
around 10 X 10 ix, and nitrogen and oxygen in a section a couple
of microns thick within an area of 50-100 /x-. A particular advantage
of this technique is that the tissue is not used up, and hence may be
subsequently employed for histological study by the usual methods
so that direct correlation may be made between the chemical
composition and the microanatomical structure. Furthermore, the
analysis is independent of the chemical structure in which the ele-
ment may be bound, and the physical state of the specimen is
unimportant, e.g., fixed tissue, dry powder, and in certain cases,
paraffin-embedded tissue, or solutions, may be used. A number of
elements can be determined on the same sample.

The advantages of being able to determine the total quantity of
a tissue element in situ without regard to its chemical form and state
of valence, or affiliations with other elements, are hardly to be
minimized. The method is confined to the quantitative determination
of elements having an atomic number of 6 (carbon) or greater. This
would include all elements of biological importance with the
exception of hydrogen.

Engstrom (1946) has pointed out that roentgen spectroscopic
methods based on emission analysis are not well suited for elements
with atomic numbers less than 20, and the emission methods cannot
be applied very well to the small surfaces involved in histo- and
cytochemical studies. This applies to the earlier procedure of von
Hamos and Engstrom ( 1944) in which tissue is subjected to roentgen
radiation and the secondary radiation is measured for the quantita-
tion of constituent elements.

* See Bibliography Appendix, Refs. 23, 24, 25, 26, and 27.

128

MICROSCOPIC TECHNIQUES

The use of the roentgen absorption method requires an under-
standing of the theoretical basis of roentgen spectroscopy. All that
can be included here are a few of the more salient features of the
theoretical treatment which Engstrom ( 1946, pages 19-50) applied
to his method, a description of the apparatus, and consideration of
certain other practical aspects. This information can serve merely to
acquaint the reader with the method, the actual use of which will
depend on the understanding of roentgen spectroscopy mentioned
l)reviously and a detailed study of the presentation of Engstrom
(1946).

O

1

I

o
o

z
o

o

h

y

<

CO
CO

<

/

1

^,<

abs

WAVE LENGTH

Fig. 15. Schematic representation of an absorption jump.
From Engstrom (1946)

1. Some Theoretical Aspects

Quantitative analysis based on the roentgen absorption utilizes
the absorption discontinuities which appear at a characteristic
wavelength for every element. The absorption of an element near
the K-absorption edge is indicated in Figure 15. The determination
of the element depends on the measurement of the absorption of
monochromatic radiation wnth wavelengths on each side of, and
close to, the absorption edge of the element to be determined.

The mass (x), in g./cm.^, of the element to be analyzed is given by
the following equation:

In

X =

A

Ml
P

(S)'

In

M2
P

©'

ROENTGEN ABSORPTION HISTOSPECTROSCOPY

129

TABLE III

Wavelength of the Absorption Edges for Certain Elements and Suitable
Analysis Lines (Engstrom, 1946)

Atomic
num-
ber

K-

absorption

edge,

X.U.

Xi

Xj

• Ele-
ment

Line

Wave-
length,
X.U.

Line

Wave-
length,
X.U.

C

X

Xa

Mg

P

S

CI

K

Ca

Fe

Cu

Zn

As

Br

Ag

I

6
7
8
11
12
15
16
17
19
20
26
29
30
33
35
47
53

43700

31000

23300

~11500

9496

5775

5009

4384

3431

3063

1740

1377

1280

1043

918

484

373

20 Ca Ka
22 Ti La
24 Cr La
32 Ge Lai
34 Se Lai
41 Xb Lai
44 Ru Lai
46 Pd Lai

20 Ca Kai

21 Sc Kai
70 Yb Lai
77 Ir La,
79 Au La,
81 Tl La,

92 U La,
51 Sb Kai
57 La Kai

36270

27370

21530

10415

8972

5712

4836

4359

3352

3025

1668

1348

1274

1013

909

469

370

6C Ka

21 Sc La
23 V La
11 XaKai
33 As Lai
40 Zr Lai

16 S Kai
45 Rh La,
51 Sb La,
20 Ca Ka,
68 Er La,
76 Os La,

78 Pt La,

79 Au L/3,
37 Rb Ka,
50 Sn Ka,
56 Ba Ka,

44540

31370

24310

11885

9652

6057

5361

4588

3432

3352

1780

1389

1310

1081

924

490

384

Atomic
number

Liii

absorption

edge,

X.U.

X,

X2

Ele-
ment

Line

Wave-
length,
X.U.

Line

Wave-
length,
X.U.

Ca
Cu
Ag

I

Hg
Bi

20
29
47
53
80
83

35630

13150

3691

2714

1008

922

21 Sc La
30 Zn La
50 Sn Lai
57 La La,
34 Se K^i
92 U La,

31370

12250

3592

2660

990

909

20 Ca La
29 Cu La

19 K Ka,

22 Ti Ka,
35 Br Ka,
37 Rb Kai

36270

13600

3734

2743

1038

924

where (mi/p) and {m/p) represent the mass absorption coefficients
for the element at wavelengths Xi and X2, respectively, p, the specific
gravit}^ of the absorbing substance, and /x, the linear absorption coef-
ficient ; ii and h are the intensities of transmitted radiation of wave-
lengths Xi and X2, respectively, and /i and I2, the corresponding in-
tensities of the incident radiation, P is a function of atomic number
and wavelength. When the wavelengths are selected close to the
absorption edge, and the intensities of the incident radiation of both

130

MICROSCOPIC TECHNIQUES

wavelengths are equal, i.e., /i = h, the equation may be simplified
as follows :

ii = 42e-(Mi/p-M2/p)a: = ^26-**

Values for (i) and (/) are determined experimentally, and (m/p) can
be obtained from tables of mass absorption coefficients for different
elements and wavelengths, or it may be calculated.

10-

10'

10-

10'

10'

1
z

K Edge 31
L,,, Edge
Mv Edge

\

1

\

\

*

L,„

My

N

^

X

^

\

\

\

V-

■s.

^

10 ' 20 30 40 50 60 70 80 90 ^

]3,06 1.28 0.69 0.42 0.29 0.20 0.15 0.11 A

9M 5.56 3.15 1.90 1.39 1.01 0.76 A

5.33 3.72 A

10-'

10"

E
^10"'

o 10

<

10"
10"

6C 7N 80 llNa 12 Mg 15P 16 S 17 CI 19 K 20Ca 26re

Fig. 16. Value of k, that is, (mi/p)
— (m2/p), for different absorption edges
and different elements. For 15 P the
wavelength for the L edge is Ljj

1

,

^

>

X

/

/

K i I

^

/

•^i

/

1

\ \

i 1

1

10 20 30 40 50 60 70

ATOMIC NUMBER, Z

80 90

Fig. 17. The smallest determinable
amount of an element.
From Engsirom (1946)

-L

III-

Fro7n Engstrom (1946)

From the preceding equation it follows that the smallest deter-
minable quantity of the element per unit of surface is a function of k
or (mi/p) — (m2/p) . Accordingly, an absorption edge should be chosen
at which the difference between the mass absorption coefficients is
as large as possible. The variation of k with elements of different
atomic numbers (Z) is shown in Figure 16, which illustrates that the
K-absorption edge must be used for elements of low atomic number,
the Lni edge for elements in the middle of the periodic table (the
Lni is of greater magnitude than Lj or Ln), and the My edge for the

ROENTGEN ABSORPTION HISTOSPECTROSCOPY

131

TABLE IV
Elementary Composition of Muscle (Engstrom, 1946)

Atomic
number

m. X 10»

20 Ca, Ka

3353 X.U.

51 Sb. L<

« 3432 X.U.

Element

g. X cm. -2 X

p

m. X - X 10'
p

p

m. X ^ X 10'

c

6

11500

50

5750

55

6330

N

7

3600

80

2880

85

3060

8

5000

115

5750

120

6000

Na

11

70

265

190

285

200

Mg

12

20

350

70

370

70

P

15

170

635

1080

680

1160

S

16

200

790

1580

840

1680

CI

17

60

885

530

945

570

K

19

370

1215

4500

125

460

Ca

20

2

145

3

150

3

Fe

26

20

310

60

330

70

S2.23910-3

21.960-10-3

heaviest elements (the My is the greatest of Mj-y). The smallest
quantity of an element, in g./mm.^, which can be determined is given
in Figure 17. This is based on the fact that the smallest intensity
difference between ii and i2 which can be measured with certainty is
about 5%.

1.0

0.8

?ti:o.6

0.4

02

:\^^=t:S=^

::^

-^

-i

..

V

V 1
\ 1

N

^ '

u

\

A

^^

^-

N

\

\

\ ^

\\'

I

1 \

s^

\

\

\

\'"

rim.

\

\

i\

100f\

\

k

\

\

\

\

V ^^--I^^

WAVE LENGTH, A

Fig. 18. Absorption of roent-
gen radiation of different wave-
lengths in paraffin of varying
thickness. From Engstrom
(1946)

a. 10'

8 10

6C 7N 80 llNal2Mg 15P 16S 17CI 19K20Ca 26Fe

Fig. 19. Appropriate layer thicknesses
for determinations of different elements
in muscular tissue. The solid columns
indicate the layer thickness when the
K-absorptibn edge is used; the hatched,
the Ljjj edge. From Engstrom (1946)

132

MICROSCOPIC TECH N IQUKS

2. Thickness of Sections

The tissue to be subjected to elementary analysis must be frozen -
dried and infiltrated with paraffin by a procedure such as that of
Packer and Scott (page 5). Removal of the paraffin from sections
of this tissue would involve extractions and displacements of ele-
ments by the action of the solutions reciuired. Therefore, it is pref-
erable to carry out the absori)tion analysis without removing the
paraffin. The absorption of the paraffin itself is demonstrated in
Figure 18. The curves were plotted for C2.iH52 (sp. gr. 0.90) ; they
show that, up to 3 A, the paraffin may be used in a layer up to
50-100 fx, while for 5-6 A the layer should not exceed 10 ju,, etc.

A summary of Engstrom's calculations of appropriate section
thickness for analyses of muscle tissue is given in Figure 19. It is
considered prerequisite that the quantity of an element to be
analyzed be about twice the smallest quantity which can be deter-
mined. Figure 19 was derived from data such as that in Table IV,

2

o

I—
<

<

1
0,8

0,6
0.4

S02

z
<
a:

0.1

~ —

1 . 1

^*'*^*s-^

^^^^^

4

^^

::;-.^

^<::^

>^

-^■^^2

1^

0.1

0.2 0.3

THICKNESS, mm.

0.4

0.5

Fig. 20. Analysis of
19 K in muscular tis-
sue: (1) Uh; (2)
i^/h.; (3) ii/U for
/, = /.; (4) ii/io for
wavelengths v e r y
close to the absorp-
tion edge and /, =1-2.
See Table IV.
From Engalrdi)).
(1946)

which gives the elementary composition of muscle and the mass
absorption coefficients and absorption capacity of the elements with
the wavelengths used. The wavelength of the K-absorption edge
for potassium (atomic number 19) is 3427 X.U. (1 X.U. = 0.001 A),
and hence, for the potassium determination analysis lines on each
side, 3353 and 3432 X.U. are used in the table. A specific gravity of
1.0 for the tissue has been used in the calculation, and the layer

ROENTGEN ABSORPTION HISTOSPECTROSCOPY

133

thickness is 1 />.. The data in the table have been used to obtain the
curves in Figure 20, and from these it appears that sections 0.2-0.3
mm. thick are appropriate for potassium analysis. Thus, in a volume
of 0.001 fx\. of muscle tissue having a surface of 0.01 mm.- and a
potassium content of 0.3% the quantity analyzed will be 3 X 10"^ g.

Fig. 21. Schematic picture showing the arrangements for analysi.s in a
microcuvette : S, slit in roentgen tube; C, the crj^stal; K, the microcuvette,
F, the photographic fihn. In place of K a microscopic section can be used.
From Engstrom (1946)

Fig. 22. Roentgen tube for primary excitation.
From Engstrom (1946)

3. Apparatus

The arrangement shown in Figure 21 enables the simultaneous
determination of the incident and transmitted radiation intensity.
The former is proportional to the blackening on the upper part of
line L, and the latter to the lower part. The roentgen tubes employed
by Engstrom to produce the radiation were operated by a direct-
current unit manufactured by G. Schonander Co., Stockholm. This

134

MICROSCOPIC TECHNIQUES

unit was designed to develop 1.5 kilowatts at 50 kilovolts or 4.5
kilowatts at 15 kilovolts. The evacuation of the roentgen tubes was
accomplished by a three-stage mercury diffusion pump connected to
a two-stage mechanical forepump. A cooled trap was placed between
the tubes and the pumps.

Roentgen Tube for Primary Excitation. In order to obtain a
line spectrum of great intensity it is best to solder the element whose
line spectrum is desired to the anode. This cannot always be done,
but for elements which lend themselves to this procedure, the anode
is made with six surfaces having a different element or suitable alloy
of it on each surface. A diagram of the tube is shown in Figure 22.
The forged brass body (A) has four openings and bored channels
for cooling. The water-cooled cathode (F) with filament E is sepa-

Fig. 23. Roentgen tube for secondary excitation.
From Engslrom (1946)

rated from A by rubber packing. The anode (B) is insulated from
A, which is grounded, by the porcelain tube C. A rubber packing
separates the cone (D) from C. The anode may be turned to present
its different surfaces to the cathode without breaking the vacuum
by virtue of an Apiezon grease packing between it and D. The
anode is water-cooled. The roentgen beam is passed through slit G
which is covered with an aluminum foil (9 /a thick) fastened on
with Apiezon grease. The flexible tube H connects to the vacuum
apparatus. The filament (E) consists of a 0.25 mm. platinum wire
winding which is coated with an oxide layer (made by burning off
sealing wax) .

ROENTGEN ABSORPTION HISTOSPECTROSCOPY

135

Roentgen Tube for Secondary Excitation. When the element
whose line radiation is desired cannot be soldered to the anode, the
oxide or the powdered metal must be used. However, the difficulties
of suitably incorporating these powders in the surface of the anode
led Engstrom to the use of secondary excitation outside of the high
vacuum. The resulting reduction in intensity was compensated in
some measure by the use of greater wattage and by obviating the
passage of the radiation through a window or membrane.

The roentgen tube is illustrated diagrammatically in Figure 23.
The brass body (A) is 90 X 90 X 100 mm. The cathode filament
consists of a flat platinum spiral with an oxide coating. An iron
cylinder surrounds the filament to direct the electron stream. The
tube is evacuated through H, which is 42 mm. in diameter. The
surface of the anode {B) forms a 12° angle, and the porcelain tube

Fig. 24. Spectrograph for microanalysis.
From Engstrom (1946)

C insulates the anode. Adjustment of the anode position under
vacuum is obtained by a bellows arrangement (D). Water-cooling
is employed for the body of the tube, the cathode, anode, and slit G.
An aluminum foil window ( 9 fi thick.) is placed over the slit, and a
plug (P) is used to hold the element whose line radiation is desired.
This plug and all of the vacuum connections are sealed with rubber.
The composition of the anode is chosen to give the greatest yield of
secondary radiation from P. Hevesy has shown that the greatest
yield of characteristic radiation results when the incident linp radi-

l.'iC)

MICROSCOPIC techniquj:.s

ation has a wavelength 200-600 X.U. shorter than that for the
absorption edge of the element whose secondary radiation is desired.
Accordingly, it is preferable to use K-radiation from copper ( atomic
number, 29) to excite K-radiation from iron (26). If there are no
suitable lines for the secondary radiation the continuous radiation
from an element of high atomic number such as tungsten (74» or
platinum (78) is used.

Spectrograph for Microanalysis. For wavelengths shorter
than 2.5 A it is not necessary to enclose the spectrograph in a vaccum,
since the absorption due to air becomes appreciable only for wave-
lengths greater than 2.5 A. The spectrograph for microanalysis is

^

■» 1-^ niiTi

+

i

1 ^ \

E

1

1

E

E

CM

F

5 mrr

E

}

d

1 1 1

t

',■,■';';

\ — / \ ^ \

1

Fig. 25. Microcuvette. The hatched area is the volume employed.

From Engstrom (1946)

used with these shorter wavelengths; the diagram of the instrument
is shown in Figure 24. Either photographic or ionization measure-
ments may be made of the radiation intensity with this apparatus.
The roentgen tube (.4) delivers a stream of radiation through the
spectrograph slit (B) to the crystal (C) which is mounted in the
holder (D). The monochromatic beam produced passes both above
and through the microcuvette (E) and onto the photographic film
in the holder (F) . The cuvette can be moved relative to the film by
the micrometer screw (G). An ionization chamber (H) may be
substituted for the photographic film; the chamber is directly con-
nected to an electrometer by its central electrode. (An Edelmann
string electrometer was employed.) A Bakelite bar (7) operates a
cog by which the cuvette may be moved into or out of the path of

ROENTGEN ABSORPTION HISTOSPECTROSCOPY

137

the beam when the ionization chamber is used. Adjustments on the
circular scale can be made to some hundredths of a degree, and stops
on the scale can be set to enable rapid and accurate changes for
different wavelengths.

The construction of the metal microcuvette is shown in Figure 25.
The walls in the direction of the radiation are made of thin aluminum
foil, glass, or cellophane sheets. The cuvette capacity is 0.2 fx\. or
greater. Crystals of calcite (d = 3029 X.U.) and rock salt
(d = 2814 X.U.) were used and the ionization chamber was
charged to 150 A'olts.

10 cm

Fig. 26. Vacuum spectrograph for histochemical analysis.
From Engstrdm (1946)

Vacuum Spectrograph for Tissue Analysis. The construction
of the vacuum spectrograph is shown in Figure 26. The roentgen
tube (yi) is joined to the spectrograph by an air-tight rubber gasket.

138

MICROSCOPIC TECHNIQUES

The radiation enters through the adjustable slit (B) and falls on
the crystal (C), whose angle may be adjusted by the micrometer
screw (D). The monochromatized beam from the crystal passes
through the tissue section in holder E. The holder can be moved in
relation to the photographic film placed behind the tissue by means
of the micrometer screw (/) . The film carriage can be adjusted along
scale F at a chosen distance from C. The angle of the tissue holder
can be set on scale G. The spectrograph whose dimensions are
10 X 20 X 6 cm. is evacuated through H. The lid is sealed to the
chamber with rubber and clamped tight. The films used were
12 X 12 mm. and the crystals employed were gypsum {d = 7578

Fig. 27. Mounting of a preparation.
.A is a part of the preparation holder.
B is the preparation itself. C is a cross
of Wollaston wires (platinum) used
to obtain points of reference.

From Engstrom (1946)

X.U.) and mica (d = 9930 X.U.). The tissue section is mounted
over a hole (0.2-2 mm.) in a sheet of metal, Figure 27. The distance
between tissue and film is 1.5-2 mm.

4. Measurement of Density of Photographic Film

The intensity of the radiation is measured by the degree of the
blackening of the photographic film. The blackening may be deter-
mined by photometric estimation of the proportion of visible light
absorbed by the film or by measurement of the proportion of the
silver that has been reduced.

The photometric estimation has been effected by two methods in
Engstrom's (1946, pages 60-64) work. One procedure utilized a
self-recording microphotometer (Siegbahn type) with a thermo-
element and Moll's microgalvanometer, and the other employed the
Caspersson photoelectric apparatus for the determination of light

ROENTGEN ABSORPTION HISTOSPECTROSCOPY

139

absorption in very small areas. The latter method was previously
described in connection with Caspersson's absorption technique for
cells (page 114) and Norberg's technique for fluids (page 120).

1.00
0.90
0.80
0.70
0.60

0.50
0.40

030

0.20

—I 0.01

0.10
0.09
0.08
0.07
0.06

0.05
0.04

003 -

X =

H

Mi _ t^(h\'

p P \X2/

Ml
P

^lg/mm^

0.02

0.01

//g/mm.^

26 Fe 100 ^g/mm."

20 Ca 40

0.02

0.03

0.04

0.05

0.06

0.07

0.08
0.09 ^

0.10 ■^--

Fig. 28. Nomogram for calculating analytical results by equation

From Engstrom (1946)

0.20

030

040

0.50

0.60
0.70
080
0.90
— ' 1.00

shown.

The light absorption in areas less than 1 fi^ can be accurately
measured with this apparatus; however, the size of the grains,
even in the finest films, makes it necessary to measure the blacken-
ing in areas 10 X 10 fi.

140 MICROSCOPIC TECHNIQUES

The raetluxl for tlie estimation of the j)roportion of reduced silver
is based on counting the silver grains in the photographic emulsion
on the film, following the procedure of Glinther and Wilcke (1926i.
The method is only adapted to low densities (upper limit d = 0.27) .
A microscopic enlargement of 600 X is used for coarse-grained film
and 1350X (immersion objective) for fine-grained film. The counts
are facilitated by the use of a netted ocular having 100 squares.
The number of grains in an unexposed film area is determined in a
region immediately adjacent to that exposed.

In a study of the properties of photographic emulsions, Engstrom
(1946, pages 65-72) pointed out that it is necessary to obtain a
curve of the relationship of photographic density to radiation
intensity for every wavelength, emulsion, and set of development
conditions in order to arrive at a suitable working arrangement.
Engstrom investigated the properties of Agfa, Laue, and Printon
films and Ilford High Resolution plates and presented curves of both
density and number of grains as functions of intensity.

Nomogram

Engstrom ( 1946) has published a nomogram for calculating the
analytical results according to the equation shown in Figure 28:

"In this nomogram are included the most important elements, and the analj^-
sis assumes the employment of the K-absorption edge. The wavelengths for the
analysis lines used are seen [Table III]. The two outer pillars in the nomo-
gram indicate the quotient between the intensity of the transmitted and incident
roentgen radiation in the two wavelengths Xi and X2. The amount sought for
(A') of the respective elements is marked out on the vertical lines in the
centre. The figures after the respective elements indicate the amount of the
element in ciuestion at the end of the scale, e.g., 30 Zn 100 /ug./mm.- The
following example shows how it is employed: In determinations of nitrogen,
it is, e.g., found that ii//i is 0.06 and that i^Hi is also 0.06. The straight line
which joins these two points on the outer pillars cuts the ciu-\e for nitrogen
at the point A'. The end point on the nitrogen scale is 1.5 /xg.N/mm.^, and the
scale is divided up into 15 parts, from which it appears that the amoimt of
nitrogen sought for is 0.49 /xg./mm.^"

E. MICROINCINERATION

Microincineration is a valuable technique for the faithful repre-
sentation of the total mineral distribution in tissue sections. In its
present state of development, its reliability is evidenced by the fact

MICROINCINERATION 141

that motion pictures of incinerating sections of skeletal muscle and
of anterior horn cells at 700-800 X magnification reveal no distor-
tion, Scott (1943). The advantage of microincineration over chemi-
cal tests for the determination of the anatomical disposition of
mineral constituents lies in avoiding inevitable displacements and
losses resulting from the use of solutions. In addition, the danger
of fortuitous adsorption of reagents on colloidal protoplasmic sur-
faces is circumvented. The present limitations of the technique lie
in its essential morphological character, which leaves not only
quantitative but qualitative chemical considerations very largely
in the dark. Only a few elements can be detected in incinerated
preparations, and only a rough estimate of the quantity of ash in a
given location can be made.

The microincineration technique was originally developed by
Policard and co-workers in France and was introduced in America
by Scott. The more recent refinements have resulted chiefly from
the careful and extensive researches of Scott and collaborators.
The earlier reviews by Policard (1931-1932) and Policard and
Okkels (1931) and the later ones by Scott ( 1933a,1937,1943) and
Gage (1938) thoroughly cover the development and applications
of this technique.

Engstrom ( 1944) carried out a very nice study on the localization
of mineral salts in striated muscle fibers by employing ultraviolet
absorption followed by microincineration of the same section. By
correlation of both techniques he was able to conclude that the
intensely absorbing isotropic segments which contain the adenylic
acids yielded the ash, w^hereas practically no ash was derived from
the weakly absorbing anisotropic segments.

1. Preparation for Incineration

Some of the earlier work dealt with the use of various solutions
for the fixation of tissue in preparation for microincineration, and
it was found that absolute alcohol or an alcohol-formalin mixture
w^as best since their use resulted in a smaller loss of mineral matter
than was observed with other fixatives. In Scott's (1937) hands the
intracellular distribution of minerals was preserved remarkably
well when the tissue was fixed for 24 hr. in a solution of 9 vol.
absolute alcohol and 1 vol. neutral formalin followed by treatment

./

142 MICROSCOPIC TECHNIQUES

with absolute alcohol, clearing in xylol, and embedding in paraffin
in the usual fashion. But Scott was quick to point out that the
advantages of freezing-drying the tissue are particularly important
in studies of this nature and hence freezing-drying is the procedure
of choice.

Paraffin sections, 3-5 ^ thick, yield the most satisfactory cyto-
logical details. The sections are placed directly on ordinary glass
slides of good quality and no adhesive is required to make them
adhere to the glass. While the presence of water is scrupulously
avoided, a drop of absolute alcohol or liquid petrolatum may be
employed to flatten the sections if necessary (Policard and Okkels,
1930). If alcohol is used, it is allowed to dry; if petrolatum, it is
drained from the slide before the sections are incinerated. The
greatest care must be exercised at all times to avoid contamination
with dust. Absolutely clean paraffin must be used; and the slides
should be washed in distilled water repeatedly, rinsed with filtered
alcohol, dried with a clean lint-free cloth, and stored in a dustproof
container.

It is good practice to cut serial sections using alternate ones for
incineration and the others for controls to be stained and mounted
in the usual manner. Scott ( 1937) pointed out that the use of the
cold knife for sectioning, as recommended by Schultz-Brauns ( 1931) ,
is undesirable since condensation of moisture on its surface may
result in some wetting of the tissue.

2. The Incineration Furnace

The furnace used for the incineration of tissue sections is simply
a quartz tube electrically heated by windings of resistance wire.
Ordinary laboratory muffie furnaces can be used if their tempera-
tures can be properly regulated and if sufficient care is taken to
protect the slides from possible contamination inside the furnace.

Scott (1937) constructed a very convenient furnace capable of uni-
form and reproducible performance. It consists of a quartz tube 24
in. long that is wound with three separate 600 watt heating units
and the whole covered with asbestos insulation. Each heating unit
is controlled by a 44 ohm, 3.2 amp. rheostat. The slides are slowly
moved through the furnace tube on quartz slabs by means of an
electric motor operating through a speed-reducing worm gear

MICROINCINERATION 143

system. A rack running through the tube and extending from both
ends serves to support the quartz slabs as they are moved along.
Further details of the apparatus have not appeared.

3. Scott Incineration Procedure

Scott (1937) gave the following directions for incineration:

1. Gradually bring to 200° over 10 min.

2. Gradually elevate to 280° over the next 5 min.

3. Gradually elevate to 385° over the next 5 min.

4. Gradually elevate to 480° over the next 5 min.

5. Gradually elevate to 580° over the next 5 min.

6. Gradually elevate to 650° over the next 3-5 min.

7. Shut off furnace and let cool for 5-10 min.

8. Remove slides from furnace and place cover slip over section
as soon as cool enough to handle. Seal edges with a mixture of 1 part
paraffin, 1 part beeswax, and 1 part resin (by weight). The use of a
cover slip permits observation with an oil immersion objective, and
it prevents absorption of moisture and efflorescence of the ash.
Greatest care is advised to avoid any air current between the time
of removal from the furnace and sealing the cover slip, since the
ash is easily disarranged.

The practice of covering the ash with collodion or Canada balsam
is undesirable, since it involves the danger of disarrangement and
disturbance of optical properties.

Variation in the above procedure may be necessary for particular
tissues. The greatest shrinkage occurs between 60-70°, and especially
in tissues rich in elastic and fibrous material such as blood vessels.
In order to produce practically all of the shrinkage in advance,
Policard and Ravaut ( 1927) place fixed tissues in absolute alcohol
and bring slowly to the boiling point. However, this procedure is
not advised by Scott for cytological studies because there is the
possibility of dissolving salts.

The passage of a stream of nitrogen through the tube during
incineration was recommended by Schultz-Brauns (1929), and
Tschopp (1929) suggested a similar use of oxygen. Policard (1933b)
employed nitrogen containing a small concentration of oxygen, which
he claimed effects more rapid oxidation. Although satisfactory
results have been obtained with these methods, and they seem to

144 MICROSCOPIC TECHNIQUES

permit lower incineration temperatures which result in less volatili-
zation of chlorides, nevertheless it is sufficient as a rule to carry out
the treatment in air.

4. Microscopic Exainiiiation and Interpretation

Observations of incinerated sections should be made under the
microscope with dark-field illumination provided by a cardioid
condenser. A proper light source is an important factor and Scott
(1937) pointed out that a carbon arc seems to produce excessive
longitudinal aberration, while a Spencer illuminator fitted with a
500 watt projection lamp or a Zeiss Point-0-Light lamp, with proper
centering of the condenser and adjustment of the mirror, are suitable.
Unscreened light is best for observation of minute particles, but
the use of a daylight filter or ground glass is more restful to the eye.
Mercury vapor lamps enable high resolution of small particles but
make recognition of colors difficult.

The use of a comparing ocular with two microscopes is recom-
mended for simultaneous observation of both an incinerated section
and its stained control. Of course, the stained section is illuminated
in the ordinary bright field.

If the incineration has been carried out properly, there will be no
black or brownish carbon deposits. The topographic disposition of
the ash no doubt fairly represents the distribution of mineral con-
stituents in the fixed tissue. That this distribution is exactly main-
tained in the living tissue cannot be said with certainty; however,
Scott ( 1932 ) has found the parallelism that histological sites in
living tissues that absorb ultraviolet radiation (275 m/x) are those
which yield large amounts of ash on incineration.

The only elements that can be identified in the ashed sections
with any measure of certainty are considered in the following
sections.

Sodium and Potassium. It has been assumed that sodium and
potassium yield a fine-grained, faintly bluish-white ash. Policard
and Pillet (1926) attempted the identification of sodium and
potassium as the sulfates by exposing sections, before incinerati(^n,
to the fumes of sulfuric anhydride in order to convert the chlorides
to sulfates. The sulfates are resistant to volatilization during the
ashing, while the chlorides are apt to be lost.

MICROINCINERATION 145

Calcium and Magnesium. The dense white ash seen in the
dark field is due chiefly to calcium with a smaller amount of mag-
nesium. Unless spectrographic means are employed, magnesium
cannot be identified in the presence of calcium in incinerated
sections. As a test for calcium, Moreau (1931) suggested dissolving
ash in a ''microdrop" of 0.1 N hydrochloric acid followed by tlie
addition of a "microdrop" of 0.1 A^ sulfuric acid in order to form
the needle-shaped crystals of calcium sulfate.

Silicon. The identification of silicon can be made with assur-
ance since silica retains its typical crystalline structure during
incineration, and its double refraction when examined with polarized
light serves as an additional means of characterization (Policard
and Alartin, 1933) . The tendency of certain constituents in the ash
to combine with the silica in the glass slide during the heating may
give rise to a misleading appearance. The probability that silica and
calcium salts combine in the incineration process must also be
kept in mind.

Iron. The oxidation of iron that occurs produces a color in the
ash which may vary from yellow to deep red making the identifica-
tion of this element relatively simple. Scott ( 1937) cautions that
care must be taken to avoid contamination with iron from the
microtome knife. A newly sharpened knife will be apt to cause the
most trouble. After 40-50 sections have been cut the number of
particles of iron left in the subsequent sections is practically negli-
gible.

Lead. Exposing incinerated sections to hydrogen sulfide gas
has been employed by Tada (1926) and Okkels (1927) for the
identification of lead as its black sulfide. However, sulfides of other
metals are also black and the possibility of an interference of this
nature should be kept in mind. It is necessary to make sure that
carbon particles due to faulty incineration are not present before
the ash is subjected to the gas, since these particles and the black
sulfide can be easily confused.

Uranium. Policard and Okkels ( 1930) claimed to have detected
uranium, by its fluorescence under ultraviolet radiation, in ashed
sections from animals poisoned by this element. Fluorescences may
be produced by impurities too, and hence this criterion is not a very
rigorous one.

146 MICROSCOPIC TECHNIQUES

5. Quantitative Estimation of Ash

Attempts at the quantitation of the relative amounts of ash left
by various structures were made by Schultz-Brauns (1931) based
on a standardized development of photomicrographs. However, this
method has many inherent difficulties and can yield little.

Scott (1933b), and Williams and Scott (1935) developed a photo-
electric apparatus to measure the intensity of light reflected from
the ash. This light intensity is roughly proportional to, and serves
as an approximation of, the quantity of the mineral residue. While
the method obviously leaves much to be desired, as Scott would
no doubt be the first to admit, it is capable of furnishing some
information that at present can be obtained in no other way.

An idea of the sensitivity of the apparatus, as assembled and used
by AVilliams and Scott, may be gained from the fact that, with a
magnification of 700X, the ash produced in a 5 ju, section by a single
hepatic cell nucleus results in a galvanometer deflection of about
half-scale (25 cm.).

Williams and Scott Photoelectric Apparatus. The microscope
illumination is furnished by a 6 volt, 108 watt ribbon filament
projection lamp enclosed in a ventilated housing. This lamp is
supplied through a constant-voltage regulator and a step-down trans-
former. The slide is held by a special mechanical stage which has
rack and pinion adjustment laterally and vertically and a fine screw
adjustment axially; and the microscope, mounted horizontally, is
fitted with a Zeiss aplanatic 1.2 condenser, a Leitz No. 3 objective,
various oculars, and a clamped-on 90° reflecting prism. A fixed
diaphragm made of a disc slightly larger than the aperture of the
objective is fitted into the ring on the condenser lens. The light
emerging from the ocular is reflected by the prism to a mirror which
in turn reflects it to the photocell.

The gas-filled photocell is mounted in a light-tight copper box
which serves as an electrostatic shield as well. The cover of this
box carries rotatable and interchangeable 3 in. white cardboard discs
with various size openings to determine the illuminated area on the
photocell. A shutter under the disc permits exposure of the photocell
when desired. Within the copper box containing the photocell, a
rD54 Pliotron tube is mounted with its 10^ ohm high-resistance
shunt. The photocell is connected directly to this tube, and the

ANALYTICAL ELECTRON MICROSCOPY 147

connections to the DuBridge and Brown (1933) amplifier and the
photocell B battery are made thi'ough a shielded flexible cable, which
also carries the galvanometer leads. The amplifier is powered by-
storage cells ( 12 volts) . A Leeds and Northrup type R galvanometer
is used. The entire apparatus is mounted in a dark room, and black
felt is used to prevent stray light of the apparatus from reaching
the photocell.

Procedure. The image of the lamp filament is sharply focused
on the center of the cardboard screen with the dark-field diaphragm
removed and the slide in place. It is convenient to place a 12 power
convex lens in front of the lamp housing to enlarge the image to
about 4 in. The dark-field diaphragm is inserted, and, with the room
almost completely dark and the felt curtains in place to eliminate
stray light, the image of the ash is focused on the screen with the
axial adjustment of the mechanical stage. With the other adjust-
ments of the stage, the desired area is brought to the center of the
screen and the photocell box is moved so that the aperture is
properly set. Then the shutter is pulled and the galvanometer deflec-
tion is observed. The reading obtained when the clear glass of the
slide is placed in the optical field is substracted from the first
deflection, and the resulting figure is proportional to the quantity of
ash on the area taken when this quantity is small. Similar data
obtained for other areas enable a comparison of relative amounts
of ash without regard to the absolute quantities involved.

F. ANALYTICAL ELECTRON MICROSCOPY

The higher resolving power of the electron microscope has
enabled finer morphological differentiations in biological material
than were hitherto possible with the best optical microscopes. How-
ever important this advantage in other fields, as used today its
purely morphological value limits its contributions to cyto- and
histochemistry. However, a beginning has been made toward the
use of an analytical electron microscope, not merely as a means of
obtaining high magnifications, but as a tool for the identification
and localization of certain metallic elements in biological prepara-
tions. To date only calcium and/or magnesium can be identified and
localized. These elements can be detected with a sensitivity of about
1 X 10"^^ g- per kilogram wet weight of tissue (muscle). For the

I*.

148 MICROSCOPIC TECHNIQUES

conception of this possibility and the ingenuity to carry through to
fruition, credit is due to Gordon H. Scott and his associates, J. H.
McMillen and D. JVI. Parker, wlio carried out their investigations
at Washington University.

Scott and co-workers utiHzed the well-known fact that when
metals or certain of their compounds are heated in vacuo they emit
a number of electrons depending in part upon both the nature of
the metal and the temperature. Hence identification of the metals
might be possible on the basis of their differential emission of
thermally excited electrons. The localization of these metals in tissue
sections would be possible since the electrons emitted could be
focused by the magnetic lenses of the electron microscope to yield an
image, on a fluorescent screen, of the topographical disposition of
the emitting substances.

The first apparatus developed for this purpose was that of
McMillen and Scott ( 1937) but, since a number of changes have
been made, only the later apparatus of Scott and Packer ( 1939a)
will be described.

The Scotl-Packer Analytical Electron Microscope

A diagram of the instrument is given in Figure 29. Basically it is
an electron microscope of relatively low magnifying power fitted
at one end with a chamber (C), lined with a material capable of
fluorescence by electron streams for the visualization of the electron
image, and at the other end with a special cathode (B), carrying
a tissue support (.4) to hold the paraffin sections that are employed.
The microscope tube is made of brass, and is 1 m. long and 63 mm.
in diameter. Two magnetic lenses (Lj and Lo), swung on gimbals,
surround the tube and are free to move along it as well as rotate
around it to some degree. Each lens is composed of 1550 turns of
No. 22 enameled, single-cotton-covered copper wire wound on a
copper inner ring 75 mm. in diameter. The coils are enclosed in
sheaths of soft iron having a wall thickness of 3 mm. The lenses have
an axial width of about 43 mm. It is apparent that the object-image
distance is fixed; therefore focusing for any magnification must be
accomplished by altering the power of the lenses. This is done by
varying the current passing through the lens coils. The power for
these coils is supplied from two 30 volt banks of storage cells in
parallel.

ANALYTICAL ELECTRON MICROSCOPY

149

The tissue holder assembly is made of a 25 mm. glass tube fitted
snugly with a brass sleeve soldered to a heavy brass plate. A
shoulder on the plate fits into the microscope tube and the joint is
rendered vacuum-tight with Apiezon Q sealing compound. Proper
centering alignment is maintained by brass cylinders {A and ^i)
which, fit over the glass tube. The flat top of the nickel cathode
cylinder (B) serves as support for the tissue sections and the
cylinder itself is held by a wire fixed to its wall. The cathode is
heated by a coiled filament (H), which consists of 10 mil tungsten
wire supported by a ceramic core not indicated in the diagram. The

Fig. 29. Diagrammatic sketch of the essential features of the electron
microscope. A, tissue holder (cathode) support; Ai, inner shell of same; B,
nickel cathode; C, fluorescent screen; D, diaphragm; H, heating filament for
cathode; Li, objective magnetic lens; L2, ocular magnetic lens; P, pumping
port; T, transformer. Other letters and symbols are standard usage in vacuum
tube technique. From Scott and Packer (1939a)

cathode filament is supplied with current from two 6 volt storage
batteries in series which are insulated from ground for 15.000 volts.
The insulation is required since the tissue holder assembly and
batteries are at a potential of 6000 volts with respect to the grounded
microscope tube.

The fluorescent screen (C), which is sealed to the microscope
tube with wax, is a commercial oscillograph type. The bare portions
of the inner walls are coated with colloidal graphite (Aquadag) to
furnish electrical conductivity from the screen to the ground. This
prevents the accumulation of charge, which would distort the image
on the screen.

150

MICROSCOPIC TECHNIQUES

The diaphragm (D) has an aperture of 25 mm. over which are
placed a few fine wires to serve as fiduciary marks in focusing the
lens Lo. The image is formed with lens Li (Fig. 30) on the aperture
of the diaphragm in order to reduce spherical aberration. The half-
wave rectifier and filter system (Fig. 29) supplies the potential
difference required for accelerating the electrons. The current pass-
ing through the 10^ ohms resistance placed across the high-voltage
leads is measured by the microammeter from which the magnitude
of the accelerating voltage can be obtained. The rectifier is supplied
by the secondary of a transformer that is insulated from the primary
for 15,000 volts. A variable resistance in the primary circuit of this
transformer permits the choice of the high D.C. voltage employed.

Fig. 30. Diagrammatic representation of the electron path and consequent
image formation. A, cathode and support; C, fluorescent screen; D, diaphragm;
E, batteries for heating filament of cathode; F, schema of magnetic field of
the lenses Li and La; V, high-voltage source. From Scott and Packer (1939a)

Constancy is maintained in the accelerating voltage in order to
prevent a distortion, similar to chromatic aberration in optical
systems, by the use of a voltage regulator, of the saturated-core
transformer type, in the primary circuit of the transformer. Another
precaution taken to avoid distortion of the image is the placing of
the batteries, high-voltage supply, and all iron objects at quite a
distance from the microscope.

In order to compensate, at the magnification used (< 150x),
for the deflection of the electron stream by the earth's magnetic
field, the first lens (Li) is tilted.

Water-cooling coils of copper tubing are employed to remove heat
from the microscope tube. One coil is wound around the tube on the
image side of lens Li and another at the junction of the tube and the

ANALYTICAL ELECTRON MICROSCOPY 151

object holder. The latter is essential to absorb the 50-60 watts of
power given off by the heated filament, which would tend to soften
the sealing compound and weaken the vacuum.

The microscope is evacuated through a port (P, Fig. 29) by
means of two double-stage mercury vapor pumps in parallel employ-
ing a Cenco Hyvac oil forepump. A vapor trap ccioled with carbon
dioxide in butyl alcohol is placed between the microscope and the
pumps, and the glass-to-metal connection is sealed with black
vacuum wax. Pressures are measured with the ionization gauge of
Montgomery and Montgomery ( 1938) employing a No. 47 radio
tube of Radio Corporation of America. A portable power pack is
used so that it can be employed with several vacuum systems to
obviate duplication of expensive meters. A pressure of 10"^ mm.
mercury or less is sufficient for the electron microscope, and the
sensitivity of the pressure gauge is about 7 X 10"^ mm. mercury per
microampere ion current.

Manipulation

Preliminary experiments with salts of sodium, potassium, calcium,
magnesium, and iron demonstrated that very bright images were
formed on the screen by calcium and magnesium, weak ones by
sodium and potassium, and none by iron. By maintaining a cathode
temperature of 700-800° for an hour the sodium and potassium
were volatilized so that the bright image could be safely assigned
to calcium and magnesium. For these experiments, ashless gelatin
impregnated with the chlorides was hardened in 10% formalin,
dehydrated in alcohols and embedded in paraffin. Paraffin sections
( 10 fi.) were placed directly on the prepared cathode.

The nickel cathode is prepared for use by polishing with optical
rouge and washing with water and nitric acid. It is then coated
with a mixture of 40% barium carbonate and 60% strontium
carbonate in a 2% solution of nitrocellulose in amyl acetate. The
barium-strontium mixture serves to increase the emission of the
calcium and magnesium in the tissue by about 1000%, by activa-
tion, and at the same time the mixture emits electrons itself to give a
contrast background on the screen. When completely dry, a 10 /x
section of tissue embedded in paraffin, prepared by the freezing-
drying technique (see page 3), is placed on this surface and
smoothed down with a steel needle. The cathode is then inserted into

152 MICROSCOPIC TECHNIQUES

the microscope tube and the vacuum pumps are started. When the
pressure falls to 10 •'' mm. mercury, or less, the cathode-heating
filament is turned on and the temperature is gradually elevated
over an hour or more to the operating level. The slow heating is
essential to avoid distortions in structure, to minimize or prevent
curling on the cathode, and to approximate microincineration condi-
tions in order to permit a comparison. The practice is followed of
heating the cathode higher than operating temperature for a short
time to volatilize the sodium and potassium and to initiate active
emission of electrons. When this activation period is over, and the
operating temperature obtained, the high-voltage and lens currents
are turned on. The electron image formed on the screen is compared
with stained or incinerated control sections. When the tissue curls
away from the cathode and is improperly oxidized, dark areas appear
in the image and carbonization is evident by optical examination.
The magnification is altered by changing the position of the mag-
netic lenses on the tube and then bringing to focus by adjustment
of the lens current. Electron-accelerating voltages of 5000-6000 were
employed. Various portions of the section are focused on the center
of the screen by virtue of the mobility of the lenses.

Photographs of the image are made with a high speed camera
(/ = 2.9) at a camera magnification of 0.73 on Eastman Kodak
Superspeed Ortho Portrait or Panchro-Press film. Exposures of 1-5
sec. are usually employed. The films are developed with Eastman
Kodak D-72 developer. Some photographs obtained in this fashion
are shown in Figures 31 and 32.

G. RADIOAUTOGRAPHY

The novel and thus far very limited technique of radioautography
has been employed in a few instances for the localization of radio-
active elements in tissue sections. The use of these isotoi)es as tracers
in biochemical investigations, particularly as the result of the
pioneering work of Hevesy, is a well-established device; however
histological distribution cannot be determined quantitatively, as yet,
on the basis of radioactivity, by any very satisfactory means,
since the order of the intensity of the radiation produced in the
quantities of tissue commonly employed for histological examina-
tions is far too small to permit suitable measurements M'ith the

RADIOAUTOGRAPHY

153

Geiger-Miiller counter or the electroscope. Radioautography, based
on the ability of emanations from radioactive elements to affect the
photographic plate, is an attempt toward the solution of this difficult
problem. Tissue sections containing radioactive elements leave their
"autographs" on photographic plates when placed in contact with

Fig. 31. Emission electron micro-
graph (X300) showing calcium and
magnesium distribution in rectus ab-
dominus muscle of cat. Note strong
cross-bandings in muscle fibers. From
Scott (1943)

Fig. 32. Emission electron micro-
graph of cat gastric mucosa (fundus)
showing calcium and magnesium dis-
tribution (left), compared with a
microincinerated section from the
same animal (right). Magnification
about X75. From Scott (1943)

them for a sufficient period. When developed, these "autographs"
indicate to some degree the relative distribution of the substances
responsible for the radioactivity.

Historically, the first use of radioautography was made by
Lacassagne and Lattes ( 1924) for the demonstration of polonium in
tissue. Since that time the usefulness of this technique, as well as all
others employing radioactive tracers, has been greatly expanded by
the recent revolutionary developments which have made possible
the preparation of radioactive isotopes of elements that occur in
living systems. Limiting factors in regard to the suitability of a
radioactive isotope for studies by radioautography are the nature
and intensity of its radiation and its half-life period. The duration
of the photographic exposure will depend on these factors as well as
on the concentration of the isotope in the tissue. The half-life
periods of the principal artificial radioactive elements that might be
used in tracer studies are given in Table V.

Perhaps, the greatest deficiency of the technique of radioautogra-

154

MICROSCOPIC TECHNIQUES

TABLE V. Principal Artific^ial Radioactive Isotopes Used as Trace Elements
as Compiled by Pool and Kurbatov (1943)

Radioactive

Atomic

Intensity of

Type of

element

Half-life

weight

activity

radiation

2 . 1 min.

15

Strong

+

N

9.93

13

Strong

+ 07

Mg

10.0

27

Strong

-0y

Co

11

60

Strong

-07

C

21.0

11

Strong

+0

Ag

24.5

106

Strong

+

I

25.0

128

Strong

-0y

CI

37.5

38

Strong

-0y

In

54

116

Strong

-0y

Zn

57

69

Strong

-0

Ba

1.42 hr.

139

Strong

-0y

F

1.8

18

Strong

+

Se

1.81

75

Weak

+

Y

2.0

88

Strong

+0

Cr

2.27

55

Weak

-0

Mn

2.59

56

Strong

-0y

Si

2.60

31

Strong

-0

Ni

2.6

63

Strong

-0y

Ti

3.1

45

Strong

+

Sc

4.0

43

Strong

+ 0y

Br

4.45

80

Weak

y

Ab

7.5

211

Weak

(xy

K

12.4

42

Strong

-0y

I

12 6

130

Strong

-0y

Cu

12.8

64

Strong

-0, +0

Au

13.0

196

Weak

-0

Pd

13.0

109

Weak

-0

Zn

13.8

69

Weak

y

Na

14.8

24

Strong

-0y

Pt

18

197

Weak

-0

Co

18.2

55

Strong

-\-0y

W

1.01 day

187

Weak

-0y

Sn

1.05

121

Weak

-0

As

1.11

76

Strong

-0,+0y

La

1.70

140

Strong

-0y

Ni

1.5

57

Weak

+

Br

1.66

82

Strong

-0y

R A DIO A U TOG R AP H Y

155

TABLE V (Concluded)

Radioactive

Atomic

Intensity of

Type of

element

Half-life

weight

activity

radiation

Cd

2.3 days

115

Strong

-0y

Au

2.7

198

Weak

-Py

Mo

2.8

99

Weak

-/3

Ag

7.5

111

Strong

-0

I

7.8

131

Strong

-M

Ca

8

41

Weak

y

Ag

8.2

106

Strong

y

Sn

10.0

123

Weak

-/3

P

14.5

32

Strong

-0

V

16

48

Weak

+ 0y

As

16

74

Strong

-^, + l3y

Rb

19.5

86

Strong

-/3

Cr

26.5

51

Weak

y

Be

43

7

Weak

y

Fe

47

59

Weak

-fiy

Sr

55

89

Strong

-0

Sb

60

124

Strong

-0y

Zr

63

93

Weak

-0

Ti

72

51

Weak

-0y

W

74.5

185

Weak

-0

Sc

85

46

Strong

-0y

S

88

35

Weak

-0

Ta

97

182

Weak

-0y

Y

105

86

Strong

y

Ca

180

45

Weak

-0y

Zn

250

65

Weak

+ 0y

Mn

310

54

Weak

y

Cs

1.7 yr.

134

Weak

-/3->

V

1.71

47

Weak

7

Na

3.0

22

Weak

+ ^7

Co

5.3

60

Weak

-0y

H

31

3

Weak

■ -&

Ra

1590

226

Strong

a

C

10000

14

Weak

-0

U

7.1 X IQs

235

Weak

a.

U

4.5 X 109

238

Weak

a.

Rb

1 X 10"

87

Weak

-0

156 MICROSCOPIC TECHNIQUES

phy lies in its inability to reveal distribution in the finer structures,
and to this lack of resolving power must be added the further draw-
back that, quantitatively, only a rough approximation is possible.
However, there is the considerable advantage, inherent in the use of
radioactive elements regardless of whether a histological or gross
tissue study is involved, that very small quantities of an element
introduced into a biological system can be followed without reference
to, or interference from, the large stores normally present. The
amount of a radioactive element that can be detected is fortunately,
very minute. According to Hamilton (1941) a total of 2 X 10^
(i particles, with an average energy of at least 150 Kev., are
required to strike each cm.^ of photosensitive surface to yield a
satisfactory image.

Reviews dealing with radioautography have been presented by
Hamilton (1941-1942) and Simpson (1943), and important physical
data have been furnished in a review by Kurbatov and Pool ( 1943) .

Preparation of Radioautographs*

Both fresh-frozen and paraffin sections of tissue have been used
to obtain radioautographs. In general the paraffin sections give the
hest results since they can be cut thinner and are less subject to
distortion. It is essential that the sections be of uniform thickness
and free of wrinkles. There would be a particular advantage in the
use of freezing dehydration (page 3) for the preparation of the
paraffin-infiltrated tissue. The diffusion of the radioactive substances
during fixation and dehydration in solutions would be eliminated
and a more authentic "autograph" could be obtained.

As examples of procedures which have been used the following
may be cited: Hamilton, Soley, and Eichorn (1940), in a study of
radioactive iodine in thyroid tissue, removed the paraffin from 3-5 /x
sections with xylol, dipped the slide containing the sections in dilute
celloidin, allowed it to dry, and obtained a celloidin film over the
sections about 1 fx. thick. The sensitive surface of the photographic
film was placed in contact with the celloidin surface. Harrison,
Thomas, and Hill ( 1944) in an investigation of the distribution of

* See Bibliography Appendix, Refs. 20, 21, 22, 2S, and 30.

RADIOAUTOGRAPHY 157

radioactive sulfur in the wheat kernel, employed paraffin sections
25-50 p. thick which were covered directly by a layer of aluminum
foil 0.8 jx thick. The sensitive photographic surface was placed in
contact with the foil.

The greater the distance between the tissue and the photosensitive
surface, the poorer the resolution in the radioautograph due to
scattering of the radiation. It is preferable that this distance be kept
under 1 mm. Ultraspeed x-ray film has been extensively used, but
it has the disadvantage of producing grainy enlargements. Harrison,
Thomas, and Hill (1944) recommend a fine grained panatomic film
when it is possible to have longer exposures. The photographic film
over the sections on a glass slide is covered with another slide and
the whole bound firmly together with cellulose tape. All of these
operations are carried out in a dark room, of course. After wrapping
the slide in light-tight black paper, it may be placed in a cold place
for the duration of the exposure. It is advisable to keep the sections
cold to inhibit any tendency toward diffusion of the radioactive
element. After exposure, the sections are stained in the usual manner
to bring out their morphology, and compared to the developed
"autographs" with the aid of a dissecting microscope.

More recently, Belanger and Leblond (1946) extended the useful-
ness of radioautography by the ingenious expedient of spreading a
photographic emulsion directly on the sections. This not only permits
a more intimate contact between the tissue and the photographic
surface, but it obviates the matching of the "autograph" to the cor-
responding histological detail, which is particularly difficult at higher
magnifications. The possibilities of this technique merit a more
detailed description of the procedure.

Belanger and Leblond Technique

Preparation of Photographic Emulsion. Soak lantern slide
plates (medium contrast, Eastman) in distilled water at room
temperature. When the gelatin swells, remove from the water. With
a glass knife scrape off the gelatin, and melt it in a beaker placed
in a 35-40° bath. Carry out this procedure and all others in which
the emulsion is used in a dark room. A Wratten "Safelight — No. 1"
(Eastman) may be used at a distance of about 3 ft.

158

MICROSCOPIC TECHNIQUES

m. int

h.int.

X""-

mol. tub

Nj. ^m. int.

O

i

">!.

2A

3 A

V4 «»^

3B

■n js.

4 A

Fig. 33. Radioautographs (A) and corresponding stained sections (B) (X8).
White areas in radioautographs are exposed parts of film. (1) Thorax of adult
mealworm, transverse section, paraffin; (2) abdomen of adult mealworm, trans-
verse section, paraffin; (3) abdomen of wax moth larva, longitudinal section,
frozen; (4) abdomen of wax moth larva, transveise section, frozen; (g)
ganglion; (gon.) gonad; (hy.) Iwpodermis; (mal. tub.) malpighian tubule;
(m. int.) midintestine; (mus.) muscle; (h. int.) hind intestine; (rep. org.)
reproductive organs; (s. g.) silk gland. Fi(ii)( Lindsay and Craig (1943)

1 !

Mk

L .''

■•i«

W

I^:' .

^

'^^:v^

"/.*'■'

- ii

■M^

0^ ■

• ■

*

't.

_.i-^

J

Fig. 34. Radioautographs of adult rat tissues. (1) Cross section through the
lower limb (X50). The radiophosphorus is in the diaphysis of the tibia and
the fibula. Arrow A points to heavy periosteal layer in the tibia, (i?) Para-
median longitudinal section of thoracic vertebra (X50). Arrow B indicates
phosphorus deposit in the ossifying neural arch. (5) Cross section of the lower
jaw (Xl7). The deposition of radiophosphorus clearly outlines the mandible.
In the right portion of the bone an incisor tooth is developing and is also
impregnated with the radioactive element. (4) Section of the thyroid from an
adult rat (X70), treated with radioiodine. The tracheal cartilage is visible at
the bottom of the figure. The thyroid follicles show a reaction due to radio-
iodine. Friim Belnnger and Leblond (19.'/i)

160 MICROSCOPIC TKCHNIQUES

PROCEDURE

1. Prepare 10 /x })araffin sections and attach to slides with egg
albumin.

2. Dry, deparaffinize with xylol, and carry through absolute
alcohol to 1% celloidin in alcohol-ether.

3. Drain off excess celloidin by standing the slides in an empty
Coplin jar.

4. Place in 70% alcohol for about 1 min. to harden the celloidin,
and dry at room temperature.

5. In the dark room, pipette 5 drops of the melted photographic
emulsion on to each slide and spread evenly with a camel's hair brush.
Carry out this operation a little below 40° on a hot plate to prevent
premature gelling of the emulsion. (The temperature should be held
below 40° with this emulsion to reduce the fogging.)

6. Allow to cool and dry, and place the slides horizontally in a
dustjiroof, light-tight box for the duration of the exposure.

7. x\fter the exposure develop for 3-4 min. in Kodak developer
D72 at 18-20°. Wash rapidly in water and fix for about 10 min. in
5% thiosulfate at the same temperature. Wash finally in cold run-
ning water for about 30 min. The black silver deposit indicates the
site of the radioactive element.

8. Counterstain in Coplin jars cooled in running water. (Methyl-
ene blue and Harris hematoxylin may be used for radiophosphorus
"autographs" and Harris hematoxylin for those of radioiodine.)
With methylene blue place slides in a 1% alkaline soln. for about
30 min. and rinse in running water until the stain is removed from
the gelatin coating. With Harris hematoxylin, stain lightly to avoid
the need of differentiating. (Artifacts caused by gelatin swelling and
disengagement of the sections disappear when the slides are thor-
oughly dried after each operation.)

9. Pass through several changes of 95% alcohol, absolute alcohol,
and xylol, and mount in Canada balsam. (Clearing in oil of origanum
before mounting will also give good results.)

Discussion

Points on various procedures required for particular studies may
be obtained by referring to some of the applications already made.
To date, most of the investigations employing radioautography

RADIOAUTOGRAPHY 161

deal with the isotope P^-. Thus its distribution was studied in bones
by Dols et al. (1938) and Belanger and Leblond (1946), in tomato
fruits by Arnon et al. (1940), in squash plants by Colwell (1942),
and in insects by Lindsay and Craig (1942). "Autographs" have
been obtained in bone studies with radioactive calcium and strontium
by Pecher (1941) and Treadwell et al. (1942). The distribution of
radioactive lead in the animal body was investigated by Behrens
and Baumann (1933a,b), and thyroid studies were carried out with
radioactive iodine by Hamilton et al. ( 1940) , Gorbman and Evans
(1941), and Belanger and Leblond (1946). Harrison, Thomas, and
Hill ( 1944) employed radioactive sulfur for a radioautograph survey
of the distribution of this element in wheat. Many new applications
are constantly appearing.

CHEMICAL TECHNIQUES

"By calling attention to the cell I desired to provoke
investigators to inquire into the processes within the
cell, to define that which happens within these small-
est elementary organisms. And it was self-evident that
an exact definition could be nothing else than to find
the chemical and physical foundations upon which

vital phenomena and the activity of the cell are
based."

ViRCHOW

as quoted by Paul Klemperer in Some Recent Biologic
Investigations and Their Significance for Pathology,
J. Mt. Sinai Hosp. N. Y. 14: 442 (1947/48).

INTRODUCTION

The chemical techniques to be described are all of the quantitative
variety and they differ from their macro counterparts primarily as
regards the volumes employed and the mode of handling them. In
general, the same reactions and concentrations of reagents are used
in both. The degree to which the localization of the chemical constit-
uents in tissues and cells is limited, in these techniques, largely
depends upon the degree to which the anatomical parts can be
isolated mechanically in preparation for their separate analysis. The
precedures most commoly used are: (a) the preparation of serial
microtome sections of tissue and analysis on each of selected sec-
tions, (fc>) isolation of cells or cellular particulates by centrifugation
for their separate analyses, or (c) use of micro dissection to obtain
the part to be analyzed. It is considerably more of a problem, as a
rule, to obtain a satisfactory sample for analysis than to perform
the analysis itself. While the ultimate goal of being able to apply
quantitative procedures in situ to biological material is still essen-
tially beyond the present horizon, the use of these chemical
techniques can lead to the acquisition of knowledge which can now
be obtained by no other means.

It should be pointed out that in the interests of simplicity and
accuracy certain well-established procedures of macroquantitative
analysis are best avoided in work on the level considered here. The
procedures to be given are those of the original authors, but the
laboratory worker should introduce his own simplifications of the
following type at every opportunity : (1) Avoid quantitative trans-
fers — rather remove an aliquot. (2) Avoid dilution to a given volume
in a vessel; this necessitates the calibration and marking of the
vessel — rather dilute by adding a known volume of liquid with
a pipette. (5) Employ pipettes calibrated to deliver rather than to
contain — this obviates the necessity of rinsing the pipette. (4) Avoid
filtration when it is possible to separate a precipitate by centri-
fugation.

165

i. GENERAL APPARATUS AND
MANIPULATION

A. VESSELS, STOPPERS, HOLDERS, ETC.

Vessels. Most of the reactions employed in the various chemical
techniques are carried out in simple glass vessels. The tube shown
in Figure 35 is especially useful; it is nothing more than a small test
tube having a total capacity of 0.25 ml. (available from A. H.
Thomas Co. and E. Petersen, Carlsberg Laboratory, Copenhagen
Denmark). Norberg (1937) employed the tubes (Fig. 36) for use
in centrifugation and the apparatus shown in Figure 37 for removal
of supernatant fluid from centrifuged precipitates. By applying
suction at A the fluid is drawn into the reservoir; the low-power
microscope is used to enable careful control of the operation. As
indicated in Figure 37, the tube may be surrounded by a larger
vessel filled with a clear liquid, such as alcohol, to permit better
observation, particularly when the vessel B (Fig. 36) is used, since
the bottom part of the tube has a dark zone due to its form.

The tubes may be cleaned conveniently by immersing in the clean-
ing liquid, heating to drive the air out of the tubes, cooling to let
them fill up with the liquid, and shaking out the liquid from each
one. Usually the process is repeated two or three times. To place
films of liquid across the upper portion of a reaction tube, as in
iodometric titrations, Holter and Doyle ( 1938) employed the vessel
shown in Figure 38, which has a total volume of 0.20 ml. These
vessels must be given an inner hydrophobic coating to prevent the
liquid from spreading on the glass surface. The method of doing
this is described on page 169.

Levy (1936) used the 2.5 ml. tube illustrated in Figure 39 for the
Kjeldahl digestions and Linderstr0m-Lang, Weil, and Holter (1935)
employed the two-piece unit shown in Figure 40 for ammonia dis-

166

«Id>

^

u

Fig. 35. Reaction
vessel in holder.
From Linderstr0m-Lang
and Holler (1933a)

W V

10

'-20 mm

Fig. 36. Centrifuge

reaction vessels.
From Norberg (1937)

rO

"-10 cm.

Fig. 37. Arrangement for removing

supernatant fluid over a precipitate.

From Norberg (1937)

■ 5 mm

// w

^WA

Fig. 38.

(r\

Ml

-^ -^-S mm.

Fig. 39.

\y

Fig. 40.

Fig. 38. Reaction vessel for iodometric titration. In neck, lower film is
sulfuric acid, upper film starch solution. From Holler and Doyle (1938)

Fig. 39. Kjeldahl digestion tube. From Levy (1936)

Fig. 40. Vessel for ammonia distillations (no longer used). From Ldnder-
str0m-Lang, Weil, and Holler (1935)

168

CHEMICAL TECHNIQUES

tillations, e.g., in arginase measurements. The ammonia was distilled
from vessel A into cap B, which was coated internally with paraffin
and charged with standard acid. Ramsay grease (thick) was used
to seal the parts together. Later it was found preferable to abandon
the use of this form of vessel for ammonia distillation (Briiel et al.,
1946) (see page 283).

Glass diffusion cells for the distillation of ammonia were described
first by Conway and Byrne (1933), and later by others (Figs. 41-
43) . Ammonia diffuses from the outer well into standard acid con-

F

67 mm.-
61 mm.-
40 mm.-
35 mm-

^ "R^J

^^- ~^

"' "

Ie

in

Fig. 42, Gibbs and
Kirk (1934) diffusion
cell (cross section, one
half actual size).

D

\it ill UJ lU ll< OJ

Fig. 41. Conway and Byrne (1933)
diffusion cell. Above, top view;
below, vertical, section on line AB.

Fig. 43. Kinsey and Robinson
(1946) diffusion cells: upper, top
view; lower, side view.

tained in the center well in the types shown in Figures 41 and 42
(available from Microchemical Specialties Co.). The Kinsey and
Robison (1946) apparatus (Fig. 43) consists of a Lucite plate 0.5
in. thick with rings 1 mm. deep having inner and outer diameters of
14 and 18 mm., respectively, reamed out of the plastic for one form
of cell (A); for the other form (B), rings of the same diameter but
8 mm. deep are reamed out and a center hole 4 mm. deep and 8 mm.
in diameter is drilled. In the A form cell, two glass vials are used
alone. Ammonia diffuses into the receiving solution placed in the

GENERAL APPARATUS AND MANIPULATION 169

bottom of the outer or larger vral. When the vial is inverted and
set on the plate this solution forms a hanging drop over the liquid
which is liberating ammonia.

A small open porcelain dish {Micro chemical Specialties Co.) (Fig.
44 heknv) was used by Kirk and associates as a titration vessel.

P^ig. 44. Titration dish, actual size.
Frotn Kirk and Bentley (1936)

Coating Vessels with a Hydrophobic Layer. At times it is
desirable to coat reaction vessels with a hydrophobic layer to pre-
vent aqueous liquids from spreading on the glass surface, as in
iodometric titrations where liciuid films are placed across the lumen
of the neck of the titration tube. Linderstr0m-Lang and Holter
( 1933a) used paraffin and Holter and Doyle ( 1938) employed
ceresine. The procedure followed by the latter was to boil about 50
vessels for 5-10 min. in 75 ml. of water to which 0.1 g. ceresine was
added. After the water had cooled, the vessels were emptied and
dried for at least 3 hr. at 100-110°.

The procedure finally employed at the Carlsberg Laboratory for
paraffin coating was described by Brliel et al. (1946). The clean,
dry glass tubes are immersed in melted paraffin at 150-200° (the
synthetic paraffin used had a melting point of 82°), picked out one
at a time with forceps, quickly emptied and rotated in a clean towel
between the fingers until the paraffin solidifies. A heavy layer of
paraffin on the bottom of the tube and a thinner one on the upper
part is desirable. The outside of each of the tubes is wiped free of
paraffin and they are stored protected from dust and fumes. After
the vessels have been used, they are cleaned by rinsing first with
water, then with acetone, hot toluene, acetone, and water in the order
given.

Stoppers. For most work it is sufficient to stopper reaction tubes
with a short piece of rubber tubing one end of which is plugged with
a glass bead or short piece of glass rod. A stopper consisting of a
cap with a small hole (Fig. 45) is useful in some cases as in the
addition of alkali in the method of Linderstr0m-Lang and Holter
(1933b I for ammonia. In this same method a stopper was used

170 CHEMICAL TECHNIQUES

having a drawn-out piece of glass tubing to plug one end (Fig. 46)
so that the larger air space would prevent the displacement of the
liquid film, which was across the tube, when the stopper was fitted
on. To protect solutions from atmospheric carbon dioxide, Linder-
str0m-Lang, Weil, and Holter (1935) employed stoppers containing
soda lime tubes (Fig. 47).

Tube Holders. Perhaps the simplest holder for a small reaction
tube is a short length of thick- walled rubber tubing into which the
bottom of the tube may be placed, as in Figure 50. It is more con-
venient to use a small wooden or metal block with three flexible
metal prongs to hold the tube. For titration, where the color of the
solution is to be matched with a color standard, a single block with
prongs to hold two tubes is used (see Fig. 64, page 180; A. H. Thomas
Co. and E. Petersen, Carlsberg Laboratory).

Rediiclor. Kirk and Bentley (1936) devised a small glass volu-
metric flask of either 0.1 or 0.2 ml. capacity for use as a reductor
in their method for the estimation of iron (Fig. 48). The reductor
is made from heavy-walled 2 mm. bore capillary tubing. In the iron
method (page 277) cadmium amalgam is employed to reduce the
iron.

Tube and Pestle. For the grinding of bits of tissue, Glick ( 1937)
used a small pestle with a 250 /xl. tube having the inner bottom sur-
face ground as shown in Figure 49.

B. MICROLITER PIPETTES

Pipettes of various designs have been employed for measuring
microliter volumes. The chief among these will be described.

Fixed Pipettes. One of the pipettes developed by Linderstr0m-
Lang and Holter (1931) is shown in Figure 50 (A. H. Thomas Co.
and E. Petersen, Carlsberg Laboratory). It consists of a capillary
tube drawn out to a tip which is slightly bent so that contact can
be made with the wall of a vessel. The pipette is calibrated by first
weighing it, and then filling it with water to a little more than the
i-equired volume. The i)ipette is again placed on the pan of the
balance and water is removed by touching a piece of filter paper to
the tip. When the desired weight of water remains in the pipette, it
is removed from the balance and a mark is placed at the meniscus,
either by etching with hydrofluoric acid or using a piece of gummed

GENERAL APPARATUS AND MANIPULATION

171

Fig. 45. Reaction tube cap with hole.
From Linderstr0m-Lang and Holier (1933b)

Fig. 46. Stopper
for reaction tube with
cap of drawn-out glass
tubing. From Linder-
str0m-Lang and Holler
(1933b)

Ky

Ground
inner
surface S^^/

I

Ground
surface

Fig. 47.

Fig. 48.

Fig. 49.

Fig. 47. Soda lime tube for stoppering reaction vessel. From Linderslr0m-Lang ,

Weil, and Holler (1935)
Fig. 48. Reductor. actual size. From Kirk and Bentley (1936)
Fig. 49. Tube and pestle for grinding tissue. From GUck (1937)

172 CHEMICAL TECHNIQUES

paper. In the assembly shown in Figure 50, the pipette is filled by-
applying gentle suction through the tube S with H closed and K
open. When the liquid is a little above the mark, K is closed and the
slowly falling meniscus is observed through the low-power micro-
scope (M). The moment the meniscus reaches the mark, the vessel
of liquid is quickly lowered away from the tip and the capillary
forces will prevent the liquid from running out of the pipette. The
vessel into which the liquid is to be delivered is brought up so that
the pipette tip touches the vessel wall near the bottom and H is
opened. P leads to a source of compressed air, and the pressure
regulator ( T) enables the hquid to be forced out of the pipette under
constant pressure. Usually a 20 cm. column of water gives the
required pressure; the emptying time should not be less than 5 sec.
H is not to be closed until the delivered liquid has been lowered away
from the pipette. With pipettes having a capacity of 7 ix\. the error
of pipetting was found to be less than 0.3%.

Hand Pipettes. A hand pipette (A. H. Thomas Co. and E. Peter-
sen, Carlsberg Laboratory), Figure 51, having an accuracy of about
1 % was also used by the Carlsberg group. The instrument is filled or
emptied by sucking or blowing through the attached rubber tubing.
The tip of the pipette is fine enough to prevent liquid from running
out unless a slight pressure is applied through the rubber tubing.
Hand pipettes, in which the suction or pressure is applied by a glass
syringe, have been used by Kirk's group, Kirk and Craig (1932),
Sisco, Cunningham, and Kirk (1941) (Figure 52) {Microchemical
Specialties Co.). A rubber gasket fixed to the end of the syringe
barrel receives the large end of the pipette, or the metal syringe
fitting of a hypodermic needle is cemented to the pipette in order to
permit easy attachment to, and separation from, the syringe.

Constriction Pipettes. The preceding types of pipette have been
displaced very largely by the constriction pipette (Levy, 1936;
Linderstr0m-Lang and Holter, 1940) shown in Figure 53 (A. H.
Thomas Co. and E. Petersen, Carlsberg Laboratory) . In this pipette
the calibration mark is replaced by a constriction in the lumen of
the capillary. Liquid is first sucked up over the constriction and a
slight pressure is then applied which causes the liquid to fall down
to the constriction but not past it. To deliver the charge, a momen-
tary greater pressure is applied to force the meniscus through the
constriction and then gentle pressure can be employed to empty the

173

H

-r- -

"&

c=a=^

[M]

M

Fig. 50. Fixed pipette.
From Linderstr0ni-Lang and Holler (1931)

Fig. 52. Capillary pipette and syringe
control. From Sisco, Cunningham, and
Kirk (1941)

n

Fig. 51. Hand pipette.
From Linderstr0m-Lang
and Holler (1933a)

-o-

-n

=TJ

<]

Fig. 53. Constriction pipette:

(A) proper form; (B) faulty form.

From Linderslr0m-Lang and Holler (1940)

}

^^

Fig. 54. Constriction pipette

with water jacket.
From Holler and Doyle (1938)

174 CHEMICAL TECHNIQUES

pipette. If employed in the assembly shown in Figure 50, a quick
squeezing of the rubber tubing over K will be sufficient to initiate
the emptying process. Greater accuracy is obtained by adjusting the
dimensions so that the pipette delivers automatically without apply-
ing excess pressure w^hen the tip touches the vessel or the liquid.
The tip and the constriction should be constructed as in A (Fig. 53) ,
and not as in B. Holter and Doyle ( 1938) employed a constriction
pipette surrounded by a water jacket to control the temperature of
the liquid being pipetted (Fig. 54).

In the procedure for the determination of total nitrogen (pages
234 and 283) the pipettes used must meet certain dimensional re-
quirements as defined by Briicl et al. (1946). Thus, the pipette used
to transfer the digested sample must have a tip, the opening of
which is not so narrow as to become blocked by small crystals or
other particles; but neither must it be so wide as to make it difficult
to empty the pipette without blowing air through the tip, which
might cause spattering of the liquid delivered. Furthermore, the
pipette stem must be thin enough for use in a narrow tube without
causing the liquid to be drawn up between the tube and pipette, and
yet it must be thick enough to have mechanical strength. A suitable
pipette is illustrated in Figure 55, and the allowable variation in the
dimensions is shown in Figure 56, which gives the dimensions of a
rather thin and a rather thick pipette, either of which may be used.

The dimensions of a suitable constriction pipette for pipetting the
acid used to absorb ammonia are given in Figure 57.

Placing a water seal of known volume across the lumen of a reac-
tion tube is best performed with the type of constriction pipette
shown in Figure 58. Water is drawn up to the point X; the entire
amount is blown out to form the seal, and then the excess water is
sucked back into the pipette to Y. The amount left in the seal is
then the volume between X and Y in the pipette. Acid-selenium mix-
ture is pipetted with the horizontal pipette illustrated in Figure 59.
Each division corresponds to 1 //.I. These pipettes are available from
E. Petersen, Carlsberg Laboratory.

Automatic Pipettes. An automatic pipette was designed by
Linderstr0m-Lang and Holter (1931) and it was used by them for
the accurate delivery of 20—40 /xl. alcoholic acid to stop enzyme ac-
tion (Fig. 60) {A. H. Thomas Co. and E. Petersen, Carlsberg Labo-
ratory). The pipette consists of a narrow glass tube drawn out to a

175

0.4

1 1 1 1 ,

1.0

1

0.2

0.2

ry////////^

0.8
0.6

\'////////Z^.

0.4

1 0.6

-

io.2

<■

UJ 0.4

J UJ

D < 0.2

^^"

DIAMETER,
o

E

E

j^^_

0.4

o

0.2

\^^/7/77777y>^

0.6

0.4

_^^^^^^^^22^%

0.8

0.6

(

. , , , "

1

3 10 20 30 40
LENGTH, mm.

30

10 20 30 40 50 60 70 80
LENGTH, mm

Fig. 55.

Fig. 56.

Fig. 57.

Fig. 55. Pipette for transfer of digested sample in nitrogen determination.
From Brilel et al. (1946)

Fig. 56. Allowable variation in dimensions of pipette in Fig. 55. From
Brilel et al. (1946)

Fig. 57. Dimensions of constriction pipette for pipetting acid used to absorb
ammonia in nitrogen determination. From Brilel et al. (1946)

y

-X

■ 45 n\.

-Y

-A/il.

Fig. 59. Horizontal mouth-operated

pipette with vertical delivery tip.

From Brilel et al. (1946)

Inner diameter 0.35 mm.
Outer diameter 0.70 mm.

Fig. 58. Constriction pipette
for placing liquid seal across

neck of reaction tubes.

From Brilel et al. (1946)

Fig. 60. Auto-
matic pipette.
From Linderstr0m-
Lang and Holler
(1931)

176 CHEMICAL TECHNIQUES

capillary at both ends so that, with one atmosphere pressure, the
fineness of the tips will prevent liquid from running out. 22 is a
siphon arm connecting to a reservoir of the liquid to be pipetted
which is placed about 50 cm. above the instrument. The pipette is
filled as follows: Close L and open K and H to fill the outer cham-
ber. When the level is a few mm. over the upper tip of the pipette,
close H; the pressure from the reservoir will fill the pipette. Then
close K and open // and L to bring the level of the liquid in the outer
chamber below the upper tip. Deliver the pipette charge by closing
H and L and opening K, which compresses the air in the chamber and
forces the liquid out. A pipette of this type having a capacity of 30
^il. was found to have an error of measurement of less than 0.1%.

Accurate syringe pipettes have been employed which use a screw
(Krogh and Keys, 1931; Krogh, 1935) or a micrometer spindle
(Trevan, 1925) to move the plunger of a small hypodermic syringe.
The micrometer syringe pipettes are essentially the same as the
micrometer burettes of Dean and Fetcher (1942) and Hadfield
(1942) (page 255). The Krogh-Keys instrument is manufactured
by Macalister Bicknell Co. It has a delivery precision of 0.1 ix\.

Devices for Drawing Finer Pipettes. Finer pipettes which are
used under a microscope may be drawn by hand, but mechanical
devices for making them are considerably more efficient. DuBois
(1931) described an automatic device for drawing very fine micro-
pipettes and microneedles which has been made available commer-
cially (Leitz). A capillary tube is clamped in two arms of the device,
and between the arms the tube passes through a small electric heater.
When the glass softens in the heater the spring tension on the arms
pulls them back, thus drawing out the tube into a pair of pipettes.
Rachele's device, described by Benedetti-Pichler and Rachele ( 1940) ,
operates in a similar manner except that gravity is used to pull out
and lower the arms when the glass is softened by the electric heater.

C. FILTERS

Sintered-glass filters for small volumes of liquid were used by
Kirk and co-workers for the quantitative collection of precipitates
for various determinations. The filter described by Cunningham,
Kirk, and Brooks (1941b) is made of capillary tubing (2 mm. in-

GENERAL APPARATUS AND MANIPULATION

177

ternal diameter, 6 mm. external diameter) the end of which is
tapered and contains a fused-in 4 mm. section of sintered glass at
the tip. Kirk ( 1935) discussed the preparation of sintered-glass
filters for those who cannot obtain them commercially. A layer of
fine asbestos 1 mm. thick is sucked onto the tip of the filter to form
a pad; after pressing this pad down with the fingernail, a layer of
asbestos is deposited on the pad to form a cone about 2 mm. deep
The filter is used as shown in Figure 61. Only the tip of the cone ia
allowed to be in contact with the liquid in order that tlie precipitate
may be collected entirely on this part of the asbestos. The precipitate
can be transferred quantitatively by disengaging the pad at its base.
These filters are available from Microchemical Specialties Co.

Fig. 61. Filtration detail. A represents

an asbestos filtering cone; B, an asbestos

base pad; and C, a sintered-glass plug.

From Cunningham, Kirk, and

Brooks (1941b)

Fig. 62. Ultrafilter for small

volumes of liquids.
From Johnson and Kirk (19IiO)

Bott ( 1943) employed a capillary tube filter with paper pulp (Fig.
76) in the determination of sodium. A description of the preparation
of filter and its use is given on pages 204-207.

Ultrafilters. Various devices have been employed for the ultra-
filtration of small volumes of liquids, and only that of Johnson and
Kirk (1940), designed for about 0.1 ml., will be described, since
it is one of the simpler and more efficient types. The ultrafilter shown
in Figure 62 consists of two brass tubes, A and B, drilled to fit snugly
the glass capillary tubes having a bore of 2 mm. or less and an out-
side diameter of 7-8 mm. Kronig cement ( 1 part white wax and 4

178

CHEMICAL TECHNIQUES

parts rosin melted together) is used to bind the glass to the metal.
A brass collar, C, engages B and is threaded to A. For visibility, the
center part of the collar is cut away on two sides. The ends of the
capillary tubes are flared and ground fiat, and a piece of collodion
membrane is held tightly between the ground surfaces at D. The
ultrafilter is used with positive pressure.

D. STIRRING DEVICES

Many investigators have used a stream of carbon dioxide-free
air bubbles ejected from a fine glass tip to obtain stirring in manipu-
lations of a submacro order. However, when very small volumes of
liquid are to be stirred, recourse must be had to other means. When
dealing with a drop of liquid in a capillary tube, a practical method
of agitation appears to be the simple expedient of moving the drop
back and forth in the tube by means of controlled air pressure.
Schmidt-Nielsen (1942) devised a centrifuge for sealed capillary
tubes which rotates them so that liquid contained within will be
thrown from one end to the other to effect mixing. The apparatus,

, rt R ^

■ Ampullae

63. Centrifuge apparatus for mixing and extracting small amounts
of liquid in ampules. Length of the apparatus about 25 cm.
From Schmidt-Nielsen (1942)

shown diagrammatically in Figure 63, is attached to the motor
shaft and revolves with it. During the revolutions the plate with
the ampules is slowly turned, being connected by means of a
rubber band to a small rubber wheel which in turn is being driven
by its frictional contact with the motor housing. The rubber wheel

GENERAL APPARATUS AND MANIPULATION 179

is suspended in such a way that it is held against the motor housing
by means of the rubber band. The tubes are turned once for about
each five revohitions.

Agitation of a liquid in a capillary tube was effected by Bessey
et al. (]946) by touching the side of the tube to a rapidly rotating
nail head (page 250).

For stirring small volumes of liquid in an open shallow vessel,
Kirk ( 1933) employed the tip of a glass needle drawn from the end
of a tube in which a piece of iron was sealed. The core of an electric
buzzer placed in proximity to the iron made the glass needle vibrate
and stirring was thus produced. This type of stirrer is manufactured
by Micro chemical Specialties Co.

A particularly effective and convenient stirring device for small
volumes in open or closed vessels is the electromagnetic "flea" of
Linderstr0m-Lang and Holter ( 1931) . The "flea" consists of a sealed
glass spherical shell, about 1-2 mm. in diameter, filled with ferrum
reductum; stirring is effected by an electromagnet repeatedly turned
off and on by means of an interrupter. The core of the magnet is
placed near the outer wall of the vessel, and the lifting and dropping
of the "flea" provides the agitation. The arrangement employed in
titration is shown in Figure 64. The interrupter is not shown; it is a
small glass-enclosed mercury switch mounted on a pivot which is
connected to a movable strip of iron in the field of the magnet. When
the current is turned on, the magnetized core tips the iron strip,
which tilts the mercury switch and thus breaks the current. The iron
strip falls back when the core is no longer magnetized, and in so
doing it brings the mercury switch back to its original position,
which again completes the circuit, magnetizes the core, and starts the
process over again. "Fleas" are made by blowing a small bulb in the
end of a drawn-out piece of glass tubing, tapping a little ferrum
reductum down into the bulb, and sealing off the neck with a micro-
flame. The "fleas" may be cleaned by rinsing with water and cover-
ing them with fuming nitric acid. After washing well with distilled
water, the "fleas" are allowed to dry on a piece of filter paper. The
development of a brown stain of iron oxide on the filter paper under
a "flea" indicates that it has an incomplete seal and it should be
discarded. A method of testing the "fleas" suggested by Linderstr0m-
Lang and Holter ( 1940) is to place a drop of neutral bromothymol
blue solution on each one on a glass plate. Those not properly sealed

180

CHEMICAL TECHNIQUES

will become apparent since the acid that seeped into them during
the cleaning will cause the indicator to turn yellow. The complete
stirring equipment is available from A. H. Thomas Co. and E.
Petersen.

istiH_3> -=C_M}

nT

1

(qM§)

100

98

96

Fig. 64. Microtitration arrangement

for use with magnetic "flea" stirrer.

From Linderstr0m-Lang and Holier (1940)

Heatley, Berenblum, and Chain (1939) employed steel ball bear-
ings, ^/i6 in. in diameter, given several coats of Bakelite varnish No.
V-5209/2. Each coat was polymerized by stoving before applying the
next, and then the balls were given a layer of paraffin by heating
them to 100° in a paraffin bath. After excess paraffin was removed
by rolling the bearings on hot filter paper, they were rolled in the
clean, dry palm of the hand with some well-washed kaolin to enable
them to be wetted by aqueous solutions. Of course these balls should
not be used with liquids that might attack the coating.

E. HEATING DEVICES

A simple micro muffle furnace was described by Kirk and Bentley
( 1936) which, when employed with the proper rheostat, can be
used for temperatures up to 1000°. The furnace is made by winding
Chromel A resistance wire around a porcelain cup (2 in. inside di-

GENERAL APPARATUS AND MANIPULATION

181

ameter, 4 in. deep), embedding in insulating material, and surround-
ing with iron or brass pipe.

Linderstr0m-Lang (1936) employed an incineration oven for the
ashing of samples in small tubes. The oven (Fig. 65) is made of a
solid copper block containing holes 18 mm. deep and about 7 mm.
in diameter. Two electric heaters placed at the sides of the block
enable a temperature of 440-460° to be maintained. A rheostat is
used to obtain lower temperatures and to regulate the heating. The
sides and bottom of the oven are insulated with asbestos. The tubes
used with this oven were of quartz and had an inner diameter of
3.8 mm., an outer diameter of 6 mm., and a length of 20 mm. A
solid copper block with holes drilled to accept small tubes for
Kjeldahl digestions was used with gas heat by Borsook and Dubnoff
(1939), as shown in Figure 66.

Copp'

Fig. 65. Incineration oven

to accommodate small tubes.

From Linderstr0m-Lang (1936)

8 mm. H h-

Fig. 66. Digestion rack and Kjeldahl tubes.
From Borsook and Dubnoff (1939)

Naturally, modifications in the micro furnaces and ovens may be
made, and commercially available types such as that of Micro-
chemical Specialties Co. are also often suitable.

F. MOIST CHAMBERS

When working with small volumes of liquid it is necessary in
certain instances to maintain a moist atmosphere around the liquid
to prevent evaporation. The various forms of the moist chambers
employed on the stage of a microscope have been described by
Chambers and Kopac (1937). In general these are partially en-

182

CHEMICAL TECHNIQUES

closed chambers containing strips of wet filter paper, or a layer of
water on the floor.

Holter (1945) described a large moist chamber into which the
hands may be placed for various operations, and into the top of
which a binocular dissecting microscope is fitted to enable observa-
tion of the material being manipulated. The humidity is kept con-
stant through the regulation afforded by an electrically fitted hy-
grometer connected to a circulation pump which supplies an adjust-
able proportion of wet and dry air.

Holter Moist Chamber. The chamber is illustrated in Figure
67. It is made of varnished plywood ; the dimensions of the box are
64 X 35 X 30 cm., not considering the bevelled surfaces on the ends
of the front which contain doors 12 X 12 cm. The hands may be
inserted through tight-fitting rubber cuffs attached in the openings
of the doors. A slanting glass window (20 X 20 cm.) is fitted into
the front of the box and is hinged so that larger objects can be

f^\-^

Fig. 67. Air-conditioning chamber.
From Holter (1945)

Fig. 68. Arrangement of

air-conditioning apparatus.

From Holter (1945)

placed inside. The edges of a sheet of rubber are sealed into a hole
( 15 X 15 cm.) cut in the top of the box; holes in the rubber fit tightly
around the tubes of a binocular dissecting microscope. The sheet of
rubber is rather limp and bulging so that vertical movements of the
microscope will not cause it to stretch unduly. Illumination of the
interior is supplied through a window in the back wall which, for
some purposes, should be screened with a heat-absorbing device.
A moist atmosphere is maintained in the box by means of the

GENERAL APPARATUS AND MANIPULATION 183

arrangement shown in Figure 68. An electric circulation pump (P)
sends a current of air, divided by the T-tube (a), into the chamber.
The water in the large bottle (B) is warmed by a lamp (about 40
watt) to a temperature 5° higher than that of the room, and the
copper coil attached to the arm (6) is surrounded with water cooled
5° below the room temperature. The air passing into the water in
(B) is dispersed into fine bubbles. The currents of warm and cool
air enter the chamber at the same corner and the baffle (d) permits
them to mix without allowing water droplets to be carried into the
center of the box. The box contains a thermometer (/) and a hair
hygrometer (e) fitted with an electrical contact at the percentage of
moisture desired. By means of pinch cocks on tubes g and b, the
warm air stream is first regulated so that the temperature rise in
the chamber is about 1° in 15 min. when the cool air is shut off, and
then sufficient cool air is admitted to compensate for this tempera-
ture rise. The contact on the hygrometer is connected through a relay
to the pump which automatically stops when the desired humidity
is attained and starts when it begins to fall. A thermoelectric control
of the air flow through a or 6 could be used to achieve finer regula-
tion, but Holter found it unnecessary for his work.

G. ELECTRODES

Linderstr^m-Lang, Palmer, and Holter Silver Electrode. A

simple electrode arrangement (Fig. 69) was used by Linderstr0m-
Lang, Palmer, and Holter (1935) for the micro determination of
chloride by electrometric titration. The silver wire electrodes (A)
and (^4') are employed as shown. A is fixed with a bit of picein in
the side tube of the tip of the burette (B). Before sealing it in the
side tube the wire is cleaned with a little cold dilute nitric acid, and
afterward it is kept in contact with the acid silver nitrate titration
solution. In this manner it need not be changed for months. A^
requires occasional cleaning with a little nitric acid and, if neces-
sary, fine emery cloth may be used. After titration this electrode is
dried with filter paper, taking care not to touch it with the fingers.
The tip of the burette may be protected from contact with A' by
the glass cap shown on the right of Figure 69.

Sisco, Cunningham, and Kirk Glass Electrode. An open-cup
glass electrode was used by Sisco, Cunningham, and Kirk (1941)

184

CHEMICAL TECHNIQUES

for formol titrations in the manner shown in Figure 70. The cup is
made by blowing a bulb and then sucking in a depression. The

\^

Fig. 69.

Fig. 70.

Fig. 69. Silver electrodes (A and A') arranged for chloride titration with
burette (B) and electromagnet on the left to enable stirring with the "flea" in
the vessel (see page 282). Tip of burette may be protected from contact with
A' by the glass sleeve shown on the right. From Linderstrdm-Lang , Palmer,
and Holler (1935)

Fig. 70. Cross section of the glass electrode titration vessel. B represents an
inverted glass electrode, C a burette, D a reference calomel cell, E a glass
tube. A is a cup assembly, consisting of: a, a central block; b, a lower cup;
c, an upper inverted cup. From Sisco, Cunningham, and Kirk (1941)

outside of the cup is coated with paraffin and the electrode is filled
with 0.1 A'" hydrochloric acid saturated with quinhydrone. The
chamber (A) is made of Lucite. Stirring is effected by blowing a
streari of nitrogen through the tube (E) in such a fashion as to
whirl the sample drop in the cup.

Claff and Swenson Glass Capillary Electrode. Glass electrodes
employed for measurements of the hydrogen ion concentration of
small volumes of solutions have been numerous. One of the more

GENERAL APPARATUS AND MANIPULATION

185

Lucite wafer
(insulator)

Picein
plug

Lucite wafer
(insulator)

Heated glass
rod

Lucite wafer

Paraffin

Corning 015
glass capillary

Fig. 71. Capillary glass electrode assembly.
From Claff and Swenson (1944)

186 CHEMICAL TECHNIQUES

recent of these is tluit described by Claff and Swenson ( 1944) , which
can be used for voliunes as little as 5 ix\. and is capable of repro-
ducibility in measurement of ±0.02 pH units. The apparatus is
indicated in Figure 71. The glass electrode is a capillary tube of
Corning No. 015 glass 75 mm. long, attached to two Lucite wafers
with picein as shown. The capillary assembly is fitted in a jig so that
the wafers will always be a fixed distance apart. Starting 4 mm. from
each end, the capillary is brushed with hot paraffin up to the wafers.
One end of the capillary tube is plugged with picein and the open
end dips into a drop of the saturated potassium chloride solution
from the calomel half cell. The portion of the capillary tube between
the wafers is immersed in 0.1 A^ hydrochloric acid contained in a
flattened Pyrex funnel. The funnel is filled to capacity, surface
tension preventing the solution from overflowing. The stem on the
funnel is bent upward to form a silver-silver chloride electrode, and
the funnel unit is mounted on an insulated standard so that it can
be raised and lowered or moved horizontally. The entire assembly
is electrically shielded, and shielded leads connecting to the pH meter
are used. The capillary tubes are cleaned by sucking through them in
the following order: Keego cleaner {J. B. Ford Co.) 0.1 N hydro-
chloric acid, alcohol, distilled water, and, if blood is to be used,
0.2% potassium oxalate. The tubes are then dried by drawing air
through them. The tubes are stored in 0.1 N hydrochloric acid when
not in use.

Pickford Sealed-In Capillary Glass Electrode. A permanently
mounted capillary glass electrode was described by Pickford ( 1937) .
The capillary, made of Corning No. 015 glass, is sealed into the
apparatus as shown in Figure 72. The three-way stopcock is of the
type developed by Stadie, O'Brien, and Lang (1931). Of course it
must be made of the same kind of glass as that used in the electrode
jacket. A fine bore ( 1 mm. diameter) in the stopcock plug is required
for filling in order to keep the volume of the sample as small as
possible. The bore of the filling and connecting tube is about 0.5
mm. A 1 ml. syringe (preferably of the short insulin type) is em-
ployed to fill the electrode, using the assembly shown. The syringe
may be used to obtain the sample, in which case the needle would be
removed after the sample was taken. A short piece of hemocytom-
eter tubing (3 mm. inside and 5 mm. outside diameter) is then
slipped over the nozzle of the syringe and this is connected to the

GENERAL APPARATUS AND MANIPULATION

187

One way
stopcock

Bulb CaO. 1 n-^l -

Stdndard buffer
lOf 1 A' HCI/

Capillary glass
membrane

Ag-AgCI
electrode

To calomel

cell via

KCI flusti

To calomel

cell via

KCI flusfi

Diagonal bore

for filling,

diameter 1 mm

Three way
stopcock

Rubber
i.ubing

/

iMose ol Luer syringe

Fig. 72. Sealed-in capillary-
glass electrode assembly.
From Pickjord (1937)

D

O

O O

D

O
o o

B

D D

Fig. 73. Conductivity cell. The
upper section shows the faces of the
two blocks which make the cell (actual
size), the center section, the construc-
tion, and below is an enlarged dia-
gram of the recess which holds the
fluid. Fro7n Bayliss and Walker (1930)

cup attached to the filling inlet. The electrode is calibrated with
standard buffer, washed, and then dried with alcohol and ether.
After the sample has been introduced the lower stopcock is rotated
clockwise, first to flush saturated potassium chloride through the
groove and right- angle bore (the position shown in Fig. 72) and

188 CHEMICAL TECHNIQUES

then to make liquid junction between the sample and the calomel
half cell. After use the electrode is washed with dilute salt solution
and kept filled with distilled water. The outer jacket is filled with
0.1 A^ hydrochloric acid. Pickford used a Pliotron tube amplifier
with the apparatus, and the electrodes were made by Macalister
Bicknell Co.

H. CONDUCTIVITY APPARATUS

The Bayliss-Walker Cell. A conductivity cell was designed by
Bayliss and Walker (1930) for measurements on as little as 0.5
ix\. of liquid (Fig. 73). Two blocks of vulcanite (A, B) are shaped
and drilled as shown. Two small holes are drilled near the upper
edges and in one block an enlargement is made to form a recess (C)
0.5 mm. deep, 0.75 mm. wide, and 1 mm. long in the face of the
block. In each of the small holes platinum wire (0.3 mm. diameter)
is sealed with sealing wax, taking care that the end of the wire is
flush with the bottom of the recess in the one case and with the face
of the block in the other. Glass tubes {E) are sealed into the blocks
with sealing wax, as shown, and when filled with mercury enable
electrical contact to be established. The blocks faced to fit perfectly
against each other are located by two pins and held firmly together
by a thumb screw so that a small cavity having a platinum electrode
in each face is formed. A conical hole (D) is made in the face of the
block nearest the cavity. The recess is lined with sealing wax, which
has to be replaced from time to time as the surface deteriorates.
Each electrode is coated with platinum black by covering with a
drop of 2% platinum chloride, dipping a wire into the drop, applying
3 volts to the circuit, and reversing the polarity every 10 sec. for
2-3 min. It is necessary to keep the cell in distilled water, drying it
only before use, in order to prevent a drift in the resistance during
measurements. It is also necessary to reblack the electrodes every
week or so.

The cell is filled through the conical opening with a fine pipette,
using a low-power microscope to aid in the operation. The pipettes
are made of small-bore glass tubing drawn out for about 10 cm. to
a diameter of around 0.2 mm. or a little less. Each is cut at points
20-30 mm. from the beginning of the constriction where the diameter
begins to be uniform, and the center capillary is discarded. The ends

GENERAL APPARATUS AND MANIPULATION 189

of the pipettes thus formed are bent at right angles about 5 mm.
from the small end. Care must be taken that the ends are cleanly
and squarely cut. A mercury leveling bulb connected to the pipette
with rubber tubing may be used for filling and emptying. Small
air bubbles sometimes cling to the cell walls or to the electrodes when
the cell is filled. This difficulty is overcome as a rule by withdrawing
the fluid into the pipette and refilling the cell more slowly.

The circuit used by Bayliss and Walker was the standard Kohl-
rausch bridge fed from a 1000 cycle audiofrequency generator. The
null point was determined with head phones in the usual way. Be-
cause of the small size of the electrodes and the necessity of drying
them, the null point tends to be flat. A 0.01 /xF. condenser placed
across the standard resistance makes the null point sharp. The
practice is to adjust the resistance until minimum sound intensity
is obtained with the slide wire at midpoint, or until a sharp increase
in intensity occurs symmetrically on each side of it.

I. BALANCES

The development of the instrumentation for the weighing of very
small amounts has been thoroughly review-ed by Gorbach (1936).
The commercial balances, including the torsion balances of Roller
Smith Co., which are sensitive down to about 2 fig., require no
comment here. The quartz fiber balances, which are considerably
more sensitive, are particularly useful in histochemical work and
these will be considered in detail.*

Quartz Fiber Balance. Lowry (1941) designed a simple and
serviceable quartz fiber balance that can handle a maximum load
of 200-300 fig., and that has a sensitivity of about 0.03 fig. and a
reproducibility of 0.1 jug. The functioning of the instrument (Fig.
74) depends on the measurement of the bending of a horizontal
hollow quartz fiber when a weight is attached to its free end. The
fiber (A), about 20 cm. long, is drawn from narrow quartz tubing.
One end of the fiber is fused at B to a low tripod (C) made of 1-2
mm. quartz rod. Instead of fusing the fiber to a quartz tripod, the
end can be inserted into a short Pyrex sleeve (having a lumen just

* Since this writing a new quartz fiber balance has been described by Kirk
et al.; a "Cartesian-diver" balance has been announced by Zeuthen (Bibliog-
raphy Appendix, Refs. 39, 49, and 54.)

190

CHEMICAL TECHNIQUES

large enough to hold it) fused to a Pyrex tripod. This enables easier
replacement of fibers. The Pyrex sleeve should lean toward the
front of the instrument at an angle of a little less than 90° to the
plane of the tripod. The free end of the fiber (D) is bent into a tiny
V in a plane at right angles to the fiber axis. Without a load, the
free end of the fiber should be 12-15 cm. above the tripod. The
tripod is mounted inside a metal cylinder (E), a gallon tin can will
do or a smaller instrument can be made to fit into a smaller can. The
open front of the cylinder is fitted with a removable glass plate {F).

tz^

_^.w

•10 cm-

N

Fig. 74. Quartz-fiber balance.
From Lowry (194V

The tripod is fixed in place with DeKhotinsky cement, and the
cylinder is mounted rigidly on a heavy wooden block. A cathetometer
(Q) , reading to 0.01 mm., is used to observe the positions of an
arbitrary point on the fiber tip when weights are applied. Illumina-
tion of the interior of the cylinder can be enhanced by removing
the back end of the can and replacing it with a plate of glass.
Electrostatic shielding can be increased by lining the inside surfaces
of the glass plates with metal foil from the center of which strips
have been cut to act as windows. For less accurate measurements
without a cathetometer, a narrow ribbon of graph paper running
down the center of the front window can serve for the indication of
the displacement of the end of the fiber.

GENERAL APPARATIS AND MAXIPlLATIOISr 191

The samples to be weighed are held in quartz fiber hooks (G)
1 cm. long with loops at each end 2 mm. in diameter. These are
made from 3-4 cm. lengths of solid fiber weighing about 30 /xg./cm.
One end of a piece of the fiber is held in a small oxygen flame so
that the force of the flame bends the tip as it softens. By proper
manipulation a complete circle can be made. The weight is now
adjusted by clipping off the straight end with a scissors until the
desired deflection of the balance is obtained with it. When a number
of these hooks have been brought to the same weight within a few
mm. deflection, the second loop is made in each one. The hooks are
stored on a glass rack (H) which has a series of pegs (J) 0.5 mm.
in diameter projecting from the large horizontal tube at 3 cm.
intei-vals. The fine glass springs (K) prevent hooks from falling or
blowing off. The hooks are handled by a glass rod (L) about 1 mm.
in diameter drawn out at the end to 0.2 mm. A 5 mm. bend (M) is
made in the end of this rod to slip into the loop for transfer. During
attachment or removal of hooks at the end of the fiber, a straight rod
(N) is used to hold the fiber.

After the case has been closed for 1.5-2 min., readings may be
taken; successive observations have been found to agree to 0.03 mm.
One hook is kept as a standard weight. Calibration of the balance
is carried out by accurately pipetting 3-10 /xl. of standard salt
solution into the lower loop of a hook (if 1-2 fx\. of distilled water is
placed in the loop first, the transfer of the salt solution is easier),
drying the solution, and observing the deflection given by the known
weight of salt. Deflections given by various weights are plotted to
form a calibration curve.

In order to obtain the dry weight of microtome sections of tissue,
Lowry places a 3-5 (A. drop of water in the lower loop of a weighed
hook with a fine-tipped pipette, and then places the section in the
drop with a fine rod. Hooks with sections are put on the rack, dried
at 100° for 30 min., and reweighed. For the measurement of neutral
fat, the hooks with the dried sections may be kept in ethyl or
petroleum ether for 30 min., redried in the oven, and reweighed.

Quartz Torsion Balance. A quartz torsion balance was devised
by Lowry (1944) that has a capacity of 50-100 mg. and a sensi-
tivity of ±0.1 fig. The instrument is shown diagrammatically in
Figure 75. The beam (A) is a quartz tube 25 cm. long and about 1
mm. in diameter suspended between the horizontal quartz fibers (C) .

"-v.

192 CHEMICAL TECHNIQUES

The quartz stand (B) supports these fibers. Fine quartz loops in the
ends of the beam hold quartz hooks (E) from which the aluminum
foil pans (D) are suspended. The two arms of the beam need not
be of exactly the same length. The feet (G) of the standard are
sealed to the floor of a balance case with DeKhotinsky cement. It is
convenient to employ the usual mechanism in analytical balances
that lifts the beam when loads are added to, or removed from, the
pans. Electrostatic shielding is achieved by lining the inside of the
balance case with metal foil in which windows (H) are cut. In addi-
tion, the members of the balance are metalized by coating them with
a 5% solution of chloroplatinic acid in alcohol and, after drying,
heating with a "cool" flame to effect conversion to metaUic platinum.
Particular care is required to avoid overheating the fine quartz sus-
pensions. A strip of aluminum foil is used to ground the instrument
to the case.

Fig. 75. Quartz torsion balance.
From Lowry (1944)

After the case has been closed for at least 1 min., measurement of
the displacement of one end of the beam produced by the load placed
on a pan is made with a cathetometer (F) reading to ±:0.01 mm. In a
particular balance constructed by Lowry, a load of 10.8 ^ug. pro-
duced a displacement of 1.00 mm. The cathetometer may be focused
on any convenient landmark on one end of the beam. An illuminated
piece of white paper outside the opposite end of the balance case
furnishes a background that facilitates the measurement.

Calibration of the balance is carried out by cutting 5-10 cm. of
fine wire, weighing 1.5-2 mg., into ten nearly equal lengths, weigh-

GENERAL APPARATUS AND MANIPULATION 193

ing the ten pieces together on a microbalance, and then observing
the displacement given by each piece placed separately on a pan.
The sum of the individual displacements divided by the total weight
gives the sensitivity. The process must be repeated on the other pan
if both arms are to be calibrated. The balance will accommodate
larger weights if a tare is used as a counterbalance.

//. COLORIMETRIC TECHNIQUES

A. CAPILLARY TUBE TECHNIQUE

During the course of their classical investigations dealing with
the composition of glomerular urine, Richards and his group at the
University of Pennsylvania developed a simple and clever technique
of capillary tube colorimetry which enabled them to carry out
analyses on less than 1 fx\. liquid with an accuracy comparable to
that of macro procedures. The chief problem, as stated by Richards
et al. (1933), was "to introduce the minute amount of fluid to be
analyzed into a capillary tube without evaporation or contamination,
to dilute it quantitatively with water if necessary, to introduce into
the same capillary in quantitatively accurate proportions and with-
out mixing the one or more reagents required for production of color,
to effect mixture of the fluids in the capillary tube at a given moment,
and to compare the resulting color with those developed in standard
solutions treated simultaneously in identical or equivalent fashion."

The recent introduction of microcuvettes for the colorimetry of
small volumes of liquid in photoelectric apparatus (page 216) will
very largely displace capillary tube colorimetry because of the ob-
vious advantages of greater objectivity and accuracy of the analyses,
and the greater ease of manipulation in most cases. However, the
capillary tube technique and methods are included here because
there are instances in which the equipment for the cuvette methods
is not available, or the volumes to be handled are still too small to
permit the use of cuvettes, even of the micro variety. Furthermore,
some of the capillary tube methods might be adapted to cuvette
colorimetry when the equipment for the latter is available, and in
that case the assembly of the methodology of the former would also
be useful.

195

196 CAPILLARY TUBE COLORIMETRY

1. Apparatus

Capillary Tubes. For blood collections, plasma protein precipi-
tations, and for the making of pipettes, capillary tubing having an
outside diameter of 0.8 mm. and an inside diameter of 0.6-0.7 mm.
was employed by Richards et al. (1933). The capillaries in which
reactions were produced and colors developed were 0.5 mm. outside
diameter and 0.35 mm. inside. These smaller tubes must have very
uniform bores and hence it is necessary that they be drawn mechan-
ically.

Pipettes. The pipettes are drawn from the larger capillary tub-
ing. Their slender tips should have an outside diameter of about 50
[x; the over-all length should be about 10 cm. Liquid is drawn up and
expelled in the pipettes by means of an attached piece of rubber
tubing through which suction or pressure may be applied.

Microscope. A binocular microscope giving about fifteen fold
magnification with an optical field of about 1 cm. in diameter is
recommended. For the microscopic measurements a micrometer disc
is placed in one of the oculars or the disc is cemented to the glass
stage of the microscope. The disc should have a 10 mm. scale divided
in 0.1 mm. In order to reduce the chance of evaporation of fluids,
the glass stage, with the exception of the circle visible in the optical
field, is covered with wet filter paper.

Water Manipulator. For the introduction and movement of
columns of fluid in the capillary tubes, controllable suction or pres-
sure must be applied at one end. A small syringe having a piston 3
mm. in diameter moved by a micrometer screw serves this purpose.
The tip of the syringe is connected by rubber tubing with a short
glass or metal tube drawn out at one end to a tip small enough to
enter the capillary tube. The syringe, rubber tube, and tip are filled
with colored water, care being taken to exclude air bubbles, and
mounted on a level with the microscope stage. When water is forced
out of the tip into the capillary tube, a water seal is formed which
permits the movement of water into or out of the capillary. In this
fashion columns of liquid may be introduced into the capillary tube
from the other end and their movements can be easily controlled.

Other Accessories. A small centrifuge is required that will hold
the capillary tubes. A piece of unglazed milk glass (35 cm. X 35 cm.
X 4 mm.), two desk lamps fitted with 100 watt bulbs, and a sus-

APPARATUS AND MANIPULATIONS 19?

pended lamp equipped with a 150 watt bulb and Daylight glass
filter are also needed.

2. Manipulations

1. Connect a length of capillary tubing to the water manipulator,
and fix the tube on the stage of the microscope so that it is parallel
to, and lying on, the micrometer scale with its open end near the
edge of the optical field farthest from the manipulator.

2. Force water from the manipulator into the tube until half of
its length is filled.

3. Bring the tip of a pipette filled with the solution into the
optical field and insert it into the open end of the capillary tube.
Carefully blow the liquid out of the pipette, at the same time draw-
ing it into the tube by turning the piston screw of the water manipu-
lator. The volume of liquid introduced is determined by the length
of the column as measured on the micrometer scale visible through
the tube. When the appropriate amount of the solution has been
introduced move the column inward so that its distal meniscus is
near the center of the field.

4. In the same manner introduce columns of reagents, and, when
these have been added, break off the portion of the tube containing
all the columns (about 3-4 cm. long) and seal both ends quickly in
a minute gas flame. Set aside in a horizontal position. When breaking
off the tube, caution is required to avoid including any portion of
the tube which has been wetted with the manipulator water ( a dia-
mond point is useful for cutting the tubes at the proper place) and
when the ends are sealed, care must be exercised to avoid heating
adjacent liquid columns.

5. In order to mix the solutions, briefly centrifuge the sealed
tubes to bring the separated columns together, invert, and again
centrifuge. Then repeat the inversion and centrifugation. (This may
be simplified, see page 178.)

6. When it is necessary to heat the mixture, place the tubes in
a hot water bath.

7. Since it is essential that color comparisons be made with tubes
of the same diameter, the tubes to be compared should be obtained
from the same original length of uniform-bore tubing. When many
tubes are to be compared, the various original lengths of tubing re-
quired may not have the same diameters. In this case break a single

198 CAPILLARY TUBE COLORIMETRY

30 cm. length of unifonii tubing into 2 cm. pieces. Transfer the
colored solutions to these pieces by breaking off the sealed ends of
the tubes, inserting one end of a tube into small rubber tubing held
in the mouth, and placing the other end in contact with the piece
into which the solution is to be transferred. Apply gentle pressure to
effect the transfer and quickly seal the ends of the tube with plasti-
cine, taking care to avoid contact between the plasticine and the
solution in the tube. In order to prevent blowing the liquid from one
tube right through the other, the two tubes should be held in the
position of a wide-angled V during the transfer.

8. For the comparison of blue colors, place two desk lamps fitted
with 100 watt bulbs side by side about 6 in. over the milk glass plate.
The use of two lamps prevents shadows. For colors at the red end
of the spectrum suspend a 150 watt lamp provided with color filters
over the plate. Place the standard tubes, one at a time, beside the un-
known on the plate for comparison. It is sometimes helpful to cover
the tubes with a piece of white paper in which a rectangular window
has been cut so that the visible columns are of the same length.

note: In some cases it has been found that the intensity of the color
developed when minute quantities of test solution and reagent are mixed
in capillar}' tubes is not the same as when the liquids are mixed in the
same proportions in macro quantities. However, when this difference does
not exist, it is obviously less laborious to prepare the series of standard
color mixtures in macro volumes and transfer them to capillary tubes.

3. Methods

PREPARATION OF PROTEIN-FREE SUPERNATANTS

The following procedures were employed by Richards' group
( 1933) for frog plasma. The proportion of plasma to precipitating
reagent and the final dilution may be varied to suit the particular
kind of blood used. In general it is best to keep the dilution of plasma
as low as possible for capillary tube colorimetry.

Tungslic Acid Supernalanls

1. Collect the blood directly in one of the micropipettes. A few
grains of dry sodium oxalate may be placed in the pipette in advance.

METHODS 199

2. Seal off the larger end of the pipette in a minute gas flame
and centrifuge at once.

3. Cut the tube a little above the juncture of the cells and plasma,
let the plasma flow back from the cut end by gravity, and then seal
both ends of the tube taking care not to heat the plasma.

4. Attach a large capillary tube (0.6 mm. inside diameter) to
the water manipulator and fix it on the stage of the microscope.
Draw back 5 mm. from the end of a column of Vis N sulfuric acid
5.0 mm. long.

5. Introduce a 4 mm. column of plasma and add 10% sodium
tungstate to it until the column becomes 5 mm. long.

6. The two columns now in the tube are made to oscillate back
and forth several times by means of the water manipulator in order
to effect thorough mixing of the plasma and tungstate.

7. Break off the distal part of the tube and seal the ends. The
end nearest the acid is sealed last and held in the flame long enough
to make a small bulb.

8. Centrifuge the tube, bulb end down. Reverse and recentrifuge
at least six times for complete precipitation. The final centrifuga-
tion should be thorough and the material should be left in the nar-
row end of the tube.

9. Break the tube about 5 mm. above the surface of the fluid and
draw off the protein-free liquid into a pipette. Should a zone of hazi-
ness exist between the clear fluid and the precipitate, too much oxa-
late was used.

Trichloroacetic Acid Supernatants

1. For frog plasma phosphates a 4 mm. column of plasma is
placed in a larger capillary tube followed by 1 mm. of 90% trichloro-
acetic acid ( by weight ) . Seal the tube and centrifuge with the plasma
end down.

2. Immerse in hot water for a moment and centrifuge several
times inverting the tube each time.

3. Should the supernatant fluid be turbid, separate from the pre-
cipitate by cutting the tube, draw into another pipette, seal the large
end, and centrifuge at high speed.

4. Separate the sediment by cutting off the tube. The protein-
free liquid is ready for transfer to a mixing capillary for color devel-
opment.

200 CAPILLARY TUBE COLORIMETRY

Zinc Sulfate-Sodium Hydroxide Supernatanls

1. For frog plasma chlorides a 3.0 mm. column of plasma is
drawn in 1 cm. from the end of a 10-12 cm. capillary tube followed
by separate columns of 6.0 mm. of 0.1 N sodium hydroxide and 2.1
cm. of 0.64% zinc sulfate (ZnS04.7H20, freed from excess acid by
three recrystallizations from w^ater) .

2. Draw in the columns so that at least 2 cm. from the end of
the tube is empty, break off the portion of the tube containing the
liquids, and seal both ends in a flame.

3. Place in the centrifuge with the plasma uppermost and mix the
liquids by four centrifugations.

4. Immerse the tube for 30 sec. in water at 90-95° and centrifuge
once for 5 min.

5. Cut off the tube above the fluid and then cut off the part con-
taining the protein precipitate. The protein-free fluid is then ready
for analysis.

CHLORIDE

By the use of s?/m-diphenylcarbazide ( Cazeneuve reagent) West-
fall, Findley, and Richards ( 1934) increased the sensitivity of
Isaac's (1922) method for the determination of chloride and then
adapted it to capillary tube colorimetry. Their procedure allows
chloride determination in a fraction of a fA. of liquid containing 1 /xg
or less of sodium chloride with an average error of under 3.0%. The
principle of the method is that dry silver chromate will react with
chloride to precipitate silver chloride and leave the chromate ion,
which can be estimated by its yellow color. However, a much more
intense purple-red color will develop in the presence of diphenyl-
carbazide. For other methods see pages 224 and 281.

Westfall, Findley, and Richards Method for Chlorides

SPECIAL REAGENTS

Potassium Chromate Standards. Dissolve 3.321 g. of pure dry
potassium chromate (corresponding to 2.00 g. sodium chloride) in
1 1. distilled water. Dilute to prepare standard solns. in the range
10-70 milligram per cent sodium chloride at 2.5 milligram per
cent intervals.

CHLORIDE 201

Powdered Silver Chromate. Add slowly 200 ml. 5.5% potassium
chromate to 100 ml. boiling 10% silver nitrate soln. Add drops of
the chromate soln. until a slight excess is present as indicated by a
yellow color. Cool, wash the precipitate with water, and air-dry
on a Buchner funnel.

Diphenylcarbazide Reagent. Dissolve 0.5 g. syw-diphenylcarba-
zide {Eastman Kodak Co.) in 70 ml. 95% alcohol; add 25 ml.
glacial acetic acid, and make up to 100 ml. with distilled water.
This reagent is stable at 20° for only 3 hr.

PROCEDURE

1. Since it will be necessary to measure columns of liquid longer
than the diameter of the optical field of the microscope, mount a 15
cm. steel rule, graduated in 0.5 mm., on the microscope stage.

2. Fill a 10-12 cm. capillary tube (0.35 mm. inside diameter) to
nearly half its length with water from the water manipulator, and
adjust the tube so that its open end is in the optical field over the
zero mark of the stage micrometer and adjacent to the zero of the
steel rule.

3. Introduce a 2-3 mm. column of zinc sulfate-sodium hydroxide
supernatant (page 200) or other fluid to be analyzed and measure its
length accurately, then introduce just nine times as much distilled
water. If the concentration of the unknown corresponds to less than
0.1% sodium chloride, less water will be required; if higher than
0.7%, more water must be used.

4. Draw the liquid in 2 cm. from the open end, break off the
portion of the tube containing the added liquids, seal the ends in a
flame, and mix well by eight brief centrifugations, reversing the tube
after each one.

5. Seal one end of a larger capillary tube (0.6 mm. inside diam-
eter), place a few grains of dry silver chromate in it and tap the
tube to get the material down to the closed end. Take care that none
of the substance is left near the open end; cut off a little of the tube
if necessary.

6. Cut off the end of the first (smaller) tube above the column
of liquid, insert the open end into the silver chromate tube so that it
projects into it for about 1 cm. and fasten the two tubes together
with a ring of DeKhotinsky cement.

7. Centrifuge with the larger tube down for a moment so that

202 CAPILLARY TUHE COLORIMETRY

the liquid in the small tube is forced into contact with the silver
chromate in the larger tube.

8. Warm the cement, withdraw and discard the smaller tube, cut
away any of the larger tube to which the cement is adhering, and
seal the open end in a flame.

9. Drive the silver chromate back and forth through the liquid
by eight successive centrifugations. Continue the last centrifugation
for 5 min.

10. Make a pipette from the smaller capillary tubing and draw
the supernatant fluid into it.

11. Seal the larger end of the pipette, and force the liquid into
this end by centrifuging for 5 min.

12. Examine the tube under the microscope (magnification 50X)
to be sure the fluid is free of silver chromate particles. If not,
transfer to another pipette and centrifuge again.

13. Place a new piece of the smaller tubing 10-20 cm. on the
microscope stage with one end in the optical field and the other con-
nected to the water manipulator.

14. Introduce a 2.0 mm. column of the chromate liquid obtained
in step 12, and after measuring its length accurately draw it in at
least 2 cm. from the end of the tube.

15. Introduce a column of the diphenylcarbazide reagent four-
teen times the length of the chromate fluid, draw both columns in
2 cm. from the end, break off the portion of the tube containing the
liquid, seal the ends, and place in the centrifuge with the chromate
fluid uppermost.

16. Measure 4.2 ml. diphenylcarbazide reagent into each of three
test tubes. Start the centrifuge containing the capillary tube, and
as quickly as possible measure into the test tubes 0.3 ml. of each of
three standard chromate solns. covering the range of concentration
within which the unknown lies.

17. Centrifuge the capillary tube eight times, inverting it after
each centrifugation.

18. Fill a piece of capillary tubing at least 3 cm. long, having the
same diameter as that containing the unknown, from each test tube
and seal the ends with plasticine.

19. Compare the colors of the unknown and standards on a milk
glass plate under Daylite electric bulbs.

20. The result obtained gives a first approximation of the chloride

CHLORIDE AND SODIUM 203

concentration of the unknown. It may be necessary to repeat once or
twice with fresh portions of chromate supernatant in order that the
final color comparison may be made to standards differing from one
another by the equivalent of 1.25 milligram per cent sodium chlo-
ride. Smaller differences must be estimated.

SODIUM

A method for the determination of sodium in samples containing
as little as 0.3 fig., e.g., 0.2 /A. urine, with an average error of about
3%, was described by Bott (1943). Since this method employs a 6
ml. volume for the development of color, it is obvious that smaller
quantities might be measured if the colorimetry were performed on
smaller volumes. The method depends on the precipitation of the
sodium as sodium zinc uranium acetate, and the measurement of the
zinc in a solution of the salt by means of the red color it produces
with diphenylthiocarbazone. This principle was used by Deckert
( 1935) for the determination of zinc, and the technique of Bott
could be adapted to the measurement of small quantities of zinc.
For other methods see pages 265 and 270.

Bott Method for Sodium

SPECIAL REAGENTS

Water. The water employed in preparing all of the reagents is re-
distilled from an all-Pyrex still.

20% Trichlorocetic Acid. Made from acid that has been redistilled
from an all-Pyrex still.

93% Alcohol.

Ether. Redistilled.

0.01 N Sodium Hydroxide. Carbonate free.

Zinc Uranium Acetate Reagent. According to Butler and Tuthill
(1931): Prepare a soln. of 80 g. sodium-free uranium acetate,
U02(C2H302)2.2H20, and 48 g. or 46 ml. 30% (by vol.) acetic
acid in water to make a total of 520 g. Prepare a second soln. of
220 g. zinc acetate, Zn(C2H302)2-2H20, and 24 g. or 23 ml. of the
30% acetic acid in water to make a total of 520 g. Cover and warm
both solns. on a steam bath until, with stirring, soln. is complete.
Mix while hot, let stand 24 hr., and if no yellow precipitate ap-
pears add 0.2 g. precipitated uranyl zinc sodium acetate in order

204 CAPILLARY TUBE COLORIMETRY

to saturate the soln. Shake well and filter through quantitative
paper before using.

Magnesium Uranium Acetate Reagent. According to Blanchetiere
( 1923) : Dissolve 100 g. uranium acetate in 60 g. glacial acetic
acid and enough water to make 1 1. Dissolve 333 g. of magnesium
acetate in 60 g. glacial acetic acid and enough water to make 1 1.
Combine equal vol. of the two solns. Filter the reagent through
quantitative filter paper before use.

Diphenylthiocarbazone solution. Prepare immediately before use
by shaking 100 mg. of the compound {Eastman Kodak) in 5 ml.
of the sodium hydroxide soln. for 3 min. in a glass-stoppered ves-
sel the ground surfaces of which are thinly coated with paraffin.
Filter off the excess reagent on quantitative filter paper which
has been washed in redistilled water and dried before use. Dilute
1 vol. of the filtrate with 4 vol. of the sodium hydroxide soln. A
considerable variation in the quality of the compound from one
lot to the next has been observed.

Zinc Standards. Prepare pure sodium zinc uranium acetate by pre-
cipitating the sodium of pure sodium chloride with the zinc ura-
nium acetate reagent. Dissolve 0.235 g. of the triple salt in redis-
tilled water and make up to 1 1. This stock soln. contains 1 mg. zinc
per 100 ml. and it will keep for years in a Pyrex bottle in the dark.
Frequently prepare dilute standards containing from 10 to 70
microgram per cent of zinc by diluting the stock soln. with re-
distilled water.

PREPARATION OF FILTERS

1. Cut 3 cm. lengths of capillary tubing, 0.6 mm. internal di-
ameter. Cut ends squarely or the funnel openings to be made later
will be off center.

2. Partially seal one end of each piece of tubing by twirling in a
microflame. Use a microscope to observe the result. The opening
should be funnel shaped, about 0.1 mm. at the top and less at the
bottom (Fig. 76).

3. Prepare paper pulp by teasing apart the filter paper in re-
distilled water and drying at 105°. Bits teased off the dried pulp are
placed in the filter tubes and pushed down into the funnel end by a
thin capillary tube about 6 cm. long sealed at one end. Pack each bit

SODIUM

205

of pulp separately using, alternately, the open and sealed ends of
the thin capillary tube in order to obtain a filter well packed on the
sides and in the center. Make the packed filters 0.6-0.8 mm. thick,
and see that no spaces are present around the filter mat.

Fig. 76. Apparatus for determination of sodium. A and C are
approximately actual size. B is an enlargement (X20) of the
end of a filter tube. From Bott (1943)

4. Wash and test the filters by filling the tubes, with the aid of a
syringe and adapter, with redistilled water. Place them in a round-
bottomed centrifuge tube fitted with a mat of clean dry filter paper
on the bottom, centrifuge for about 1 min., examine the tubes, and
discard any which have not drained completely. Dry the filter tubes
at 105° in a clean vessel and store in covered weighing bottles kept
in a dust-free container.

206 CAPILLARY TUBE COLORIMETRY

PROCEDURE

1. As in the procedure for chloride (page 201), mount a 15 cm.
length of capillary tubing (0.35 mm. inside diameter, and check the
outside diameter with a stage micrometer — it should be just 0.5 mm.)
on the microscope stage and fix a 15 cm. steel rule beside it so that
the zero on the rule is opposite the 35 mark on the micrometer scale.

2. To one end of the tubing attach a water manipulator and place
the other end over the 30 or 40 mark of the stage micrometer so that
a 0.2-0.4 fA. sample will be in the center of the optical field.

3. Introduce a 2-4 mm. column of sample and draw it into the
tube just far enough to give a fully curved meniscus. Measure the
length of the sample column and pull it in about 5 mm.

4. Depending on the size of the sample, introduce rapidly from
a rather coarse capillary pipette, just filled with freshly filtered re-
agent, a 30-40 mm. column of zinc uranium acetate reagent. Meas-
urement of the reagent column need not be precise but do not move
the column back and forth, since evaporation of the liquid will give
high results. Draw the column in about 10 mm.

5. Cut off the portion of the tube containing the liquids, seal both
ends in a microflame without heating the liquids, centrifuge and in-
vert the tube ten times, allow to stand at room temperature for 10
min., and then centrifuge and invert five times more. During the 10
min. intervals examine the filter under the microscope and repack
it gently.

6. Examine the empty end of the precipitation capillary to make
sure no crystals have remained there. Cut off the end of the tube
containing the precipitate and insert it into the filter tube so that
the open end is about 6 mm. above the filter mat. Seal the two tubes
together with DeKhotinsky cement as shown in Figure 76, taking
care to avoid heating the reagent. Insert the tubes in a small hole in
a rubber stopper and fit into a test tube as in A of Figure 76.

7. Lower the assembly into a centrifuge cup by means of rubber-
tipped forceps, and spin rapidly for about 15 sec. Examine the capil-
lary and filter; usually a little liquid is found above the filter. Cut
off the sealed end of the capillary without disturbing the assembly,
centrifuge again for about 6 sec, and again inspect. No fluid should
appear above the mat. Centrifuge for 2 min. more to insure com-
plete draining.

SODIUM 207

8. Set the assembly in a wooden block. Dip the end of a clean
microfimnel (about 15 /A. capacity) such as pictured at the top of C
in Figure 76 into the magnesium uranium acetate soln. Fill the funnel
completely by pinching and releasing the attached rubber tubing.
Wipe off the outside, and slip the rubber tubing over the end of the
small capillary. Centrifuge for 6 sec, take off the funnel, and re-
centrifuge for 30 sec. Now repeat the process with two funnel fillings
of alcohol and two of ether. After evaporation of the ether, clean,
dry, yellow crystals should be seen on the filter mat.

9. Transfer the stopper and capillaries to a larger tube calibrated
to contain exactly 6 ml. (C, Fig. 76). Introduce a little redistilled
water into the small capillary, remove the funnel, and centrifuge for
a few sec.

10. Cut off the capillary about 1 cm. above the DeKhotinsky
cement, and proceed as before after attaching the funnel, filling the
funnel with water six to seven times. This should dissolve the pre-
cipitate and transfer the liquid completely into the tube which is
then filled with water to the 6 ml. mark.

11. From a 0.1 or 0.2 ml. Mohr pipette drawn out to a fine tip,
add 0.1 ml. diphenylthiocarbazone soln. to each unknown tube, and
also to each of three colorimeter tubes containing 6 ml. of water and
of two standards, respectively. Mix the contents of all the tubes and
transfer the unknowns to colorimeter tubes. The blanks should be
golden yellow, and the solns. with 10-70 microgram per cent of zinc
should vary from orange to cherry-red. Make colorimetric measure-
ments immediately. With the Evelyn colorimeter use Filter 565. The
calibration curve is linear, but at least one standard should be run
every time an unknown is run since variations may occur even
though the reagent is prepared the same way every time. The blank
should be found to be negligible, since measurements carried out on
redistilled water substituted for the sample gave maximum values of
only ±0.6 microgram per cent of zinc.

12. Since the inside of the very curved meniscus is used for
measurement of samples, make a vol. correction by the addition of
0.005 fA. for each meniscus. From the corrected vol., calculate the
final dilution (in a corrected sample of 0.2 lA. vol. the dilution is
30,000 times), and apply the following relation:

Na concn. _ 23.00 Zn concn. in ^ilnfi-nn
in sample " QdM ^ final soln. ^ a^ution

208 CAPILLARY TUBE COLORIMETRY

PHOSPHATE

A colorimetric capillary tube method for inorganic phosphate was
developed by Walker ( 1933) as an adaptation of the Kuttner ( 1927,
1930) modification of the Bell-Doisy phosphomolybdic acid method.
Walker's procedure enables analysis of as little as 0.08 fA. of liquid
containing less than 1 m^tig. phosphate phosphorus with a mean error
of about +0.1% and a mean deviation of about ±2.5% for solu-
tions of known concentration. A recent discussion by Sumner (1944)
of phosphomolybdic acid methods should be consulted. For other
methods see pages 124, 226, and 280.

Walker Method for Phosphate

SPECIAL REAGENTS

Standard Phosphate Solution. Prepare standards in the range of
1.5 to 7.0 milligram per cent phosphorus differing from one an-
other by 0.5 milligram per cent and below 1.5 milligram per cent
by 0.1 or 0.2 milligram per cent.

Molybdic-Sulfuric Acid Reagent. To 1 vol. of 10 A^ sulfuric acid
(282 ml. cone, acid, 95%, sp.gr. 1.84, to 1 1.) add 2 vol. distilled
water and 1 vol. 7.5% sodium molybdate soln. Store in a brown
glass-stoppered bottle.

Stannous Chloride Stock Solution. Prepare a 40% soln. in cone,
hydrochloric acid and store in a brown glass-stoppered bottle. Do
not use longer than one week.

Stannous Chloride Working Solution. Dilute the stock soln. 1:100
and prepare fresh each day.

PROCEDURE

1. Introduce a column of about 0.2 /xl. of the trichloroacetic acid
supernatant (page 199) or other phosphate soln. into a capillary tube
(0.35 mm. inside diameter) followed by an equal vol. of the molyb-
dic-sulfuric acid reagent and withdraw the two columns 3 cm. from
the end of the tube.

2. Introduce a third equal column of the stannous chloride work-
ing soln. and seal the ends of the portion of the tube containing the
liquids.

3. In a similar manner prepare tubes with the standard phos-
phate solns.

PHOSPHATE AND PHOSPHATASE 209

4. Centrifuge all the tubes at one time with the stannous chloride
soln. up.

5. Compare the colors on a milk glass background illuminated by-
two lamps arranged to avoid shadows. The colors fade by about
10% during the first few min. but this change occurs equally in all
of the tubes and hence need not interfere with the comparisons.

PHOSPHATASE

By an adaptation of the method of King (1932) to capillary tube
colorimetry, Weil and Russell (1940) worked out a procedure for
the determination of phosphatase which could be applied to less
than 1 ix\. plasma with an average deviation of ±3.0%. Their
technique employs capillary tube procedures for the determination
of the inorganic phosphate, but the enzymatic digestion procedure
is based on the use of pipettes, reaction tubes, and other apparatus
common to the titrimetric techniques. As in the method of Siwe
(1935b) (page 226), aminonaphtholsulfonic acid is used to reduce
the phosphomolybdate. For other methods see page 226,

Weil and Russell Method for Phosphatase

SPECIAL REAGENTS

Standard Phosphate Solutions. Prepare standards in the range of
0.02-1.00 /xg. phosphorus/15 /A. differing from one another by
0.02 or 0.04 ixg.

Molybdic-Sulfuric Acid Reagent. 5% ammonium molybdate con-
taining 15% by vol. cone, sulfuric acid.

Ajninonaphtholsulfonic Acid Solution. Dissolve 0.5 g. of the 1,2,4
acid, 30 g. sodium bisulfite, and 6 g. crystalline sodium sulfite in
water by shaking, and make up to 250 ml. Filter and, if filtrate
is not clear, leave overnight and again filter. Prepare fresh
every 2 weeks.

Veronal Buffer, pH 9.0, with magnesium. 0.0015 M magnesium
chloride in buffer consisting of 9.36 ml. 0.1 M sodium diethyl
barbiturate + 0.64 ml. 0.1 N hydrochloric acid.

Substrate Solution. 0.1 M sodium-/3-glycerophosphate.

10% Trichloroacetic Acid.

210 CAPILLARY TUBE COLORIMETRY

PROCEDURE

1. Pipette 3 fA. plasma into 21 /xl. water in a reaction tube of
250 /xl. capacity, and add 7 /xl. Veronal buffer with magnesium and
7 /xl. substrate soln. Mix with a magnetic stirring "flea" (page 179).

2. Set up a control experiment in which the substrate and buffer
are placed as a separate drop on the side of the tube where it
cannot touch the enzyme soln.

3. Place tubes in a rack in a desiccator containing water in the
bottom and a small bottle of chloroform to produce a vapor inhibit-
ing growth of microorganisms. The desiccator is kept at 37° and
the digestion is allowed to proceed for 4 hr.

4. Stop the reaction by setting the tubes in ice water, and add
10 /xl. 10% trichloroacetic acid to each.

5. Centrifuge, and pipette 15 /xl. of the supernatant into another
tube.

6. Add 7 /xl. of the molybdic-sulfuric acid reagent to the super-
natant.

7. Pipette 5 /xl. aminonaphtholsulfonic acid soln. on the side of
the tube as a separate drop.

8. Set up standards with 15 /xl. of the known phosphate solns.,
following steps 5-7.

9. Mix the drops on the side of the tubes with the rest of the
liquid using stirring "fleas."

10. Draw the solns. into capillary tubes of uniform lumen (inside
diameter 0.65 mm., length 30 mm.) and seal the ends with Duco
cement.

11. Compare the colors as in the Walker method.

REDUCING SUBSTANCES

Sumner's ( 1925) dinitrosalicylic acid method was adapted to
capillary tube colorimetry by Walker and Reisinger ( 1933) . In
this manner, quantities of glucose of the order of 0.1 /xg. in 0.2 /xl.
liquid (50 milligram per cent) can be determined with a maximum
error of 3 milligram per cent in duplicate measurements of solutions
of known concentration. For other methods see page 296.

REDUCING SUBSTANCES AND CREATININE 211

Walker and Reisinger Method for Reducing Substances

SPECIAL REAGENTS

Standard Glucose Solutions. Prepare solns. in the range 10-100
mg./lOO ml. in 5 mg. steps.

Sumner Reagent. Add 22 ml. of 10% sodium hydroxide to 10 g.
crystallized phenol. Dissolve in a little water and dilute to 100 ml.
Add 69 ml. of this soln. to 6.9 g. sodium bisulfite; then add a
soln. containing 300 ml. 4.5% sodium hydroxide, 255 g. Rochelle
salt (KNaC4H40r,.4H20) and 880 ml. 1% dinitrosalicylic acid.
Store in well-stoppered bottles and prepare fresh each week.

PROCEDURE

1. Introduce a 1.5-3.0 mm. column of tungstic acid supernatant
(page 198) or other unknown soln. into a capillary tube (0.35 mm.
inner diameter) followed by a second column (three times as long)
of the reagent. Seal both ends of the tube and mix by centrifuging.

2. Mix the standard solns. with reagent in test tubes employing
1 ml. glucose to 3 ml. reagent.

3. Immerse the capillary tubes and the test tubes together in
boiling water for 5 min.

4. Transfer the standard color solns. to capillary tubes.

5. Compare the colors on a white background under light
screened with Daylite glass.

CREATININE

Bordley, Hendrix, and Richards (1933) adapted Folin's method
for the determination of creatinine to capillary tube colorimetry.
As finally worked out, this adaptation enables analysis of about 0.5
fj}. of liquid containing 10-30 m/y.g. creatinine with an error of a few
per cent. For other methods see page 239.

Method of Bordley et al. for Creatinine

SPECIAL REAGENTS

Standard Creatinine Solutions. Prepare solns. containing 2.0, 2.5,
3.0, 3.5, 4.0, 4.5, 5.0, and 6.0 milligram per cent creatinine in 0.01
A^ hydrochloric acid. Add toluene as a preservative.

2 12 CAPILLARY TUBE COLORlMETftY

Saturated Picric Acid Solution. Prepare from pure picric acid.
10% Sodium Hydroxide. Prepare from Merck's reagent "from

sodium."
Folin Reagent. Freshly prepare before use by mixing 5 vol.

saturated picric acid with 1 vol. of the 10% sodium hydroxide.

PROCEDURE

1. Introduce separate columns of saturated picric acid (25 mi-
crometer divisions), tungstic acid supernatant (page 198) or other
unknown soln. (60 divisions), and 10% sodium hydroxide (5 divi-
sions) in that order into a capillary tube (0.35 mm. inside diameter) .

2. Seal off the ends of the tube and place in a closed box until
the other tubes are prepared.

3. Darken the room for the following operations.

4. Prepare as rapidly as possible the standard color solns. in
test tubes by adding 1 ml. Folin reagent to 2 ml. standard soln.

5. With no loss of time mix the liquids in the capillary tubes
by repeated centrifugations and plan to begin color comparisons
10 min. after the first centrifugation.

6. In this 10 min. interval transfer the contents of each capillary
tube and a portion of each standard mixture to pieces of capillary
tubing of uniform bore, and seal the ends of each piece with
plasticine.

7. Place each sealed piece of tubing in a labeled space on a
milk glass plate for color comparison.

8. Compare the colors in a dark room under a 200 watt bulb
equipped with a straw-colored light filter, or illuminate the milk
glass plate from underneath using a straw-colored filter between the
plate and the light source.

note: The particular order of procedure given must be followed since
Folin reagent darkens at a faster rate and more extensively in capillary
tubes than it does in larger volumes in test tubes. Furthermore, this change
is mtensified and accelerated by daylight, which makes it imperative to
protect the solutions from light. The yellow picric acid color interferes
with the comparisons of the red colors developed, and hence it is necessary
to use a straw-colored light filter. The color produced by 2.0 milligram
per cent creatinine is about the palest which can be reliably estimated in
the tubes used. The most advantageous colors are those produced in the
range 2.5 to 5.0 milligram per cent.

XJEIC ACID 213

URIC ACID

Bordley and Richards ( 1933) adapted Folin's ( 1930) method for
the determination of uric acid to capillary tube colorimetry with
the result that 0.03-0.5 fil. liquid containing 3-10 m/xg. uric acid can
be determined in solutions of known concentration with an average
error of about 5%. For other methods see page 239.

Bordley and Richards Method for Uric Acid

SPECIAL REAGENTS

Standard Uric Acid Solutions. Prepare stock soln. containing 1
mg./ml. Transfer 1 g. uric acid to a 1 1. volumetric flask. Shake
0.6 g. lithium carbonate in 150 cc. water for 5 min. to dissolve,
and filter. Heat the soln. to 60°, pour into the liter flask with
the uric acid, and shake for 5 min. Cool under running cold tap
water. Add 20 ml. 40% formalin and half fill the flask with
distilled water. Add a few drops of methyl orange soln. followed
by 25 ml. 1 N sulfuric acid added slowly and with shaking. The
soln. should turn pink before the last 2-3 ml. acid is added.
Dilute to vol., mix well, and store in a tightly stoppered bottle
protected from light. Prepare a series of standards containing 0.6,
0.8, 1.0, 1.2, 1.4, 1.6, and 2.0 mg./lOO ml.

Cyanide-Urea Solution. Dissolve 50 g. sodium cyanide in 700 ml.
water, add 300 g. pure urea and, when dissolved, transfer into a
2 1. flask. Add 5-6 g. calcium oxide and shake for 4-5 min. Filter
through Whatman No. 41 or similar paper. Add up to 1 g. finely
divided disodium phosphate; shake and filter. The soln. may be
used for at least 2 months if stored at room temperature and much
longer if kept cold.

Uric Acid Reagent. Dissolve 100 g. sodium tungstate in 200 ml.
water. Add slowly with stirring and cooling 20 ml. 85% phosphoric
acid. Pass a slow stream of hydrogen sulfide through the soln.
for 20 min. but after the first 3-4 min. add 10 ml. more of the
85% phosphoric acid. Filter through Whatman No. 41 or similar
paper, refiltering the first 40 ml. Transfer the filtrate to a separa-
tory funnel and shake for a few min. with 300 ml. alcohol. Transfer
the lower layer into a previously weighted 500 ml. flask and add
water to a total of 300 g. liquid. Boil a few min. to remove the

214 CAPILLARY TUBE COLORIMETRY

hydrogen sulfide. Add 20 ml. 85% phosphoric acid and slowly
boil for 1 hr. under a reflux condenser. Decolorize with a few drops
of bromine, boil off the excess bromine, and cool. To 12 g. lithium
carbonate and 25 ml. phosphoric acid, add slowly 150 ml. water.
Boil off the carbon dioxide and when solution is complete, cool
and mix with the cone, uric acid reagent and dilute to 1 1. Keep
in well-stoppered bottles protected from light.

PROCEDURE

1. Introduce into a uniform capillary tube (0.35 mm. inner
diameter) a 5 mm. column of tungstic acid supernatant (page 198)
or other soln. to be analyzed, 5 mm. cyanide soln., and 1 mm. uric
acid reagent, keeping the three columns separated by air spaces. Seal
off both ends of the portion of the tube containing the liquids.

2. In a similar manner prepare tubes with the seven standard
solns.

3. Mix the liquids simultaneously in all of the tubes by centri-
fugation.

4. Four min. after the mixing immerse the tubes for 1 min. in
boiling water.

5. Compare the colors.

UREA

Walker and Hudson ( 1937) adapted the capillary tube apparatus
to the determination of urea by the hypobromite method. This
adaptation enables the analysis of 0.3 /xl. liquid containing 2 to 25
milligram per cent urea nitrogen with an average deviation from the
macro method of ±3.8%. The measurement of known quantities of
urea added to dialyzed horse serum could be made with an average
error of 2.3%. For other methods see page 286.

Walker and Hudson Method for Urea

SPECIAL REAGENTS

Sodium Hypobromite Solution. Prepare according to Stehle
( 1921) : In a 50 ml. Erlenmeyer flask, mix 2 ml. of a soln. contain-
ing 12.5 g. sodium bromide and 12.5 g. bromine/100 ml., with 2 ml.
of a soln. containing 28 g. sodium hydroxide/100 ml. Gently

UREA 215

revolve the mixture in -the flask for 1 min. and set aside for 30
min. before use.

PROCEDURE

1. Fix a 15 cm. length of uniform capillary tubing (0.35 mm.
inner diameter) to the stage of the binocular microscope so that one
end is in the optical field and attach the water manipulator to the
other end.

2. Introduce a 6 mm. column of distilled water with a capillary-
pipette and draw it back 1 mm. from the end.

3. Introduce a 3 mm. column of tungstic acid supernatant (page
198) or other urea soln. to be analyzed, draw it in away from the
end, and seal the end with plasticine.

4. Accurately measure the length of the air column between the
two liquid columns with a filar micrometer temporarily substituted
for the right ocular which contains a disc micrometer. Two successive
readings must agree within 1 micrometer scale division (5 fx).

5. Replace the right ocular, cut off the end of the tube sealed
with plasticine, move the urea column back to the end of the capil-
lary with the water manipulator, and add 3 mm. of sodium hypo-
bromite soln. from a blunt-tipped pipette freshly filled just before
use. Should gas bubbles appear immediately upon the addition of
the reagent, discard the tube, and prepare fresh reagent.

6. Move the liquid column away from the end of the tube and
seal with plasticine.

7. Carefully remove the tube from the water manipulator by
cutting it about 6 cm. from its end, revolve it between thumb and
forefinger for a few sec. and set aside in a nearly vertical position
upon a plasticine mount.

8. After 2 hr. again revolve the tube for a few sec, place on the
microscope stage and accurately measure the length of the air column
with the filar micrometer.

9. Run a blank determination with distilled water and subtract
the increase in the length of the air column from that found above.
For each milligram per cent of urea nitrogen the increase averages
four scale divisions (20 /a) ; the increase in the blank averages five
divisions. Hence a soln. containing 10 milligram per cent urea
nitrogen should give an increase of 45 divisions.

216 CUVETTE COLORIMETRY

HYDROGEN ION CONCENTRATION

Capillary tube colorimetry has been employed for the measure-
ment of the hydrogen ion concentration of less than 1 /xl. liquid by
Montgomery (1935), who used quartz capillary tubes having an
internal diameter of 4-5 mm. When comparisons of indicator colors
given with protein-free buffer solutions were made, the error of
measurement was less than 0.02 pH. However when applied to
biological fluids, the indicator color may not be an accurate indica-
tion of the pH value. Montgomery (1935) observed that the capil-
lary tube method gave values for blood plasma from frogs and
Necturus which were consistently lower by an average of 0.11 pH
than those obtained with a glass electrode, a deviation which he
ascribed to the protein error of the indicator. It may be possible in
some cases of this nature to apply a correction factor. For electro-
metric measurements see page 183.

B. CUVETTE TECHNIQUE

1. Apparatus

General. Cuvettes for the colorimetric measurement of small
volumes of liquid have been designed for use with certain standard
colorimeters. Zeiss cuvettes having a capacity of 0.2 ml. are made
for the Pulfrich step photometer. The Evelyn photoelectric colorim-
eter has a micro attachment made to accommodate cells which re-
quire 0.15 ml. Adapters which enable 0.2 ml. cuvettes to be used with
the Coleman Junior spectrophotometer (model 6) are obtainable
from &. Ash (Lowry, Lopez, and Bessey, 1945). Quartz cuvettes
permitting the use of volumes of 0.05 ml., or less, with a special
adapter for the Beckman quartz spectrophotometer have been de-
scribed by Lowry and Bessey ( 1946) . With their adaptation meas-
urements can be carried out on 0.05 ml. volumes from about 225 to
1050 m^ with spectral widths of no more than 3 m/x. With 0.025 ml.
volumes a range of 235-935 m/x can be utilized with the 3 m^u,
spectral bands. The cuvettes and adapters may be obtained from
Pyrocell Manufacturing Co.*

* Since this writing a capillary absorption cell has been described by Kirk
et al. (see Bibliography Appendix, Ref. 40; see also Ref. 42).

APPARATUS

217

The degree of light absorption, and consequently the response in
the eye or in a photocell, is proportional to both the concentration
of the color substance and the length of the light path through the
solution. Therefore, a given quantity of color substance will effect
the same light absorption whether it is contained in a volume of
0.002 ml. and a 0.35 mm. light path is used (as in Richards' tech-
nique, page 195), or in a 0.060 ml. volume with a 10.5 mm. path.
Of course, the greatest absorption would be obtained by employing
the smallest volume with the longest light path.

i

Top %

I

t^ Light

^ Penny"

Light
Cuvette

^

^

Cuvette

; © ;

s y

Block

Diaphragm
(side)

Diaphragm (Type A)
(face)

°o*

— Holes for precision pins — '
of Beckman

o

■f

0o O ))

O

Adj

/
ustable

Diaphragm (Type B)

Fig. 77. Microcuvette and diaphragms.
Frovi Lowry and Bessey (1946)

Lowry and Bessey Adaptation of Beckman Spectropho-
tometer to Measurements on Small Volumes. The special cu-
vettes used have the same 1 cm. light path as the macro variety, but
the width of the chamber has been reduced to 2 mm. or less (Fig.
77) . A 0.05 ml. volume of liquid will fill the cuvette to a height of
about 2.5 mm. The height of the cell is 25 mm. and its outside cross-
sectional dimensions are the same as those of the macro vessel. The
inner cross-sectional dimensions of the macrocuvette are 10 X 10
mm. Cuvettes having an internal measurement of 1 X 10 mm. have
also been used; they require 0.03 ml. liquid, but their use is more
difficult.

218 CUVETTE COLORIMETRY

A diajihragm is placed in front of the cuvette to obtain a light
beam confined to a cross section of less than 2X2 mm. A beam of
this size can pass through the liquid without touching the meniscus
or the walls of the cuvette. The diaphragm (type A, Fig. 77) has
a metal disc the size of a penny through which a 1.0 to 1.4 mm. hole
is drilled about 1 mm. off center. Before the disc is fastened to the
metal sheet, it is held in the oi)ening from which the light enters
the cuvette and turned until the beam passes precisely in the middle
between the walls of the chamber when the cuvette is in place. The
disc is soldered at this angle to the sheet of metal (about 6X9
cm.) so that the hole coincides with a 3-4 mm. hole in the sheet 2.5
cm. from one end. The top of the sheet is bent at a right angle to
form a flange which lies on the top of the instrument. Wooden
blocks are used to raise the cuvettes so that the light beam just
misses the bottom of the chamber. The diaphragm is inserted and
removed by loosening the bolts which hold the phototube housing.
The carriage for the cuvettes should be oriented to bring the cuvettes
as near the diaphragm as jwssible. The cuvettes are numbered and
always set in the holder with the same orientation.

The type B diaphragm (Fig. 77) contains a sliding strip of brass
with pinholes which can move in a channel cut in the sheet metal.
The diaphragm is inserted between the cuvette carriage and the
body of the instrument and the sliding strip is moved until a pinhole
coincides with the center of the cuvette. The stop on the strip is
then adjusted with a bolt so that the pinhole can be brought to the
same position each time. The different-sized pinholes can be brought
into position without disturbing the adjustment. Blocks are used
to raise the cells as with the type A diaphragm. By removing the
brass strip the instrument can be used with macrocuvettes without
disturbing the metal sheet.

To obviate the effect of "play" in the cuvette carriage, the cells
should be moved into position from the same direction. In use, the
microcuvettes are left mounted in the carriage. Samples are intro-
duced with fine-tipped pipettes, and removed by suction with fine
tipped glass tubes. A macro cell may be used in the first position in
the carriage for the solvent or other blank solution.*

*See Bibliography Appendix, Ref. 32.

CALCIUM 219

2. Methods

CALCIUM

Sendroy (1942b) adapted iodometric reactions, previously used
in titrimetric measurements of calcium, to colorimetry. Using an
Evelyn macro photoelectric colorimeter, the method was applied
to volumes of serum down to 20 /xl. (about 2 fig. calcium). Further
refinement could be obtained if the colorimetry were carried out with
smaller volumes in microcuvettes. The calcium is precipitated as
the oxalate; the latter is washed, dried, dissolved in acid, and reacted
with an excess of eerie sulfate. The excess of eerie ion is made to
liberate iodine from potassium iodide and the yellow color thus
developed is measured. A precision of ±2^0 has been reported. An
alternative method is measurement of the blue color formed when
starch is added to the iodine solution. However, the many factors
which influence the blue color make it a less desirable choice even
though the relative color intensity of the blue is about 100 times that
of the yellow (Sendroy and Alving, 1942). The method to be de-
scribed was developed for serum, but it can be adapted to other
fluids. A thorough study of different procedures for the determination
of serum calcium was made by Sendroy (1944). For titrimetric
methods see page 272.

Sendroy Method for Calcium

SPECIAL REAGENTS

Saturated Ammonium Oxalate (about 3.5%). Prepare at room
temperature using analytical reagent grade of the salt.

2% Ammonium Hydroxide. Dilute 2 ml. cone, ammonium hy-
droxide (26% analytical reagent grade) to 100 ml.

Water-Alcohol-Ether Mixture. Mix equal vol. distilled water,
absolute ethyl alcohol, or redistilled 95% alcohol, and ethyl ether
(analytical reagent grade, or absolute, or redistilled U.S.P. grade).

1 N Sulfuric Acid (approx.). Dilute 27 ml. cone, acid, sp. gr. 1.84,
analytical reagent grade, to 1 1.

0.2 N and 0.1 N Sidfuric Acid (approx.). Prepare from the 1 N
soln.

220 CUVETTE COLORIMETRY

0.1 N Ceric Bisulfate (approx.). Dissolve 29 g. anhydrous eerie
bisulfate in 1 A^ sulfuric acid to make 500 ml. (The greater ease
of solution of the bisulfate makes it preferable to the sulfate. The
bisulfate is obtainable in about 92% purity from G. Frederick
Smith Chemical Co.) Store in amber bottles and protect from
light.

In preference to ceric sulfate, Kochakian and Fox (1944) have stressed the
greater sharpness of the end point in titration with ammonium hexanitrato-
cerate using setopaHne C as the indicator. Prepare a 0.01 A^ soln. by dissolving
6 g. ammonium hexanitratocerate, reagent grade, in about 200 ml. 1 N per-
chloric acid and dilute to 1 I. with the acid. Do not heat during preparation,
and store in a black bottle in the dark. Prepare a 0.05% setopaline C soln. by
adding 50 mg. of the dye (Eimcr and Amend) to 100 ml. distilled water and
warming on a hot plate or bath. Precipitation occurs on cooling, therefore the
soln. must be warmed before use and used while warm.

0.0035 N, 0.001 N, 0.0007 N, and 0.00035 N Ceric Bisulfate
(approx.). Prepare these solns. just when needed from the 0.1
N soln. Use 0.2 A^ sulfuric acid to dilute the 0.1 A'' to 0.0035 A^.
Use 0.1 N sulfuric acid for dilution to the weaker concentrations.
Prepare the 0.007 A^ and 0.00035 A^ solns. from the 0.0035 A^ soln.
Store in amber bottles and protect from light.

Standard 0.1 N Sodium Oxalate. Dissolve 3.3498 g. sodium oxalate,
analytical reagent grade, in 52 ml. 1 A^ sulfuric acid and add water
to make 500 ml. Store in amber bottle; the soln. is stable for at
least 6 months.

Standard 0.0005 N, 0.00025 N, and 0.0002 N Sodium Oxalate. Pre-
pare fresh when needed from the 0.1 A'' soln. by dilution with
water.

0.5% and 1% Potassium Iodide (approx.) Prepare fresh for use
from analytical reagent grade of the salt. (When tested with
starch, no trace of free iodine should be present.)

95% Ethyl Alcohol. Filter through two layers of ashless filter paper
on a Buchner funnel.

2% and 1% Starch (Lintner Soluble). Prepare the 2% soln. in
saturated sodium chloride soln. every 2 weeks by making a paste,
diluting to vol. and boiling for 5-10 min. Prepare the 1% soln.
fresh for use from the 2% soln. by diluting with water.

CALCIUM 221

PROCEDURE

1. Mix 5 vol. distilled water with 1 vol. serum in a 12-15 ml.
centrifuge tube; run in duplicate. Add 6 vol. distilled water to
another centrifuge tube to be treated in a parallel manner as a
standard; run in duplicate.

2. Add 1 vol. saturated ammonium oxalate to each tube, stir by
tapping, cover tubes to keep out dust, and let stand at least 16 hr.

3. Centrifuge for 5 min. at 2600 R.P.M. and carefully siphon
off all but about 0.2 ml. of the supernatant with an upturned capil-
lary, the tip of which is kept immersed.

4. Wash the entire inner surface of each tube with 3 ml. 2%
ammonia added slowly from a pipette moved around the top of the
tube.

5. Tap the tubes until the precipitates just begin to move up;
then again centrifuge and withdraw the supernatant.

6. Add 1 ml. of the water-alcohol-ether mixture; stir and mix
well. Add 3 ml. more of the mixture and mix gently to keep a mini-
mum of the precipitate in the upper portion of the liquid. Centrifuge
and withdraw the supernatant.

7. Repeat the washing with the water-alcohol-ether mixture as
in step 5.

8. Place the tubes in an oven at 100-110° at an angle of about
15° for 0.5-1.0 hr. to dry completely.

9. To each of the two standard tubes add different vol. of the
standard oxalate soln. (See Table VI.) Add the sulfuric acid to
all the tubes (see table) and heat for 5 min. in a beaker of water
kept below boiling.

10. Remove tubes, let cool to room temperature and add the eerie
bisulfate (Table VI). Mix well, cover the tubes, and let stand at
room temperature for 30 min. or in a water bath at 70° for 10 min.

11. Transfer solns. to vessels in which color development and
dilution to final vol. (Table VI) are carried out. To facilitate trans-
fer, coat a part of the outer rim of the tube with a thin film of
paraffin. Wash out tubes with 4 ml. portions of water to make the
transfer quantitative.

12. Add potassium iodide (Table VI) with a minimum of
agitation necessary to mix well. After 60 sec. add filtered alcohol
and then water to bring to final vol.

222

33

to

s

>

(N

O

o

a
w

01

O .

o S

w a

3

S
O

CO

3
T3

«
o

c

s
O

■ a

o

O "5

o o

Q O

I

00

o a

ffi

«

o o o

(M (N C<l

05

o o o

00 GO 00

o >ra
o o

Q O

I

00

o o o

o c^

o o

Q o

I

CO
00
1/5

o o o

o
o

o
o

CD

o o o

'^ tJ^ tJ<

5> (M C^ (N

^

oO •

°* g
o

^ o o

O "^ ^

•s-9 «■

(M

O O iC
(M C<i (N

o feO .

^^ 03 CI —

o c 2; °

o

o
o

d
in

3
t>

=0 "H

— • 33

s C

^ 33

e

33

c

03

CO CO CO

odd

LO IC

d d

o o

o o o

■^ "^ "^

CO CO CO

<6 'S <6

CO
o
o
o

d

o o

^

o

CO
83

lO lO

d d

lO lO IC

d d d

b- o o

o

O ^H 1— I

o
d

lO

(N

o
o

o

3 ^

I §

• ^ ^

!2 S 33

o ^ n

d c» Ph

<

d

o o o
-- — (N

! ^

o o

g o

8 °

3 ^

^ - ^

CO ^ ^

_: 33 3
■3-^4)

S 33

d CK Ph

d
o"

CO

O !»

.i a.

^ 3

03 -

ci ^~^

Co

"-o

00-*

?§s

o 2

^ a

00-3

(M 2

CO 2

^ a

o^

OJO

MM

3^

2S

«^o

■^co

£ =»

■3 <o

•- 3

^t

■*^ ,s

"l^-s

<B.S

.■t^.3

a-^

tc :s

c

?^:£

-S u)

• <-i

u ^^

Oi .^

-iJ .^

a a

tc

CO o3

03 ^

tiD(-5

a>

T30

CD CD

03

.3 .

!0

>

a-T3

03

c

.^-^

-^- 3

3 to

.— .

^ '^

m. th
0. 40
ivelv

f^oo

aZ"£

»3 IC

3 .

.n £ a

•-

^^ S CO

co;§2J

fi^

.^•7

ooy2 >

0) s

^^ a

> *s

No.
(icon
0)m

wa

s 3

33'"

"^.3

i: 3cD

3 &

-Ui /vJ

let Ul
lynd
imum

3 S

3

II

ji X
L- > 33

>H a

bC ^

r/3 "•

Cornin
to 635

-1

a;cD

iJ «

^COqO

CQii

H 3"^

r^

' i-g

A c

CALCIUM 223

13. Prepare simultaneously reagent blanks containing sulfuric
acid, potassium iodide, and alcohol in the same concentrations as in
the standards and unknowns. Use these blanks to set galvanometer
at 100 just before reading the standards and unknowns.

14. Read 10 ml. portions at 25 ±5° in the Evelyn colorimeter
with filters indicated in the table. The yellow color may be read at
any time within 1 hr. after addition of the potassium iodide. Data
are given in the table for use of blue color with starch if filters for
yellow color are unavailable. In the latter case, add water to about
80% of vol. after the potassium iodide has been added, add the
starch slowly to the soln. at 25 ±1°, mixing by rotation, add water
to vol., and read color promptly at 25 ±1°. The blue-color method
is not reliable for samples of serum smaller than 0.05 ml.

Result. Since blue color readings are not reproducible from day
to day, a new calibration curve must be obtained for each day's
work. Galvanometer readings plotted semilogarithmically are linear
functions of oxalate concentration. For calibration of yellow-color
readings, the percentage transmission readings of 90, 80, 70, 60, 50,
40, 30, 25, 20, 15, and 10 correspond to the respective values of
iodine, in milliequivalents per liter in the color solns., of 0.0022,
0.0048, 0.0077, 0.0111, 0.0151, 0.0200, 0.0267, 0.0308, 0.0362, 0.0433,
and 0.0544. To redetermine the calibration curve, treat 0.5 to 0.7 ml.
of 0.133 mM potassium iodate with 2.4 ml. of 0.085 M phosphoric
acid and 1.2 ml. of 5% potassium iodide, dilute to 40 ml. with
filtered 95% alcohol, and then to 100 ml. with water. Read with
filter No. 586-5.

Calculate the concentration of the calcium (in milliequivalents
per liter) in the unknown from:

j-C-C^s^, _ ^j^^^-j^

where Si, So, and u refer to standards and unknown sample, C =
(C204^")si + {C204^~)s2, and D = V/v where V is the volume (in
ml.) at final dilution of color soln. and v is the volume (in ml.) of
original sample used. The C and D values are given in the table and
the (lo) values are obtained from calibration curve data correspond-
ing to the readings observed.

224 CUVETTE COLORIMETRY

When 0.02 ml. samples of serum are used it has been found neces-
sary to correct the yellow color readings for the effects of traces of
residual serum. The correction varies with the reading and is to be
subtracted from the latter: for readings of 15, 20, 30, 40, 50, and 60
subtract corrections 0^-^, 0-, 0^, l'^, V, and P, respectively.

example: Analysis of 0.02 ml. samples of a serum gave yellow color readings
of 36\ 36^ for the standards, 23^ and 55". The former were corrected to 35^ by
subtracting 1°. Values for iodine in milliequivalents per liter from the calibra-
tion curve were 0.0230 for the serum, and 0.0325 and 0.0130 for the standards.
Then the calcium concentration in the original serum (in milliequivalents
per liter) was:

[( °°^°° + 0.0325 + 0.0130 ^ _ „^23g-| ^ ^ ^ ^^^

In the case of blue-color readings a semilogarithmic plot is made with
milliequivalents of oxalate per liter from to 0.0171 as abscissa and galvanom-
eter readings from 10 to 100 ordinates. A straight line is drawn between
the points representing the readings of the two standards, Si and S2, at concen-
trations of 0.016 and 0.004 milliequivalent oxalate per liter. Oxalate values for
serum analyses, obtained by interpolation of their galvanometer readings on
this line, times D give directly milliequivalents calcium per liter in the sample.

example: Analysis of 0.05 ml. samples of a serum gave blue color readings of
32^ and 32*, for the standards, 2V and 49^. A straight line was drawn through
the two latter values located at 0.016 and 0.004 milliequivalent oxalate per liter,
respectively. Interpolated values for the serum were 0.01005 and 0.01013 milli-
equivalents per liter. Then the average calcium concentration in the original
serum was:

0.01008 X 500 = 5.04 milliequivalents per liter.

CHLORIDE

Colorimetric chloride methods have not been specifically adapted
to histochemical work. However, the procedure of Sendroy (1939b,
1942a), which was designed for use with the macro Evelyn photo-
electric colorimeter, could be adapted to the smaller quantities
sufficient for use with microcuvettes. Even with the macro apparatus,
10 lA. serum is adequate for analysis. The principle of the Sendroy
method is conversion of the chloride in acid solution to its silver
salt by shaking with solid silver iodate; the iodate liberated by the
chloride is made to act on potassium iodide, and the yellow color of
the iodine set free is measured using filter No. 420 with the Evelyn
instrument. For other methods see pages 200 and 281.

CHLORIDE 225

Sendroy Method for Chloride

SPECIAL REAGENTS

Approxwiately 0.085 M Phosphoric Acid. Test to make sure soln.
is halide free.

Caprylic Alcohol.

Silver lodate Powder, C. P. Test for presence of potassium iodate
according to Sendroy (1939a): (1) Solubility measurement —
lodate analysis of a saturated soln. of silver iodate should give a
value not exceeding 0.21 mM/1. (2) Analysis of a standard
chloride soln. — Analyze a known 100 mM chloride soln. diluted
twenty times with 0.085 M phosphoric acid.

5% Potassium Iodide. Test with starch to be sure no trace of free
iodine is present.

PROCEDURE

1. Dilute the sample chloride soln. in 0.085 M phosphoric acid,
or in tungstic acid soln. if protein is present, at between pH 2.0 and
3.0 to a final concentration of between 3 and 12 mM/\.

2. Add solid silver iodate (10 mg./ml.) to duplicate portions in
15 ml. centrifuge tubes, and shake vigorously for 2 min.; then either
filter through halide-free paper or centrifuge for 1 min. at over
3000 R.P.M. The soluble iodate is now equivalent to the chloride in
the sample.

3. Sendroy (1942a) recommended that the system given in
Table I, B, of the paper by Sendroy and Alving (1942) be followed
for the measurement of chloride in serum and blood. In this schedule
0.5-13 ml. iodate soln. containing 0.8 mM is added to 2.4 ml. 0.085 M
phosphoric acid and 1.2 ml. 5% potassium iodide; the vol. is made
up to 100 ml. with water.

4. Promptly after the color has been developed, transfer the
tubes to a 25° water bath for 3-5 min. Wipe the tubes clean and
dry; set the Evelyn instrument with filter No. 420 to read 100 with a
blank soln., and then read the unknowns. The blank soln. may be
prepared by omitting the iodate from the reaction mixture.

5. A calibration curve has been given by Sendroy and Alving
(1942), but it is usually well to obtain one's own curve with the
particular reagents and instrument used. The calibration may be
made with standard potassium iodate solns.

226 CUVETTE COLORIMETRY

PHOSPHATE AND PHOSPHATASE

Siwe (1935) has described the use of the Pulfrich step photometer
with filter No. 72 (red) for the colorimetric measurement of in-
organic phosphorus in small amounts of blood by conversion to
phosphomolybdic acid and reduction by aminonaphtholsulfonic acid.
The same reaction was employed by Weil and Russell (1940) in
their phosphatase method ( page 209) . At about the same time Lund-
steen and Vermehren (1936) developed a micro procedure for the
determination of inorganic phosphate and alkaline phosphatase in
blood plasma based on Miiller's (1935) amidol reduction of phos-
phomolydic acid. The Pulfrich instrument with filter No. 72 was also
used in this case, and for most measurements the 10 mm. cells were
employed, but the 20 mm. cells were required for weaker colors. 50
fA. of blood are needed for a duplicate determination, and since the
procedure might be adapted to tissue extracts as well, it will be
described.

Conditions for the determination of inorganic phosphate in the
presence of labile phosphate esters, such as phosphocreatine, acetyl
phosphate, and ribose-1-phosphate, were established by Lowry and
Lopez (1946). As these authors pointed out, the usual procedures
for measurement of inorganic phosphate in tissue extracts represent
the sum of the inorganic phosphate and the phosphate of the labile
esters hydrolyzed by the reagents employed in the determination.
The procedure of Lowry and Lopez is based on the reduction of
phosphomolybdate by ascorbic acid at pH 4.0.

Bessey, Lowry, and Brock (1946) utilized as the substrate p-nitro-
phenyl phosphate, which had been studied by King and Delory
( 1939) , and applied to phosphatase determinations by Ohmori
(1937) and Fujita (1939). Bessey et al. were able to determine the
phosphatase in as little as 5 ju.1. serum using 0.5 ml. of solution for
the colorimetry. The advantage of this substrate is that it is color-
less and yields the yellow salt of p-nitrophenol when the phosphate
group is split off. Thus the color develops in proportion to the degree
of the hydrolysis and no additional reagents are required for the
color development. This advantage is also to be found in the use
of phenolphthalein phosphate, which was employed by Huggins and
Talalay (1945). However, alkaline phosphatase splits the p-nitro-
phenyl phosphate 25-30 times faster than the phenolphthalein com-

PHOSPHATE AND PHOSPHATASE 227

pound, 15% faster than phenyl phosphate, and 2-3 times more
rapidly than glycerophosphate, according to Bessey, Lowry, and
Brock. Either acid or alkaline phosphatase may be determined with
the substrate; it is only necessary to carry out the colorimetry in
alkaline solution, since the free nitrophenol, which would exist in
acid solution, is colorless. For other methods see pages 124, 208, 209,
and 280.

Lundsteen and Verniehren Method for Inorganic Phosphate
and Phosphatase

SPECIAL REAGENTS

Substrate. Combine 8 ml. 1 A^ ammonium hydroxide, 12 ml. 1 N
ammonium chloride, 1 g. disodium-;S-glycerophosphate, 2 ml. 1 M
magnesium chloride, and make up to 100 ml. with water.

10% Trichloroacetic Acid.

Acid-Molybdate Solution. Combine 100 ml., 7.5% ammonium
molybdate, 45 ml. 10 A^ sulfuric acid, and 105 ml. water.

Amidol Solution. Dissolve 15 g. sodium sulfite and 1.5 g. Amidol
(Agfa) in 100 ml, water. Store in the dark and cold, and dilute
five times before use. After about 2 weeks it turns red and can no
longer be used.

PROCEDURE

1. If blood is used, pipette 50 ^1. into 1 ml. of 0.9% sodium
chloride soln. and centrifuge out the cells.

2. To 200 ix\. of the supernatant fluid or a tissue extract add 200
fj\. substrate soln. and place in thermostat for the digestion period
(for plasma 24 hr. at 37°).

3. To another tube containing the same ingredients add 300 /*!.
10% trichloroacetic acid for a control experiment.

4. Stop the reaction by adding 300 fA. 10% trichloroacetic acid,
and centrifuge out the precipitate from both the enzyme and control
tubes.

5. To 400 /xl. of the supernatant in each case ^dd 100 /xl. acid-
molybdate reagent and 100 fA. Amidol soln.

6. Measure the color intensity after the soln. has stood for 15
min., and obtain the quantity of free phosphate from a previously
determined calibration curve constructed from measurements with
known amounts of phosphate.

228 CUVETTE COLORIMETRY

Lowry and Lopez Method for Inorganic Phosphate in Presence
of Lahile Phosphate Esters

SPECIAL REAGENTS

Protein Precipitant. 5% trichloroacetic acid, or 3% perchloric
acid, or (with very labile esters) saturated ammonium sulfate
which is 0.1 A^ with respect to acetic acid and 0.025 A^ with respect
to sodium acetate (pH 4).

0.1 N Sodium Acetate.

Acetate Buffer (pH 4) . 0.1 X to acetic acid and 0.025 N to sodium
acetate.

1 % Ascorbic Acid.

1% Ammonium Molybdate in 0.05 N Sulfuric Acid.

0.05 mM Standard Phosphate Solution.

PROCEDURE

1. Deproteinize the sample with ice-cold protein precipitant. If
either of the acid precipitants is used, bring the extract rapidly to
pH 4.0 to 4.2 by adding 4 vol. of 0.1 .V sodium acetate. (Most labile
esters are fairly stable at this pH.)

2. Dilute the extract with the acetate buffer until the inorganic
phosphorus concentration is 0.015 to 0.1 mM (0.05 to 0.3 milligram
per cent) . Dilute ammonium sulfate extracts at least five times.

3. Add 0.1 vol. 1% ascorbic acid and 0.1 vol. 1% acid-molybdate
soln. to each vol. of extract. If used within 15 min. of their mixing,
the ascorbic acid and molybdate may be combined.

4. Carry out the colorimetric reading at 5 and again at 10 min.
after the addition of molybdate using light between 650 and 950
vcijx (maximum absorption at 860 m/x) .

5. Take simultaneous readings of the standard soln. and blank,
which should be prepared parallel with the unknown. Should a
difference be observed in the readings of the unknown at 5 and 10
min. compared to the standard, extrapolate the values to zero time.

note: Lowiy and Lopez have found the reaction to be delayed in the
presence of certain tissue extracts. In these cases standardization must be
obtained by adding a known quantity of inorganic phosphate to an aliquot
and using the difference in the readings effected by the added phosphate
for the standardization. Dilution overcomes the inhibitory effect to some
degree. Thus, brain and muscle extracts should be diluted to a volume
150-200 times that of the tissue, and, in the case of liver 300-500 times.

PHOSPHATE AND PHOSPHATASE 229

Furthermore, acceleration in color development may be accomplished by
increasing the molybdate to 1.5% in 0.05 A'' acid, and the ascorbic acid
concentration to values that do not exceed a final concentration of 0.2%.

Bessey, Lowry, and Brock Method for Phosphatase

SPECIAL REAGENTS

Buffer-Substrate Solution. (pH 10.3 to 10.4). Prepare soln. A by
dissolving 7.50 g. (0.1 mole) glycine and 95 mg. (0.001 mole)
magnesium chloride in 700 to 800 ml. water; then add 85 ml. 1 N
sodium hydroxide and dilute to 1 1.

Prepare soln. B, which is 0.4% disodium p-nitrophenyl phos-
phate in 0.001 A^ hydrochloric acid. (The authors reported that the
Eastman Kodak Co. product contained about 50% inert material;
hence twice the quantity of this preparation should be used.
Purification may be carried out by recrystallization from hot S7%
alcohol.) Adjust the pH of soln. B to 6.5 to 8.0 with acid or base,
if necessary. Test for free nitrophenol by diluting 1 ml. with 10
ml. 0.02 A^ sodium hydroxide and measuring the absorption at
415 m^. If the extinction is greater than 0.08 {i.e., transmission
less than 83% for 1 cm. liquid, or 70% for 2 cm.) remove the
free phenol by extracting two to three times with equal vol. water-
saturated butyl alcohol followed by three extractions with water-
saturated ether. (All butyl alcohol must be removed since it in-
hibits phosphatase activity.) Aerate off the traces of ether and
store in the cold.

Mix equal vol. of solns. A and B; adjust the pH to 10.3 to 10.4
if necessary with strong sodium hydroxide or hydrochloric acid,
and store in the cold, or better, in the frozen state. When 2 ml.
diluted with 10 ml. 0.02 A^ sodium hydroxide has an extinction
greater than 0.1 for 1 cm., either discard or extract it with butyl
alcohol, as above, and readjust the pH.

Standard Solutions. Prepare solns. containing 1, 2, 4, and 6 milf
p-nitrophenol (molecular weight 139.1) per liter.

PROCEDURE

1. Place 5 ^1. serum in the bottom of a 6 X 50 mm. serological
tube, immerse in ice water, and rapidly add 50 jxl. of the ice-cold
buffer-substrate soln. with a constriction pipette. Alix by tapping
with the finger.

230 CUVETTE COLORIMETRY

2. Digest at 38° for 30 min., and then place in ice water and add
0.5 ml. 0.02 A^ sodium hydroxide with sufficient force to mix the solns.

3. Transfer to a cuvette and measure the color intensity using
light at 400-420 m/^.

4. Add 2-4 ix\. cone, hydrochloric acid and take a second colori-
metric reading. The difference in the optical densities gives the
corrected density of the unknown.

5. Run standards and blanks by treating 5 /xl. vol. of the
standards and distilled water in the same manner as the serum.
Construct a standard curve from the corrected optical densities. If
the same pipettes are used for both the standards and unknowns, the
exact pipette vol. need not be known.

Bessey et al. employ a "millimole unit" which is defined as the
phosphatase activity which will liberate 1 milf nitrophenol/liter/hr.
1 millimole unit is approximately equal to 1.8 Bodansky units. For
sera weaker in phosphatase, such as those from adults, the vol. serum
and reagent may be doubled without increasing the vol. alkali; this
will yield a color of nearly double the intensity.

NITROGEN AND AMMONIA

General

As in the more macro methods, the determination of nitrogen in
histochemical investigations involves conversion of the total nitro-
gen to ammonium sulfate by digestion. The digest may be Ness-
lerized directly, or it may be alkalized and the liberated ammonia
absorbed in acid and measured either colorimetrically or titrimetri-
cally. However, the determination of the very small quantities in-
volved requires unique treatment. Since the preliminary procedures
are common to both the colorimetric and titrimetric methods, it
will perhaps contribute to greater integration and clarity if the
chief developments in both forms of analysis are considered together
in chronological order at this point.

At about the same time, Linderstr0m-Lang and Holter ( 1933b)
and Conway and Byrne (1933) reported methods for the micro-
estimation of ammonia which depended on the transfer, by diffu-
sion, of the ammonia from an alkalized solution to one of standard
acid followed by titration. Linderstr0m-Lang and Holter employed a

NITROGEN AND AMMONIA 231

tube having a total capacity of about 250 /xl. for the diffusion process.
The sample was placed in the bottom of the tube and the ammonia
from it was allowed to diffuse into a drop of standard acid placed
in the upper part of the tube to form a seal across the lumen (Fig.
46). Conway and Byrne used a special diffusion cell (Fig. 41) re-
quiring considerably more of the liquids; hence it was not suitable
for the accurate measurement of much less than 14 jug. ammonia
nitrogen. The Linderstr0m-Lang and Holter method had a precision
of 0.005 fxg. nitrogen and it could be used for the determination of up
to about 28 /xg. The following year Gibbs and Kirk ( 1934) employed
a modified Conway-Byrne procedure, which they used for the esti-
mation of from 1.5-8.3 fig. ammonia nitrogen. Conway (1935b)
subsequently described a refinement of the diffusion cell method
which had an ultimate standard deviation of 0.02 /xg. ammonia
nitrogen.

Levy (1936) developed a technique for the determination of total
nitrogen based on the direct Nesslerization of the acid-digested
sample. The complete treatment was carried out in the same vessel
and the final solution was transferred for colorimetry to a micro-
cuvette having a capacity of 0.2 ml. Levy's method was adapted for
quantities of nitrogen in the range 0.5-6.0 /xg. and the average devia-
tion observed was 0.03 jug.

A titrimetric method for total nitrogen employing features of both
the Linderstr0m-Lang and Holter and the Conway techniques was
published by Needham and Boell ( 1939) . These investigators used
a single vessel with a special cap for all the operations, i.e., digestion,
diffusion, and titration. The method was adapted to 1-20 /xg. of
nitrogen and the standard deviation in control experiments was 0.3
fig. In order to refine the earlier colorimetric method of Borsook
(1935), Borsook and Dubnoff (1939) also borrowed features of the
Linderstr0m-Lang and Holter technique as well as the diffusion cell
of Conway and Byrne to develop a method for total nitrogen, am-
monia, and other nitrogenous compounds. The procedure of Borsook
and Dubnoff for 5-10 fig. total nitrogen involves acid digestion,
transfer of an aliquot to a diffusion cell, and finally electrometric
titration of the excess acid. The standard deviation was around
0.05 fxg.

Levy and Palmer ( 1940) adapted the hypobromite method for am-
monia to the iodometric estimation of nitrogen without diffusion.

232 CirV'ETTE COLORIMETRY

The principle of this method, which was originally proposed by
Rappaport (1935), is iodometric measurement of the excess hypo-
bromite remaining after reaction with ammonia according to the
equation: 3 NaOBr + 2 NH3 -^ N2 + 3 H2O + 3 NaBr. The pro-
cedure of Levy and Palmer can be used for total nitrogen in the
range 500-5 /xg. or, if the microvolumetric techniques of Linder-
str0m-Lang and Holter are used, down to 0.5 [xg. After digestion in
small test tubes the material is diluted with water, made alkaline
with a neutralizing reagent, treated with an excess of hypobromite,
and this excess is measured by the addition of potassium iodide and
titration with thiosulfate. The entire treatment can be carried out in
one small tube.

As an improvement on the diffusion procedure of Bentley and
Kirk (1936), Tompkins and Kirk (1942) described a specially con-
structed unit for both digestion and diffusion designed for the meas-
urement of samples containing 0.5-20 /xg. nitrogen. A probable error
of not more than about 1% was claimed by the authors, but this was
challenged by Hawes and Skavinski (1942), who stated that the
diffusion time employed by Tompkins and Kirk is sufficient to trans-
fer only about 90% of the ammonia in samples containing less than
10/xg. nitrogen.

Hawes and Skavinski ( 1942) described another modification of
the diffusion cell technique. They employed a test tube for both di-
gestion and diffusion. A small helix of platinum wire, which was
sealed to a glass tube held in a rubber stopper, was used to hold a
drop of the ammonia-absorbing solution. AIM primary sodium
phosphate solution, rather than the commonly used saturated boric
acid solution, was employed to absorb the ammonia because of the
greater absorption capacity of the former. The volume of phosphate
solution need not be measured ; it is sufficient merely to dip the helix
into the stock solution. After the absorption is complete, the helix is
immersed in water and the solution is titrated with standard acid
to an end point between pH 4.3 and 4.7. Electrometric titration is
advantageous but an indicator may be used; in the latter event a
bromocresol green-methyl red mixture is superior to the methylene
blue-methyl red combination. As used by Hawes and Skavinski, the
method enables the determination of from 10 to 100 fig. nitrogen.

Russell ( 1944) , by utilization of the Conway-Byrne diffusion cell
for the collection of ammonia, refined the phenol-hypochlorite

NITROGEN AND AMMONIA 233

inetliud for the colorimetric determination of ammonia, which had
been used earlier by Van Slyke and Hiller ( 1933) and Borsook
(1935). By the reaction of ammonia with phenol and hypochlorite
in alkaline solution an intense blue product is formed, which is be-
lieved to be indophenol or a closely related compound. The method
has been applied to 1.5 ml. ammonia solution containing 0.5-6.0 ixg.
nitrogen. Larger or smaller volumes may be used to extend the range
of the method.

Boell (1945) employed a method for the estimation of 1-50 /xg.
tutal nitrogen which utilized a digestion similar to that of Levy
( 1936) , transfer of an aliquot of the digest to a diffusion unit made
up of ordinary laboratory glassware rather than the special cells of
Conway and Byrne, and titration with a simplified microburette.
The standard deviation was about 0.1 [xg. nitrogen.

The method of Linderstr0m-Lang and Holter (1933b) for am-
monia has been adapted to the determination of total nitrogen and
subjected to exhaustive critical trial and investigation in the Carls-
berg Laboratory. Changes have been introduced as dictated by
experience and a description of the procedure used in 1939 was
given in a publication of Bottelier, Holter, and Linderstr0m-Lang
(1943). A final summary of the method after further improvements,
with a full treatment of each step, was the subject of a paper by
Briiel, Holter, Linderstr0m-Lang, and Rozits (1946). In principle,
the final method consists merely of the digestion of the sample,
transfer of the digest to the bottom of a paraffin-coated tube, and
titrimetric measurement of the ammonia in a manner similar to
that originally employed by Linderstr0m-Lang and Holter but with
improvements in certain of the details.

The preceding survey indicates that various procedures may be
used for every step in the nitrogen determination. The choice of
method may depend to some degree on the prejudices and prefer-
ences of the experimenter, but the method of Briiel et al. ( 1946) is
recommended as the most foolproof and reliable.

Titrimetric Methods. The Levy and Palmer (1940) method
has the great advantage of employing a simple small tube for all of
the chemical steps, and the diffusion process is avoided altogether.
Certainly the digestion of the sample can be performed in a small
tube regardless of the ensuing procedure. If an ammonia diffusion is
to be employed, it is simpler to use small tubes for it rather than

234 CUVETTE COLORIMETRY

the specially constructed vessels of Conway and Byrne (1933),
Needham and Boell (1939), or Tompkins and Kirk (1942). In this
regard, advantages might be claimed for the method of Briiel et al.
(1946), in which both diffusion and titration are carried out in the
same simple tube, or the Hawes and Skavinski (1942) method, in
which the digestion and diffusion are conducted in the same tube.
However, the former requires transfer of digest and the latter
transfer of acid. The method of Briiel et al. ( 1946) is the most precise
of all (0.005 /xg. nitrogen), and the one most thoroughly tested for
its reliability. This method is described on page 283.

Colorimetric Methods. Levy's (1936) method for the direct
Nesslerization of the digest has the advantage that all of the chemi-
cal operations may be carried out in the same vessel and the diffu-
sion process is eliminated. On the other hand, the phenol-hypo-
chlorite reaction used by Russell (1944) is highly sensitive, and
ammonia, in the quantities that can be determined by Levy's
method, can be analyzed with ordinary macro equipment. However,
diffusion of the ammonia from the digest is essential for a proper
phenolhypochlorite reaction.

Digestion of Sample for Determination of Total Nitrogen

The digestion can be conveniently performed in small tubes of
resistant glass. In the earlier methods after the sample had been
introduced, a small Hengar granule {Hengar Co.) was sometimes
added to prevent bumping, the digestion mixture was introduced,
and the tubes were placed in a drying oven at 120-130° for a few
hours, or at 105-110° overnight, to drive off most of the water. The
digestion was continued by heating the tubes in a sand bath, or in
a copper or aluminum block with holes drilled so that the tubes
could be inserted to a depth not exceeding one third of the tube
length.

Briiel, Holier, Lintlerstr0m-Lang and Rozits Procedure for
Digestion of the sample. In the procedure of Briiel et al. ( 1946) ,
which should be given preference, the digestion tubes are 4 cm.
long, 1.8 mm. inner diameter, and 2.4 mm. outer diameter. The
tubes are cleaned by boiling in sulfuric acid-chromic acid mixture,
rinsed thoroughly with distilled water run through each tube by
means of a thin capillary extending to the bottom, and the water is

NITROGEN AND AMMONIA 235

removed by suction through a thin capillary. The tubes are dried
in vacuo at 100° for 10 min. The rims of the tubes are never touched
with the fingers after removal from the cleaning solution and the
clean tubes are stored in a desiccator or in petri dishes in a room
where no ammonia is allowed and smoking is prohibited. To the
sample in a tube about 4 fx\. of the following mixture is added with
a constriction pipette, properly centered so that the solution deliv-
ered to the bottom of the tube is not drawn up between the pipette
and the inner wall:

1 g. copper sulfate (CuS04-5H20)

10 g. potassium sulfate

0.2 g. sucrose

5 ml. cone, sulfuric acid

Dilute to 100 ml.

(The sucrose is added to make sure that the reduction of the nitrog-
enous impurities in the reagents also takes place in the blanks.)
The constriction pipette is emptied under constant pressure, which,
for the greatest accuracy, is so adjusted that the pipette automati-
cally empties when the tip touches a wall or liquid surface. Before
use the pipette is rinsed inside and out with distilled water. After the
digestion solution is added to the sample the tubes are placed in a
vacuum desiccator over phosphorus pentoxide and the water is
allowed to evaporate at room temperature. To prevent creeping the
bulk of the water is removed at 150 mm. mercury (about 24 hr.) ;
then the drying is continued at 0.1 mm. mercury (another 24 hr.) .

By means of a mouth-operated horizontal pipette with a vertical
delivery tip, 1 [A. cone, sulfuric acid containing 10 mg. selenium/ml.
is added to the charred sample in the digestion tube. (About each
10 mm. of the graduated length of the pipette corresponds to 1 [A.;
a strip of graph paper may be used for the graduation. The selenium-
acid solution is prepared by boiling until clear.) Proper centering of
the delivery tip is essential to prevent the liquid from being drawn
up between the tip and the wall of the tube. Should the liquid be
drawn up as far as the rim where contamination may occur, the
experiment should be discarded. It is also essential to rinse the
pipette both inside and outside each day before use.

The digestion is carried out in small flasks (Fig. 78), which may
be made by forming a constriction in the middle of insulin ampules.
1 ml. cone, sulfuric acid containing 0.4 g. potassium sulfate is placed

236

CUVETTE COLORIMETRY

in the bottom of the flask, and before being used for the first time
the solution is boiled in the flask to drive off water and gases. A
small glass ball (Fig. 79) is used to close the flask when not in use.
The flasks are heated conveniently in a copper block supplied with
an electric coil. Holes in the block permit the flasks to be inserted to
a depth of about 14 mm.; the temperature is kept constant at 295°.
At this temperature the acid is clear; it fumes but does not boil.

; u

/l2 mm.\

) C

)

> (. ;/A

)

14-5 mm.

Fig. 78. Arrangement for
digestion.

Fig. 79. Glass Fig. 80. Removal

stopper for diges- of tubes from diges-
tion flask. tion flask.

From Brilel et al. (1946)

To insert the digestion tubes the flask is removed from the block,
allowed to cool for 30 sec, and then the tubes are inserted by means
of forceps and arranged in a row around the circumference of the
flask (Fig. 78). After replacement of the flasks in the block the
digestion is watched for a few minutes to see whether liquid seals
form across the lumen of the tubes and rise upward. If they rise
higher than the middle of a tube the flask is removed from the block
for a minute or two and the seals collapse. After the initial stage the
digestion is allowed to proceed unattended for 5-6 hr. The tubes are
removed from the slightly cooled flasks by means of a conical glass
rod as indicated in Figure 80. They are rinsed on the outside with
distilled water, dried with a clean towel, and stored in a desiccator
over phosphorus pentoxide until ready for the determination of
ammonia.

NITROGEN AND AMMONIA 237

Digestion Mixtures Used by Other Workers. 1 ml. cone, sul-
furic acid, 0.625 g. potassium sulfate, and 0.075 g. selenium; dilute
about 5 times. — Levy (1936).

1% selenium dioxide and 1% copper sulfate (CuSO4.5Ho0) in
concentrated sulfuric acid and water (1:1) ; after digest is colorless,
cool, add 1 drop saturated potassium persulfate, and continue diges-
tion for 15 min. after the solution becomes water clear. — Borsook
andDubnoff (1939).

3 g. copper sulfate, 1 g. potassium sulfate, and 0.1 g. selenium
dioxide to 300 ml. cone, sulfuric acid. — Needham and Boell ( 1939) .

Cone, sulfuric acid and water (1:1) ; when fuming starts, add 1
drop of 30% hydrogen peroxide. After digest is clear, cool, add a
couple of drops of saturated potassium persulfate, and continue
heating; destroy peroxides by adding water, after cooling, add
another Hengar granule, and place in an oven at about 120° for
30 min. — Levy and Palmer (1940).

Cone, sulfuric acid and water (1:1) saturated with potassium
sulfate and made 0.1% with respect to copper selenite. The latter is
prepared by mixing a strong copper sulfate soln. with one of sodium
selenite and collecting the precipitate. — Tompkins and Kirk (1942).

18 A^ sulfuric acid containing 0.1% selenium dioxide and 0.1%
copper sulfate; use 30% hydrogen peroxide if needed. — Hawes and
Skavinski (1942).

I g. copper sulfate, 1 g. potassium sulfate, 1 g. selenium dioxide,
and 100 ml. 50% sulfuric acid; add saturated potassium persulfate
to the cooled mixture if required. — Boell ( 1945) .

Levy Nesslerization Method for Nitrogen

SPECIAL REAGENTS

Digestion Mixture. See above.

Nessler Reagent (Folin and Wu). Add 15 g. potassium iodide and

II g. iodine to 10 ml. distilled water and introduce an excess of
mercury ( 14-15 g.). Shake well for 7-15 min. or until the dissolved
iodine has nearly all disappeared. Cool in running water when the
solution begins to become pale and continue shaking until the
greenish color of the double iodide (HgIo.2KI) appears. Decant
the solution and wash out vessel with distilled water. Dilute the
solution and washings to 200 ml. Add 75 ml. of this solution and

238 CtfVEtTE COLORIMETRY

75 ml. distilled water to 350 ml. 10% sodium hydroxide (within
1% of 2.5 A^) to make the final Nessler reagent.

PROCEDURE

1. Place a microtome section of tissue in 7 jxl. water in the bottom
of a digestion tube (Fig. 39, page 167). This tube has a total capac-
ity of about 2.5 ml. Add 50 fA. of the Levy digestion mixture and
drive off most of the water at 130° (about 1.5 hr. required). Digest
until water- white (Levy uses a small mobile gas flame),

2. After the tube has cooled, hold almost horizontally and pipette
700 fA. distilled water directly into the digest, delivering the water
with as much force as possible. If complete solution is not attained
at once, it will occur on standing.

3. Add the Nessler solution in a standardized manner (Levy,
1936) : First, "A piece of glass tubing is drawn out to a long capil-
lary and clamped vertically above a platform which may be raised
and lowered by a rack and pinion. It is connected to a source of
air supply at 20 cm. water pressure. The capillary is of such diam-
eter that a rapid stream of air bubbles passes through when im-
mersed in water," Then, the reaction tube is placed in a holder and
set on the platform under the capillary tube with air flowing through
it. Fill a 300 fA. constriction pipette with Nessler reagent and place
one end of a rubber tube connected to the pipette in the mouth, and
hold the pipette in the right hand. "The pinion of the platform is
now turned with the left hand to raise the tube about the bubbler so
that a stream of air bubbles passes from bottom to top of the solu-
tion. Just as soon as the bubbler reaches the bottom the Nessler
reagent is blown in. The platform is then lowered. The entire process
takes five seconds." Tilt the tube back and forth to collect any
solution in the bulb,

4. Make the colorimetric measurement from 10-90 min. after
Nesslerizing. Prepare a blank by following the preceding steps, but
omitting the sample. (Levy used a Pulfrich photometer with filter
S43. The microcuvettes which hold 0.2 ml. were employed.)

Russell Plienol-Hypoclilorite Method for Ammonia

SPECIAL REAGENTS

Alkaline Phenol Reagent. Add some water to 25 g. crystalline
phenol and pour in with stirring 54 ml, 5.0 A^ sodium hydroxide.

UKIC ACID, CREATINE, CREATININE, AND ALLANTOIN 239

Make up to 100 ml. and store in a brown bottle in a refrigerator.

Hypochlorite Solution. Dissolve as much as possible of 25 g. of
ground and sifted calcium hypochlorite in 300 ml. hot water. With
stirring, add 135 ml. 20% potassium carbonate soln. which had
been boiled to drive out ammonia. Heat briefly to about 90°, cool,
and dilute to 500 ml. Filter a little of the soln. and test for calcium
ion by adding some of the potassium carbonate soln. and heating
in boiling water a few min. In the absence of calcium ion the
solution remains crystal clear. If the test is positive, add more
carbonate until a negative test is obtained. Filter the final soln.
and store in small brown bottles in a refrigerator. This soln.
should be water clear and contain 1.30-1.40% free chlorine. Test
for free chlorine by adding 10 ml. water, 2 ml. 5% potassium
iodide, and 1 ml. glacial acetic acid to 2.00 ml. of the hypochlorite
solution. Titrate with 0.100 A^ thiosulfate; 7.5-8.0 ml. should be
required. Occasionally retest the solution. The soln. may also be
prepared from Clorox (page 58),

0.003 M Manganous Chloride or Sulfate.

PROCEDURE

1. Place 1.5 ml. sample in neutral or acid solution not stronger
than 0.01-0.02 A^ (containing 0.5-6.0 /xg. ammonia nitrogen) in a
colorimeter tube, cooled by an ice bath. Add 1 drop manganous salt
soln., 1 ml. cold alkaline phenol reagent, and 0.5 ml. cold hypochlorite
solution. Mix by gentle rotation and place at once in a boiling water
bath for about 5 min.

2. Cool, dilute to a convenient volume such as 6 or 10 ml., and
measure the color intensity with light of about 625 m/x.

URIC ACID, CREATINE, CREATININE, AND ALLANTOIN

Borsook (1935) reported colorimetric methods for uric acid,
creatine, creatinine, and allantoin which were modifications of pro-
cedures already in use, but which are in the_ category of micro
methods, even though no special micro equipment is required. The
colorimetric measurements were carried out spectrophotometrically
in cuvettes taking 3.0-3.5 ml. With the present availability of micro-
cuvettes, these methods could be adapted to the analysis of much
smaller quantities of the substances by substituting smaller test
tubes for the 125 X 9 mm. (inside dimensions) tubes used.

240 CUVETTE COLORIMETRY

The uric acid method is based on precipitation of the suljstance
with zinc, solution of the precipitate in dilute acid and water, addi-
tion of cyanide followed by Benedict's arsenophosphotungstic acid
reagent, and development of the color by a procedure which yields a
clear solution. The method was designed for the analysis of 1 ml. of
sample having less than 1 milligram per cent uric acid, i.e., less then
10 /xg. For another method see page 213.

The creatinine method involves absorption of the compound on
Lloyd reagent, removal of impurities, and liberation of the creatinine
by the same alkaline picrate in which the color is developed. Creatine
is determined by the increase in creatinine after conversion to the
latter. With the 3 ml. cuvettes which Borsook used, the smallest
quantity of creatinine which could be measured was 1 ju,g. The
absolute error with all concentrations was ±0.1 jug. For another
method see page 211.

The allantoin method depends on the enzymatic conversion of
allantoin to allantoic acid (in the presence of cyanide to prevent the
formation of allantoin from uric acid), acid hydrolysis of the
allantoic acid to urea and glyoxylic acid, and colorimetric measure-
ment of the latter. With 2 ml. of sample, the least that can be
measured is 0.05 milligram per cent (1 /xg.) with an error of ±5%.

Sure and Wilder (1941) employed the micro attachment on the
Evelyn photoelectric colorimeter for the measurement of creatine
and creatinine. Gold-plated plungers had to be used because the
ordinary cadmium-coated plungers were corroded by the saturated
picric acid. The error in the conversion of creatine to creatinine by
the procedure employed varied from —0.79 to +2.66% in test
experiments. 1 ml. blood filtrate was used by the authors for each
analysis.

Borsook Method for Uric Acid

SPECIAL REAGENTS

2.5% Zinc Chloride.

10% Sodium Carbonate.

N/14 Hydrochloric Acid.

5% Sodium Cyanide (containing 2 ml. cone, ammonium hydroxide

per liter. Prepare fresh every 6—7 weeks.
Benedict's Arsenophosphotungstate Reagent. Dissolve 100 g. pure

URIC ACID, CREATINE, CREATININE, AND ALLANTOIN 241

sodium tungstate in 600 ml. water. Add 50 g. pm'e arsenic pent-
oxide followed by 25 ml. 85% phosphoric acid and 20 ml. cone,
hydrochloric acid. Boil for 20 min., cool, and dilute to 1 liter.
Folin's Stock Uric Acid Standard Soln. Place 1 g. uric acid, weighed
to 1 mg., in a 1 1. volumetric flask. Add 150 ml. water to 0.6 g.
lithium carbonate in a 250 ml. flask and shake until dissolved.
Filter, and heat the filtrate to 60°. Warm the flask containing the
uric acid in running hot water and pour the warm lithium carbon-
ate soln. into the volumetric flask containing the uric acid; wash
down crystals adhering to the neck. Shake until the uric acid has
dissolved (about 5 min.), cool under the tap, add 20 ml. 40%
formaldehyde, and half fill the flask with water. Add a few drops
of methyl orange and then pipette in slovdy, with shaking, 25 ml.
1 A^ sulfuric acid. Dilute the pink soln. to 1 1. Store the reagent in
v/ell-stoppered bottles and keep in the dark. Diluted standards
made from this soln. will keep for several days, but do not use
them sooner than 1 hr. after they are made.

PROCEDURE

1. Pipette 1-2 ml. of sample, 1 ml. water, and 0.05 ml. 2.5%
zinc chloride into a Pyrex test tube ( 125 X 9 mm. inside) provided
with a ground-glass stopper. Mix well by inversion several times.

2. Add 0.4 ml. 10% sodium carbonate and again mix by inver-
sion.

3. Centrifuge the test tube, pour ofT the supernatant, and take
up the last drop from the lip of the tube with filter paper.

4. Add 0.5 ml. N/14: hydrochloric acid, 1.5 ml. water, and 1 ml.
cyanide soln.

5. Stopper the tube and shake well to dissolve all precipitate
and leave a clear colorless soln.

6. Add 0.2 ml. arsenophosphotungstate soln. and again mix well
by inversion.

7. Place the stoppered tube in a 37° water bath for 40 min., and
then in an ice water bath for 15 min.

8. Centrifuge, transfer some of the supernatant to an absorption
cell, and determine the color spectrophotometrically at 610 mix.

9. In a parallel manner, treat four standard solns. covering the
range to 1.0 milligram per cent uric acid in order to obtain a
calibration.

242 CUVETTE COLORIMETRY

Borsook Method for Creatine and Creatinine

SPECIAL REAGENTS

0.1 N Hydrochloric Add.

0.01 N Hydrochloric Acid.

Lloyd Reagent {Eli Lilly and Co.).

Alkaline Picrate Solution. Combine 10 parts saturated picric acid

and 1 part 10% sodium hydroxide.
Creatinine- Zinc Chloride Standard. Dissolve 1.61 g. creatinine-zinc

chloride (Benedict, 1914) in N/14: hydrochloric acid and make

up to 1 1. This soln. has 1 mg. creatinine/ml. Make up fresh at

least once a month.

PROCEDURE

1. Convert creatine to creatinine: Place 1-5 ml. of sample in a
Pyrex test tube ( 125 X 9 mm. inside) provided with a ground-glass
stopper, add one fourth the vol. of 0.1 A?" hydrochloric acid and mix
by inversion. Insert a piece of thread into the neck of the tube and
stopper. Then autoclave for 20 min. at 30 lb. (130°). Omit this step
for preformed creatinine.

2. After cooling, add 30-40 mg. Lloyd reagent and, with thread
removed, stopper tightly and shake continuously for 10 min.

3. Centrifuge, pour off supernatant, and remove the last drop
from the lip of the tube with filter paper.

4. Resuspend the precipitate in 1 ml. 0.01 A^ hydrochloric acid
and use another 1 ml. portion of the acid to wash down the stopper
and the sides of the tube.

5. Again centrifuge and discard the supernatant, removing the
final drop from the lip with filter paper.

6. Add 3 ml. sodium picrate soln. to remove the creatine from
the Lloyd reagent, and shake gently for 10 min. to develop the color
fully.

7. Centrifuge, and measure the color of the liquid spectrophoto-
metrically at 525 m/x. A linear relationship exists between the
absorption and the creatinine concentration from to 2.0 milligram
per cent.

8. Adsorb standards on Lloyd reagent and treat as above from
step 2 on.

tmiC ACID, CREATINE, CREATININE, AND ALLANTOIN 243

Sure and Wilder Method for Creatine and Creatinine

SPECIAL REAGENTS

Saturated Picric Acid.

10% Sodium Hydroxide (carbonate free).

PROCEDURE

1. Pipette 1 ml. of sample (blood filtrate) into a 15 ml. centri-
fuge tube and add 0.5 ml. picric acid soln.

2. To convert creatine to creatinine, cover the tube with lead foil
and autoclave for 40 min. at 20 lb. pressure. Omit this step for
preformed creatinine.

3. Cool, add 0.1 ml. of the sodium hydroxide, and allow to stand
for 10 min.

4. Run a control on the reagents alone.

5. Using the Evelyn photoelectric colorimeter, set the control at
100 with filter 520-M, and then obtain readings of the unknown. Do
not allow the plungers to remain in contact with the soln. for longer
than 1-2 min. in order to avoid corrosion.

6. Obtain values from a previously established calibration curve.

Borsook Method for AUantoin

SPECIAL REAGENTS

Enzyme Powder. Use urease preparation (Squibb) from soy bean
meal (Van Slyke and Cullen, 1914).

Ammonium Carbonate-Sodium Cyanide. Dissolve 1.153 g. ammo-
nium bicarbonate, 0.891 g. ammonium carbonate, and 0.46 g.
sodium cyanide in water and make up to 200 ml.

10% Trichloroacetic Acid.

2% Sodium Tungstate.

N/15 Sulfuric Acid.

0.5% Phenylhydrazine Hydrochloride in A^/14 hydrochloric acid.
Dissolve commercial phenylhydrazine hydrochloride in water and
decolorize by boiling with activated charcoal. Filter the hot solu-
tion, and, after the filtrate is cooled in an ice-salt bath, precipitate
the phenylhydrazine hydrochloride by addition of cone, hydro-
chloric acid, or by dry hydrochloric acid gas. Filter the precipitate
off by suction; wash once quickly with very cold hydrochloric acid,

244 CUVETTE COLORIMETRY

and place in a desiccator over calcium oxide in the dark. Make

up the soln. just before use.
1.25% Potassium Ferricyanide. Prepare just before use.
Standard Solutio)i. Prepare fresh each week a soln. of 1 mg. allan-

toin/ml.

PROCEDURE

1. Into a Pyrex test tube (125 X 9 mm. inside) provided with a
ground-glass stopper place 10 mg. dry enzyme powder, 0.5 ml.
carbonate-cyanide soln., 2 ml. of the sample to be analyzed, and a
drop of chloroform.

2. Stopper the tube and let stand at 37° for 12 hr. An occasional
shaking aids in the solution of the enzyme.

3. Transfer 2 ml. of the mixture into another test tube and add
0.2 ml. 10% trichloroacetic acid and 0.1 ml. 2% sodium tungstate.
Mix after stoppering by inverting a few times. Then add 0.1 ml.
iV/15 sulfuric acid and again mix by inversion.

4. Place tube in a large beaker of water at room temperature
and heat the water quickly to 90° for 5 min. Cool quickly in ice
water for 2 min.

5. Add 0.3 ml. 0.5% phenylhydrazine soln., stopper and shake
vigorously.

6. Set in a water bath at 60° for 5 min. and quickly cool in ice
water.

7. Centrifuge; carefully float a drop of alcohol on the surface
of the liquid, and again centrifuge to throw down floating particles.

8. Carefully pipette 2 ml. clear supernatant into another dry test
tube and cool it in a dry ice-alcohol mixture contained in a beaker
surrounded by an ice-salt bath maintained at — 15 to — 20°.

9. Cool some cone, hydrochloric acid in the Dry Ice-alcohol
mixture, and, after the tube has been cooling for about 10 min., add
1.5 ml. of the cold hydrochloric acid to it.

10. Stopper, and continually invert in the air until the frozen
material melts.

11. Just before the last of the frozen material disappears add 0.2
ml. potassium ferricyanide soln. and mix quickly by several inver-
sions.

12. Set in an ice-salt bath for 5 min. and then place the tube in
a beaker of water at room temperature.

ASCORBIC ACID 245

13. After 10 min., promptly measure the color (it slowly fades
and becomes turbid) spectrophotometrically at 535 m/x. From to
1.5 milligram per cent allantoin there is a linear relationship between
absorption and concentration.

14. Run the standard soln. in the same manner and at the same
time as the unknown.

ASCORBIC ACID*

The method of Roe and Kuether (1943) for assay of ascorbic acid
was adapted by Lowry, Lopez, and Bessey ( 1945) to determinations
on amounts of blood serum down to 10 fd. Measurements in the
range 0.3 to 1.4 milligram per cent have been made with a standard
deviation, in single determinations, of the order of 0.03 milligram
per cent. Ascorbic acid is converted to dehydroascorbic acid, the
latter is treated with 2,4-dinitrophenylhydrazine, and the osazone
formed is made to yield a colored dehydration product through the
action of sulfuric acid. Pijoan and Gerjovich ( 1946) pointed out that,
while this method is reliable for use on blood, its application to
tissues must be made with caution in regard to oxidation products of
ascorbic acid. This follows since the phenylhydrazine reaction is not
specific for dehydroascorbic acid but can react with structures, such
as diketogulonic acid, which bear no antiscorbutic properties. Before
applying the procedures to tissues, it would be desirable, and perhaps
necessary, to ascertain in advance whether interfering substances
were present in the tissue. A titrimetric method, employing dichloro-
phenol indophenol, which measures ascorbic acid directly and not
dehydroascorbic acid, is given on page 300.

Lowry, I^opez, and Bessey Method for Ascorbic Acid

SPECIAL REAGENTS

Osazone Reagent. Prepare a soln. of 2% dinitrophenylhydrazine
and 0.25% thiourea in 9 N sulfuric acid; centrifuge or filter
through sintered glass if a precipitate develops. Store in a re-
frigerator and discard after 1 month.

65% Sulfuric Acid. Add 70 ml. of the concentrated acid to 30 ml.
water.

* See Bibliography Appendix, Refs. 33, 34, and 43.

246 CU\^TTE COLORIMETRY

1% Suspension of Norit in 5% Trichloroacetic Acid. First treat
the Norit by placing 200 g. in a large flask, add 1 1. 10% hydro-
chloric acid, heat to boiling, filter with suction, transfer the cake
of Norit to a large beaker and add 1 1. distilled water; stir well,
filter, and repeat the whole procedure until the washings give a
negative or very faint test for ferric ion. Dry overnight at 110-
120°. Some grades of activated carbon may not require washing;
this can be determined by running a blank test on trichoroacetic
acid washings of the carbon. If these give no more color than the
acid alone, the washing of the carbon is unnecessary. Suspend 5
g. of the iron-free Norit in 100 ml. 5% trichloroacetic acid. After
the Norit has settled, decant the supernatant and restore the
volume with 5% trichloroacetic acid. Repeat several times to
eliminate some of the very fine floating carbon particles. It is
necessary to prevent carbon from getting into the final sample
since carbon contamination may result in low values. If difficulty
from floating is encountered, add 1 vol. 2% gelatin to 10 vol. of
the acid suspension just before use. Once a week or so, replace
the supernatant by fresh acid to avoid the possibility of contamin-
ation with heavy metals that may slowly leach out of the Norit.

PROCEDURE

1. Place 10 /xl. serum in the bottom of a serological tube (6 X
50 mm.) ; add 40 jxl. of the acid-charcoal suspension, and mix by
tapping the tube. In pipetting the charcoal suspension, first blow
through the pipette to suspend the material and then fill and empty
it rapidly to prevent the charcoal from settling out. Employ a con-
striction pipette with a tip and constriction two to three times wider
than normal to avoid plugging.

2. Cap the tube with a piece of Parafilm or a stopper and centri-
fuge 10 min. at 3000 R.P.M.

3. Transfer 30 fA. of the supernatant to another serological tube;
add 10 fx\. of the osazone reagent, and mix by tapping.

4. Cap the tube and set aside at 38° for 3 hr.

5. Chill the tube in ice water and add 50 fil. ice-cold 65% sul-
furic acid. Mix very well, and, after 30 min. at room temperature,
measure the color intensity at 520 m/x using the 0.05 ml. cuvette with
a Beckman spectrophotometer (page 216). If the vol. is increased
the 0.2 ml. cuvette may be used.

V

ASCORBIC ACID AND GLYCOGEN 247

6. Prepare a standard and blank by adding 4 ml. of the acid-
charcoal suspension to 1 ml. aliquots of freshly prepared 1 milligram
per cent ascorbic acid soln. and water, respectively. Centrifuge and
treat 30 /xl. aliquots in the same fashion as the unknowns. Take care
to avoid floating charcoal, which is more of a problem in the absence
of serum. Correct both standard and unknown for the blank and cal-
culate the result; only the single standard is required since the color
is directly proportional to the concentration of ascorbic acid.

note: Serum may be stored safely for several days in a refrigerator or
for several weeks at — 20° after the acid has been added. The supernatant
acid extract may be separated and stored safely in a refrigerator for at
least several weeks, and presumably at —20° for an indefinite period. The
tubes must be well sealed with rubber stoppers to prevent evaporation.
When samples are stored in the frozen state it is preferable to separate
the supernatant before freezing in order to avoid the troublesome tendency
for the charcoal to float as a result of the subsequent necessity for stirring.
The rapid loss of ascorbic acid in blood at ordinary temperatures makes it
imperative to keep the material cool until it is acidified.

In the preceding method of Lowry et al. the danger of charring
has been minimized by the use of 65% sulfuric acid instead of the
85% acid used in the original method of Roe and Keuther; however,
it is still necessary to cool the reaction mixture. Bolomey and Kem-
merer (1946) suggest the substitution of glacial acetic acid for the
sulfuric acid in order to avoid the danger of charring without the
necessity of cooling. The color intensity developed with the acetic
acid is about half that obtained with the sulfuric acid, which may or
may not be important.

GLYCOGEN

While colorimetric methods for glycogen have not been developed
specifically for histochemical studies, the procedure of Boettiger
( 1946) can be adapted to the micro scale necessary.

In the method of Boettiger, the glycogen obtained by alcohol pre-
cipitation of an alkaline digest of the tissue is dissolved, heated with
an acid solution of diphenylamine, and the color which is developed
is measured (a filter No. 635 is used with the Evelyn photoelectric
colorimeter) . This method enables duplicate determinations on 5-10
jug. glycogen, and if reduced to the volumes required in micro-

248 CUVETTE COLORIMETRY

cuvettes, will enable the corresponding refinement. The chief diffi-
culty is the erratic behavior of the diphenylamine reagent, which
necessitates a new calibration each time the determinations are car-
ried out.

The method of van Wagtendonk et al. ( 1946j is based on the color
development which occurs when Lugol solution is added to the gly-
cogen isolated from the tissue. A Klett-Summerson photoelectric
colorimeter was used with filter No. 54 and the measurements were
made in the range 0.05-2 mg. glycogen in a total volume of 5 ml.
Morris (1946) has pointed out that the color developed with iodine
varies considerably with temperature, and therefore temperature
control is required for accurate work. The concentration of iodine
is also a factor that affects the color intensity, and the importance
was stressed for standardization with glycogen obtained from the
same source as the material to be analyzed. Morris is of the opinion
that the precautions required to render the iodine method sufficiently
accurate constitute a major disadvantage. Nevertheless, if tem-
perature is controlled and care is taken to maintain a constant iodine
concentration the method should yield reproducible results. If, in
addition, the standard solution is prepared from glycogen native to
the tissue to be analyzed, sufficient accuracy should be obtained.

For work on the histochemical level, the tissue may be digested
with alkali and the glycogen precipitated with alcohol according to
steps 1-7 in Heatley's procedm'e (page 299). The glycogen thus iso-
lated can be treated in the manner used by Boettiger or van Wag-
tendonk et al. Some preliminary work will be required, no doubt, to
obtain the proper color intensities for the apparatus employed.

Boettiger Method for Glycogen

SPECIAL REAGENTS

Diphenyla7nine Reagent. The purest diphenylamine must be used.
Oxidized crystals are brownish and impart a blue color to the re-
agent. The compound can be purified by dissolving in alcohol at
55°, and crystallizing out by cooling and adding a little water. The
product is dried and stored in a glass-stoppered dark bottle in a
cool place. Glassware used for the reagent must be free of all
organic matter, hence it must be cleaned without soap and kept
dust-free. Add 100 ml. of glacial acetic acid to 3 g. diphenylamine,

GLYCOGEN 249

and, when completely dissolved, add 60 ml. cone, hydrochloric
acid, with stirring. Store in a glass-stoppered dark bottle in a cool
place, and discard when the reagent begins to acquire a bluish
color. Add a few crystals of sodium hyposulfite to improve the
keeping quality.
Standard Glycogen Solutions. Prepare from glycogen reprecipi-
tated from aqueous soln. by alcohol.

PROCEDURE

1. To the glycogen which has been centrifuged down and washed,
add water to make a solution of suitable concentration — this will
have to be determined for the particular case.

2. Add 5 vol. diphenylamine reagent to 2 vol. glycogen soln.
Mix and centrifuge to eliminate any insoluble material. Avoid get-
ting the reagent on the sides of the tube where it will evaporate
during the heating and leave a film that will not go into solution.

3. Heat the tubes exactly 40 min. in boiling water and plunge
into cold water for at least 3 min.

4. Run glycogen standards parallel with the unknowns.

5. Read color intensities (filter No. 635 with the Evelyn colorim-
eter) within 1 hr. after removal from the bath, using a water blank,
and obtain the values of the unknowns from a calibration curve de-
rived from the standards.

Van Wagtendonk, Simonsen, and Hackett Method for Glycogen

SPECIAL REAGENTS

Lugol Solution. Dissolve 1 g. iodine in a soln. containing 2 g. potas-
sium iodide in 20 ml. water. Store in a well-stoppered dark bottle.

Standard Glycogen Solution. Dissolve 25 mg. glycogen {Eastman
Kodak Co., White Label) in 25 ml. 35% potassium hydroxide.
(See note below.)

PROCEDURE

1. To a given vol. glycogen soln., diluted to an appropriate de-
gree as determined in advance, add 0.01 vol. Lugol soln., and mix
well.

2. Read the color at once (filter No. 54 with the Klett-Summer-
son colorimeter) using a blank consisting of the Lugol soln. diluted
100 times with water.

250 CUVETTE COLORIMETRY

3. Obtain the results from a calibration curve derived by using
various amounts of the standard glycogen soln. Subject the glycogen
standards to the same steps of initial precipitation and color develop-
ment as the unknowns.

note: Constancy of temperature and iodine concentration are essential,
and the glycogen standard should be prepared from glycogen obtained from
the same source as the sample to be analyzed. See page 248.

VITAMIN A AND CAROTENE

By means of 2 mm. quartz microcuvettes and an adapter for the
Beckman spectrophotometer (page 217), which was used for the ab-
sorption measurements, Bessey et al. (1946) succeeded in determin-
ing the vitamin A and carotene in as little as 35 fA. blood serum.
Various volumes of serum greater than 35 /xl. may be used as long as
proportional volumes of the reagents are also employed. The pro-
cedure to be described is based on the use of 60 ix\. The method in-
volves saponification and extraction with solvents of low volatility,
measurement of the absorption by the small volumes at 328 m^u ( for
vitamin A) and 460 m/^ (for carotene), destruction of the vitamin A
absorption by treatment with ultraviolet irradiation, and finally re-
measurement of the absorption of 328 m/x. By measurement of the
absorption of 328 m/x, before and after the destruction of the vitamin
A, absorption at this wavelength due to other substances will not
interfere with the determination.

Method of Bessey et al. for Vitamin A and Carotene

SPECIAL APPARATUS

Ultraviolet Apparatus. A diagram of the apparatus used for the
ultraviolet irradiation is shown in Figure 81. A mercury discharge
lamp (General Electric B-H4) with purple emelope-filter and
transformer are used to furnish the radiation. The brightest part
of the lamp is placed opposite to the lower half of the tube, and
the shadow of the electrode support is not allowed to fall on any
tube. The tubes must be cooled during the irradiation by a moder-
ate air current from a fan.

Mixing Apparatus. The device for mixing the liquids in the narrow
tubes is made by cutting off the head of an eight-penny nail,

VITAMIN A AND CAROTENE

251

slightly flattening the end for a distance of 10-15 mm., inserting
the nail in a small high-speed hand drill with the end projecting
about 20 mm., and mounting the drill vertically with the nail up.
The liquids are mixed by touching the side of the tube near the
bottom to the rapidly rotating nail. If this apparatus is not avail-
able mixing may be effected by adding a 1 cm. length of stainless
steel wire (0.041 in. diameter — may be obtained from Newark
Wire Cloth Co.) and shaking. For the extraction with the kero-
sene-xylol, the open ends of the tubes are sealed in a flame and
then shaken vigorously. Care is taken to prevent contamination
of the tops of the tubes by serum which would be charred when
the tubes are sealed.

Fig. 81. Arrangement for ultra-
violet irradiation. Mercury lamp (A)
is held vertically in a clamp, base up,
with the other end extending 3 or 4
cm. into a hole 8 cm. in diameter,
B-B, in a large block of wood, C,
which serves as a base. Semicircular
racks {D, D') are provided for hold-
ing the glass tubes in a circle equi-
distant from the lamp (6 cm. from the
center of the lamp). These racks may
be made from pieces of quarter inch
plywood held about 2 cm. apart, with
the upper piece drilled to hold the
tubes. Twenty or thirty holes may be
drilled in each rack along a semicir-
cular line. From Bessey et al. (1946)

' A

^

1

1
1

1

1

-t

D'

u

1 c

B

SPECIAL REAGENTS

1 N Potassium Hydroxide in 90% Alcohol. Add 1 vol. 11 A^ potas-
sium hydroxide to 10 vol. absolute alcohol. Prepare on the day it
is used. If the color develops rapidly or if the reagent gives a blank,
reflux the alcohol with potassium hydroxide and distill.

252 CUVETTE COLORIMETRY

Kerosene-Xylol. Mix equal vol. xylol, C. P., and water white odor-
less kerosene (obtainable from Einier and A7nend).
Anhydrous Propionic Acid.

PROCEDURE

1. Place 60 fxl. serum and 60 [A. alcoholic potassium hydroxide
in a test tube 100 X 3 mm. (Prepare the tubes by cutting 200 mm.
lengths of glass tubing, 3-3.5 mm. internal diameter, cleaning with
boiling half -cone, nitric acid, rinsing, drying, and dividing in the
middle with a blast lamp flame to yield two tubes.) If the liquids do
not run to the bottom, send them down with a whipping motion.

2. Mix the liquids, immerse the tube in a 60° water bath for 20
min., cool, and add 60 /A. kerosene-xylol.

3. Extract by holding the tube at a 45° angle against the whirl-
ing nail so that the contents are violently agitated for 10-15 sec.
Then, when the tube is at room temperature or a little below, centri-
fuge for 10 min. at 3000 R.P.M.

4. Cut the tube with a file just above the kerosene-xylol layer
and pipette this layer into the cuvette with a constriction pipette,
taking care to avoid the inclusion of any of the aqueous liquid which
would cause turbidity. (Use an uncalibrated 50-60 fA. constriction
pipette (page 172) with a fine tip and a fine constriction because of
the low surface tension of the organic solvents.)

5. Read the absorption at 460 and 328 m/x, and then transfer the
liquid to a soft glass test tube 40 mm. long and 2.5-3.0 mm. internal
diameter.

6. Irradiate with ultraviolet for 30-60 min. (Find the proper
time by testing with known vitamin A solns. The time should be six
to eight times that required to destroy half the vitamin in pure soln. )

7. Transfer the liquid back into a cuvette and take a second
reading at 328 m/A. Rinse the pipette with anhydrous propionic acid
before the transfer to eliminate traces of moisture which would cause
turbidity.

CALCULATION

^460 X 480 = microgram per cent carotene
(^328 — -2^328 after irradiation) X 637 := microgram per cent

vitamin A

VITAMIN A AND CAROTENE 253

where E — optical density for 1 cm. cuvette = 2 — log per cent
transmission with 1 cm. cuvette. The factor 637 is based on an E
(1%, 1 cm.) of 1720 for vitamin A palmitate in alcohol at 328 m/x,
calculated as free alcohol. The factor 480 is based on an E (1%,
1 cm.) of 2080 for /3-carotene (Sniaco) in kerosene-xylol.

///. TITRIMETRIC TECHNIQUES

A. MICROLITER BURETTES

The microliter burettes employed in histochemical procedures
fall into two general groups. In one a capillary glass tube is
calibrated so that the volume of liquid delivered can be determined
by observing the position of a meniscus. These burettes are usually
modifications of the Brandt-Rehberg (1925) instrument, which is
arranged to move the column of solution by the pressure of a
mercury thread controlled by a screw. Mercury in a reservoir is
displaced by turning in the screw and the displaced mercury moves
into the glass capillary. Instruments of this general type have been
described by Pincussen ( 1927) , Linderstr0m-Lang and Holter
(1931, 1933a), Kirk (1933), Sisco, Cunningham, and Kirk (1941),
Links (1934), and Boell (1945). In the Heatley (1935, 1939)
microburettes the pressure is supplied by leveling-bulb arrange-
ments, and both Conway (1934) and Hawes and Skavinski (1942)
employ hydrostatic pressure in their instruments. The Conway
burette was modified by Ramsay ( 1944) for use under anaerobic
conditions (page 279).

In the other general group of burettes a calibrated capillary tube is
not used, but the screw, usually in the form of a micrometer, is cali-
brated instead. These are essentially modifications of the instrument
described by Widmark and Orskov ( 1928) . Krogh and Keys ( 1931) ,
Kirk (1933), and Krogh (1935) employed a fine screw to move the
plunger of a small glass syringe for the accurate delivery of small
volumes of liquid (page 174). Trevan (1925), Dean and Fetcher
(1942), and Hadfield (1942) used the spindle of a micrometer to
operate the plunger. Probably the best micrometer burette is that
designed by Scholander ( 1942) and later improved by Scholander,
Edwards, and Irving (1943). In this instrument the spindle of the
micrometer is used to displace the mercury in the reservoir. An

255

256

TITRIMETRIC METHODS

advantage of this group of microburettes is that their accuracy is
independent of the lumen of the capillary.

Linderstr^ni-Lang and Holler Burettes. These instruments
possess an approximately fivefold refinement of the original Brandt-
Rehberg ( 1925) instrument, and they have been constructed in two
main forms. The type 1 burette ( Linderstr0m-Lang and Holter,
1931), shown in Figure 82, has a cahbrated glass capillary tube

Fig. 82. Burette, type 1.

From Linderstr0m-Lang

and Holter (1931)

Fig. 83. Burette, type 2.

From hinder At rOm-Lang

and Holter (1933a)

58 cm. long, having a total capacity of 100 ix\. and graduated in
divisions of 0.2 ix\. Estimations may be made to 0.02 /xl. When the
screw in the bottom is turned in, the mercury is forced up into the
capillary, which, in turn, forces the liquid out of the burette. The
tip of the burette is dipped into the liquid to be titrated in order that
quantities less than a drop may be added. Readings are taken from
the meniscus of the top of the mercury column. In filling the burette
the tip is dipped into the standard solution and the screw is reversed.
The top of the mercury column is in contact with the standard
solution.

MICROLITER BURETTES

257

In the type 2 (Lindersti'0m-Lang and Holter, 1933a), shown in
Figure 83, the mercury is separated from the standard solution by
an air space. This instrument is used for solutions which might be
affected by contact with mercury. When the screw S is manipulated,
the right mercury column, which is open to the air, is raised or
lowered, and this results in a small positive or negative pressure over
the left column. In this way liquid can be delivered from, or drawn
into, the burette.

The type 1 burette can be connected to a permanent reservoir of
standard solution as shown in Figure 84 ( Linderstr0m-Lang and
Holter, 1933b).

^

L^

8. Burette, type
with reservoir.
Frovi hinder str0m-Lang
and Holter (1933b)

Fig. 85. Glass bead used

to exclude air during titration.

From hinder sir 0m-Lang, Weil

and Holter (1935)

Reduction of evaporation and protection from the air during
titration is afforded by the loosely fitting glass cap around the tip
of the burette held suspended by two threads (Fig. 82). This effect
may also be obtained by passing the tip of the burette through a
glass bead, P (Fig. 85) , which rests on the top of the titration vessel
(Linderstr0m-Lang, Weil, and Holter, 1935). The glass bead may
be dipped into paraffin oil first in order to effect a better seal to the
titration tube. In order to carry out titrations in an atmosphere free

258

TITRIMETRIC METHODS

from carbon dioxide, Schmidt-Nielsen (1942) designed a soda lime
container that fits on the titration tube as shown in Figure 86.

It may be observed in Figures 64 and 84 that a titration table
is employed which is adjustable both vertically and horizontally.
An opal-glass background, illuminated by a small electric bulb, is

Reservoir

Fig. 86. Titration with "desiccator" to maintain
carbon-dioxide-free atmosphere.
From Schmidt-Nielsen (1942)

provided to facilitate the observation of the color of the solution.
The electromagnet placed to the left of the titration table is used
for magnetic stirring in the manner described in the section dealing
with stirring devices (page 179).

MICROLITER BURETTES

259

Tlie complete titration assembly is available from A. H. Thomas
Co. and E. Petersen, Carlsberg Laboratory.

Fig. 87. Burette, front and rear views:
From Sisco, Cunninghavi, and Kirk (1941)

Kirk Burette. In this instrument (Fig. 87) the mercury is
separated from the standard solution by air, and the readings are
taken from a scale behind the capillary tube rather than from

260

TITRIMETRIC METHODS

graduations on the tube itself. The burette has a total capacity of
about 0.1 ml. and is capable of a precision in reading of ±0.03 /tl.
(Sisco, Cunningham, and Kirk, 1941). (Available from Micro-
chemical Specialties Co.)

Heatley Burette. The essential differences between the Heatley
( 1935) burette and the preceding models are that the mercury
displacement is effected by means of a leveling bulb rather than a
screw, the standard solution is in contact with paraffin oil, and
delivery is made directly from a stock bottle of the solution. A
diagram of the instrument is shown in Figure 88. The tube leading

100 B

I I I I I I— I I I I I

TIT

Fig. 88. Burette.
From Heatley (1935)

from the delivery tip extends to the bottom of the stock bottle which
has a 2 oz. capacity and is lined with paraffin. The stopper of the
stock bottle is a cork infiltrated with paraffin, and the 5 mm. vertical
tube H ends flush with the bottom of the cork. The upper part of H,
the capillary connection {M) , the three-way stopcock {E), reservoir
(F), and the space (D) are all filled with paraffin oil. The space
under D and part of the capillary tube {B) contain mercury. The
leveling bulb arrangement (C), by which the pressure is regulated,
also contains mercury. An air space exists between the leveling
bulbs and the mercury thread in the capillary. No air is permitted
between the mercury under D and the delivery tip. Interchangeable
delivery tips fitted through a ground-glass joint may be used. When
not in use, the tip is dipped into a tube of the titration solution

MICROLITER BURETTES

261

covered by a layer of paraffin oil. This serves to protect the tip,
and if the protecting tube and the tip are sealed together by a rubber
connection, siphoning-over of the solution is prevented. By adjusting
C so that only a small positive pressure is applied, the surface
tension at the fine tip will prevent liquid from escaping, and delivery
will occur only when the tip is immersed in the solution to be titrated.
This, of course, is the principle used for the other microburettes
that have been described. Stock bottles can be interchanged without
affecting the capillary in any way. One of the instruments that
Heatley constructed had a capacity of 0.1 ml. over a 25 cm. scale,
and the capillary was divided in 1.0 [x\. graduations. The relatively
large volume of the stock bottle and the air space between the
mercury in the leveling bulb and that in the capillary would make
this burette particularly prone to errors arising from temperature
fluctuations during titration.

Fig. 89.