Skip to main content

Full text of "Clinical hematology, a practical guide to the examination of the blood, with reference to diagnosis"

See other formats





I. Normal erythrocytes. 
II. Large and small lymphocytes and transitional forms. 

III. Polynuclear neutrophiles. 

IV. Blood plaques. 
V. Eosinophiles 

VI. Myelocytes. 
VII. Mast cells. 

(E. F. Faber, fee.) 








Secon> Edition, IRevfsefc anfc 









C. 2>aCosta, flD.H)., 




Three noteworthy lines of advance have developed from recent 
work in hematology: the identification of new clinical entities, 
the correlation of blood pictures with a number of diseases hitherto 
unstudied or imperfectly investigated hematologically, and the 
proof of the septic nature of many of the specific infections. The 
first has established upon a clinical basis trypanosomiasis, kala-azar, 
spotted (Montana) fever, and cyanotic polycythemia ; the second 
has shown in detail the blood changes incident to chloroma, hydatid 
infection, multiple periostitis, pancreatitis, variola, arthritis defor- 
mans, sprue, leukanemia, burns, and x-ray therapy; and the last 
has proved the bacteriemic character of enteric fever, Malta fever, 
pneumonia, scarlet fever, and, possibly, rheumatic fever. Prog- 
ress of great practical Value has been made in the technic of 
blood examinations, notably in staining methods, in serum diag- 
nosis, and in blood culturing. These advances, all made during 
the past two years, have been incorporated in this edition, together 
with much other new material, gleaned by the consultation of 
more than nine hundred late references to hematological literature. 

The original data of the book are based upon the records of 
about ten thousand blood examinations, made chiefly at the Ger- 
man, the Jefferson, and the Philadelphia General Hospitals. This 
new matter relates more especially to the primary anemias, malig- 
nant disease, cholelithiasis, icterus, pancreatitis, gastric ulcer, pneu- 
monia, septicemia, and suppurative lesions. For the sake of clear- 
ness, tabulations have been avoided as far as possible, and the 
data analyzed and presented as summaries. Among the improve- 
ments in technic described are Wright's stain, Milian's method 
of estimating the coagulation time, Reudiger's serum test, medico- 
legal tests for blood, and cryoscopy. A brief account is given of 
Ehrlich's side-chain theory and its relation to immunity and to 
hemolysis. The detailed revision of the text has been supple- 
mented by the addition of a new- colored plate and numerous other 

In this revision the plan of the first edition has been adhered 
to the interpretation of the blood report as a rational aid to 


diagnosis. The author has profited by the views of his critics 
whenever they could be consistently adopted, and he begs to 
acknowledge his appreciation of the many suggestions from this 

November i, 1904. 


This book, designed as a practical guide to the examination of 
the blood by methods adapted to routine clinical work, represents 
an endeavor to recount the salient facts of hematology as they are 
understood at the present time, to correlate certain of these facts 
with familiar pictures of disease, and to apply them to medical 
and surgical diagnosis. The purpose has been to interpret the 
blood report according to its true value as a clinical sign, neither 
exploiting it as a panacea for every diagnostic ill, nor belittling it 
because of its failure consistently to give the sought-for clue in 
every instance. 

A minimum amount of theoretical discussion has been intro- 
duced in the sections dealing with the physiology and pathology of 
the whole blood and of the cellular elements only sufficient, in the 
author's judgment, to add clearness to the number of the mooted 
points of this science, which in its present transitional stage must 
still be regarded as one from which more or less hypothesis and 
conjecture are inseparable. Intimate familiarity with technic 
being an essential qualification for the comprehensive study of the 
blood, a somewhat lengthy consideration of this subject is given. 
The methods of examination likely to prove useful in everyday 
practice have been described in detail, perhaps somewhat at the 
risk of prolixity, in the hope of thus simplifying for the novice the 
minutiae of blood counting, staining, and other means of investi- 
gation. In the discussion of the primary anemias and of the 
anemias peculiar to infancy, prominent clinical features other than 
those referable to the blood have been briefly mentioned, in order 
to add clearness to the differential diagnosis. For convenience in 
reference, the various diseases included in the section on general 
hematology are arranged alphabetically, rather than grouped 
according to a traditional classification. 

The greater part of the original data referred to in the text is 
taken from the records of the Pathological Institute of the German 
Hospital, where a systematic account of all blood examinations has 
been kept for the past six years. The remaining data represent 
the writer's personal examinations in hospital and private practice 
and in the Army Medical Service, these sources of statistics 
together including about four thousand blood reports in various 
pathological conditions. 


Hematological literature has been freely consulted in the prepa- 
ration of this volume, special acknowledgment being due to 
Hayem, Ehrlich and Lazarus, von Limbeck, Rieder, Lowit, Turk, 
Grawitz, Cabot, Stengel, Thayer, Ewing, Taylor, and Coles for 
the profitable information gleaned from their writings. Due 
credit in the text has been given to these as well as to the other 
authors of whose labors use has been made. 

The colored plates and other histological illustrations, the origi- 
nals of which were made by Mr. E. F. Faber from fresh and stained 
specimens, bear evidence of the artist's technical skill and faithful 
attention to structural detail. Mr. S. Trenner has kindly furnished 
the engravings of several of the special instruments. 

The author takes pleasure in acknowledging the assistance of his 
wife and critic in revising the proofs of these pages; in crediting 
Dr. G. P. Miiller for collecting and verifying much statistical 
matter relating to hospital cases; and in thanking Dr. J. Chalmers 
Da Costa and Dr. T. G. Ashton for helpful suggestions. 

November 1901 








Obtaining the Specimen, 33 

Preparing the Slide, 35 

Microscopical Examination, 37 

Changes Affecting the Erythrocytes, 37 

Changes Affecting the Leucocytes, 37 

Increase in Fibrin, Blood Plaques, and Hemoconia, 38 

Blood Parasites, 39 

Foreign Bodies, 39 


Dare's Hemoglobinometer, 41 

Von Fleischl's Hemometer, 43 

Oliver's Hemoglobinometer, 49 

Gowers' Hemoglobinometer, 52 

Haldane's Hemoglobinometer, 53 

Tallquist's Method, 54 

Haig's Blood Decimal Card, 55 


Methods, 55 

Diluting Fluids, 56 

The Thoma-Zeiss Hemocytometer, 57 

Counting the Erythrocytes, .- 60 

Counting the Leucocytes, 64 

Cleaning the Pipette, 67 

Durham's Hemocytometer, 68 

Gowers' Hemocytometer, 69 

Oliver's Hemocytometer, ' 71 

Dry Film Method, 73 


Objects of Staining, 75 

The Anilin Dyes, 76 

Preparing the Films, 76 

Fixation Methods, 79 

Methods of Staining, 8r 

Wright's Stain, 82 

Ehrlich's Triacid Stain, 83 

Prince's Stain 85 

Staining with Eosin and Methylene-blue, 86 

Staining with Eosin and Hematoxvlin, 87 

Staining with Thionin 87 

Staining with Polychrome Methylene-blue, 88 

Differential Counting, 89 





Determann's Method, 90 


Daland's Heraatocrit, 91 

Limitations of the Hematocrit, 93 


Hammerschlag's Method, 94 


Engel's Alkalimeter, 96 

Dare's Hemo-alkalimeter, 98 


Glass Slide Method, 100 

Wright's Coagulometer, 101 


Cryoscopy in Pathological Conditions, 102 

Fontaine's Cryoscope, 104 


Hamburger's Method, 106 


The Sorby-Beck Microspectroscope, 107 


Value of Positive Findings, 109 

Methods, 109 

Blood Cultures, 109 

Staining Methods, in 


WidaPsTest, 112 

The Serum Reaction in Specific Infections 113 


Methods, 114 

Microscopical Examination, 114 

Spectroscopy, 115 

Teichmann's Hemin Test, 115 

The Guaiacum Test, 116 

The Biological Test, 117 

Hanging Drop Test, 122 





Plasma, Serum, and Cells, 125 

Salts, 126 

Extractives, 126 

Gases, 126 

II. COLOR, 126 

Normal Variations, 126 

Density and Opacity, 126 

Pathological Variations, 127 





Reaction in Health, 128 

Table of Normal Blood Alkalinity, 129 

Physiological Variations, 1 29 

Pathological Variations, 130 


Normal Range, 132 

Pathological Variations, 132 

Relation of Specific Gravity to Hemoglobin, 133 

Table of Hemoglobin Equivalents, 133 


Relation of Fibrin to Coagulation, 134 

Appearance of Fibrin in Fresh Blood, 135 

Hyperinosis and Hypinosis, 135 

Pathological Variations in Amount of Fibrin, 136 


Definition, 138 

Occurrence, 138 

Vin.jF-LETHORA, .' 138 

Definition, 138 

Permanent and Transient Polyemia, 139 

Serous Plethora, 139 

Cellular Plethora, 139 


Definition, 139 

Causes, 140 

Occurrence, 140 


Definition, 140 

Causes, . . 141 

Occurrence, 141 


Amount of Fat in Normal Blood, 141 

Definition, 141 

Physiological arid Pathological Lipemia, 142 

Tests for Fat, 142 


Definition, 142 

Occurrence, 142 


Amount of Sugar in Normal Blood, 143 

Hyperglycemia, 143 

Test for Sugar, 143 


Definition, 144 

Occurrence, 144 

Test for Uric Acid, 144 


Definition, 145 

Occurrence, 145 

Test for Bile, 145 




Definition, 145 

Occurrence, 145 

Tests for Acetone and Fatty Acids, 145 


Occurrence, 146 

Latent Infection, 146 

Blood Cultures, 147 

Bacteria Found in the Blood, 148 


Definition, 148 

Pseudo-anemia, 149 

Classification, 150 

Pathogenesis, 151 


Ehrlich's Side-chain Theory, 151 

Hemolysis, 154 

Antihemolysis, 155 

Agglutination and Precipitation, 156 




General Properties, 161 

Origin, 162 

Variations in Amount, 163 

Absolute Amount, 165 

Color Index, 165 

Hemoglobinemia, 166 

Methemoglobinemia, 167 

Carbon Mono.xid Hemoglobin, 168 


Appearance in Fresh Blood, 169 

Histological Structure, 170 

Origin and Life History, 171 

Size, 172 

Normal Number, 173 

Volume Index, 173 


Age and Sex, ^ 174 

Pregnancy, Menstruation, and Lactation, 175 

Constitution and Nutrition, 176 

Muscular Exercise, 176 

Fatigue, 176 

Digestion and Food, 177 

High Altitudes, 178 

Climate, 180 


Ameboid Motility, 180 

Alterations in Isotonicity, 180 

Hyperviscosity, 181 



Deformities of Shape and Size, 182 

Megalocytes, 182 

Microcytes, 182 

Poikilocytes, 183 

Endoglobular Degeneration, 184 

Total Necrosis, 185 

Atypical Staining Reaction, 186 

Nucleation, 187 

Normoblasts, 187 

Megaloblasts, 189 

Microblasts, 192 

Atypical Erythroblasts, 192 

Ring Bodies, 194 

Granular Basophilia, 194 

Schiiff ner's Granules, 196 

Oligocy themia, 196 

Polycythemia, 197 


Appearance in Fresh Blood, 198 

Histological Structure, 199 

Origin, 199 

Normal Number, 200 

Pathological Variations, 200 


Appearance in Fresh Blood, 200 

Histological Characteristics, 200 

Occurrence, 201 




Appearance in Fresh Blood, : 205 

Ameboid Movement, 206 

Cell Granules, 207 

Normal Number, 209 


Number and Percentage of Different Varieties, 209 

Small Lymphocytes, 210 

Large Lymphocytes, 211 

Transitional Forms, ". 213 

Polynuclear Neutrophiles, 214 

Eosinophiles, 216 

Basophiles, 217 

Myelocytes, 217 

Mast Cells, 219 

Mononuclear Neutrophiles, 222 

Neutrophilic Pseudolymphocytes, 222 

Reizungsformen, 222 

Differential Table of the Leucocytes, 223 

Origin and Development, 224 

lodin Reaction, 226 

Perinuclear Basophilia, 228 




Definition, 228 

Classification of the Leucocytoses, 229 

Physiological Leucocytosis, 230 

Character, 230 

Causal Factors, 230 

Leucocytosis of the New-born, 230 

Digestion Leucocytosis, 231 

Leucocytosis of Pregnancy and Parturition 233 

Leucocytosis Due to Thermal and Mechanical Influences, 234 

Terminal Leucocytosis, 235 

Pathological Leucocytosis, 236 

Occurrence, 236 

Degree of Increase, 236 

Differential Changes, 237 

Causal Factors, 237 

Functions, 238 

Hypoleucocytosis and Hyperleucocytosis, 239 

Leucocytolysis, 240 

Changes in the Bone Marrow, 241 

Inflammatory and Infectious Leucocytosis, 242 

Post-operative Leucocytosis, 244 

Leucocytosis of Malignant Disease, 245 

Post-hemorrhagic Leucocytosis, 246 

Toxic Leucocytosis, 247 

Ether Leucocytosis, 248 

Chloroform Leucocytosis, 249 

Experimental Leucocytosis, 249 


Definition, 252 

Differential Changes, 252 

Causal Factors, 253 

Physiological Lymphocytosis, 253 

Pathological Lymphocytosis, 254 

Experimental Lymphocytosis, 254 

Clinical Significance, 254 


Definition, 255 

Causal Factors, 256 

Physiological Eosinophilia, 256 

Pathological Eosinophilia, 256 

Experimental Eosinophilia, 257 

Diminution in the Number of Eosinophiles, 257 

Clinical Significance, 258 



Definition, 259 

Occurrence, 259 

Causal Factors, 260 


Definition, 260 

Differential Changes, 261 

Physiological Leucopenia, 261 

Pathological Leucopenia, 262 

Experimental Leucopenia, 262 





Volume, Oxygen Capacity, and Albumin, 267 

Appearance of the Fresh Blood, 267 

Coagulation, 268 

Specific Gravity, 268 

Alkalinity, 268 

Hemoglobin and Erythrocytes, 269 

Color Index, 269 

Deformed and Nucleated Erythrocytes, 271 

Leucocytes, 272 

Differential Changes, 272 

Blood Plaques, 274 

Diagnosis, 274 

Clinical Features, 275 


Volume, Oxygen Capacity, and Albumin, 276 

Appearance of the Fresh Blood, 276 

Coagulation, 277 

Specific Gravity, 278 

Alkalinity, 278 

Hemoglobin and Erythrocytes, 278 

Color Index, 279 

The Blood During Remissions, 280 

Megalocytosis, 281 

Poikilocytosis, 281 

Prevalence of Megaloblasts, 282 

Polychromatophilia, 285 

Granular Basophilia, 285 

Leucocytes, 285 

Differential Changes, 286 

Blood Plaques, 287 

Diagnosis, 287 

Clinical Features, 288 

Pernicious Anemia and Severe Secondary Anemia, 289 

Pernicious Anemia and Chlorosis, 289 

Pernicious Anemia and Bothriocephalus Anemia, 290 

Pernicious Anemia and Nitrobenzene Poisoning, 290 

Conversion of Pernicious Anemia into Leukemia, 290 

Leukanemia, 290 


Appearance of the Fresh Blood, 291 

Hemoglobin and Erythrocytes, 291 

Color Index, 291 

Deformed and Nucleated Erythrocytes, 292 

Leucocytes, 292 

Blood Plaques, 293 

Diagnosis, 293 

Clinical Features, 294 

Splenic Anemia and Myelogenous Leukemia, 295 

Splenic Anemia and Pernicious Anemia, 295 

Splenic Anemia and Hodgkin's Disease, 295 

Splenic Anemia and Splenic Tumors, 295 


Appearance of the Fresh Blood, 296 



Coagulation, 296 

Specific Gravity, 296 

Alkalinity 296 

Hemoglobin and Erythrocytes, 296 

Color Index, 297 

Deformed and Nucleated Erythrocytes, 297 

Leucocytes, 298 

Differential Changes, 298 

Blood Plaques, 298 

Diagnosis, 298 


Etiology, 299 

Immediate Effects of Hemorrhage, 299 

Secondary Effects of Hemorrhage, 300 

Degree of Blood Loss Compatible with Life, 300 

Regeneration of the Blood, 301 

Blood Crises, 302 



Varieties, 302 

Conversion of Leukemia into Pernicious Anemia, 304 

Parasitology, 305 

Myelogenous Leukemia, 306 

Appearance of the Fresh Blood, 306 

Coagulation, 307 

Alkalinity, 307 

Specific Gravity, 307 

Hemoglobin and Erythrocytes, 308 

Color Index, 308 

Relation of Erythrocyte and Leucocyte Counts, 308 

Erythroblasts, 309 

Leucocytes, 310 

Influence of Arsenic on the Leucocyte Count, 311 

The Blood During Remissions, 311 

Influence of Rontgen-ray Treatment, 312 

Differential Changes, 312 

Blood Plaques, 317 

Lymphatic Leukemia, *. 317 

Appearance of the Fresh Blood, 317 

Hemoglobin and Erythrocytes, 317 

Deformeci and Nucleated Erythrocytes, 318 

Differential Changes, 319 

Blood Plaques 321 

Acute Leukemia, 321 

Influence of Intercurrent Infections, 322 

Diagnosis, 323 

Myelogenous and Lymphatic Leukemia, 325 

Leukemia and Pathological Leucocytosis, 325 

Leukemia and Lymphocytosis, 325 

Leukemia and Hodgkin's Disease, 326 

Leukemia and Chloroma, 326 

Leukemia and Multiple Periostitis, 326 

Leukemia and Still's Disease, 326 

Leukemia and Tumors of the Spleen, Kidney, and Pancreas, 327 

Leukemia and Lymphatic Hyperplasia, 327 


Appearance of the Fresh Blood, 327 



Hemoglobin and Erythrocytes, 327 

Color Index, 328 

Nucleated and Deformed Erythrocytes, 328 

Leucocytes, 328 

Differential Changes, 328 

Conversion of Hodgkin's Disease into Leukemia, 329 

Diagnosis, 330 

Clinical Features, 331 

Influence of Rontgen-ray Treatment, 331 

Hodgkin's Disease and Tuberculous Adenitis, 331 

Hodgkin's Disease and Syphilitic Adenitis, 332 

Hodgkin's Disease and Local Lymphoma, 332 

Hodgkin's Disease and Lymphatic Sarcoma, 332 

Hodgkin's Disease and Lymphatic Carcinoma, 333 


Hemoglobin and Erythrocytes, 333 

Leucocytes, 335 

Differential Changes, 336 

Factors of the Blood Changes Following Splenectomy, 336 






Fetal Blood, 341 

The Blood at Birth, 342 


Frequency, 345 

General Characteristics, 345 

Classification, 346 

Primary Anemia, 347 

Pernicious Anemia, 347 

Splenic Anemia, 347 

Leukemia, 348 

Secondary Anemia, 351 

Mild Anemia, . . -. 35 1 

Severe Anemia, 351 

Anemias with Leucocytosis, 352 

Etiology of Secondary Anemia, 352 

Anemia Due to Syphilis, 352 

Anemia Due to Rachitis, 353 

Anemia Due to Tuberculosis, 353 

Anemia Due to Gastro-intestinal Diseases, 353 

Post-typhoid Anemia, 354 

Anemia Infantum Pseudoleukemica, 354 

Bacteriemia, 35^ 


I. ABSCESS, 3 or 

Coagulation, Fibrin, and lodin Reaction, 361 

Hemoglobin and Erythrocytes 3 01 

Factors of the Anemia in Abscess, 3 or 

Cell Deformity and Nucleation, 3 6z 



Leucocytes 3 2 

Relation of the Leucocyte Count to the Local Lesion, 362 

Range of the Leucocyte Count in Different Forms of Abscess, . 363 

Differential Changes, 3^3 

Diagnosis, 33 







Factors of the Anemia, 36 

Grade of Anemia in Catarrhal and Suppurative Cases, 366 

Hemoglobin and Erythrocytes, 3^6 

Cell Deformity and Nucleation, 367 

Leucocytes - 3^ 

Range of the Leucocyte Count in Different Forms of Appendi- 
citis, 369 

Differential Changes, 369 

lodin Reaction, 37 

Diagnosis, 37 

Clinical Value of the Blood Examination, 371 






Coagulation, 376 

Bacteriology, 376 

Serum Reaction, 377 

Hemoglobin and Erythrocytes, 377 

Leucocytes, 377 

Blood Plaques, 378 




Fibrin and Coagulation, 380 

Bacteriology, 380 

Hemoglobin and Erythrocytes, 380 

Leucocytes, 381 

Diagnosis, 381 




Alkalinity, Lipemia, Lipacidemia, and Glycemia, 383 

Williamson's Test, 383 

Bremer's Test, 384 

Hemoglobin and Erythrocytes, 385 

Leucocytes, 386 

Digestion Leucocytosis, 386 

lodin Reaction, 386 

Diagnosis, 386 




Hemoglobin and Erythrocytes, 386 

Leucocytes, 387 

Course of the Leucocytosis, 388 

Influence of Antitoxin on the Leucocyte Count, 388 

Differential Changes, 388 

Affinity of the Leucocytes for Basic Dyes, 389 

Diagnosis, 390 


Enteritis, Diarrhea, and Gastro-enteritis, 390 

Dysentery and Ulcerative Enteritis, 391 

Effect of Purges, 392 


Alkalinity and Coagulation, 393 

Bacteriology, 393 

Blood Cultures, 393 

Spot Cultures, 395 

Serum Reaction, 396 

Dried Blood Method, 398 

Fluid Blood Method, 399 

Macroscopical Method, . . . ..." 400 

The Test with Dead Cultures, 401 

The Choice of a Method, 401 

Value of the Serum Test, 402 

Positive Reactions in Non-typhoid Conditions, 403 

Hemoglobin and Erythrocytes, 404 

Cell Deformity and Nucleation, 406 

Leucocytes, 406 

Differential Changes, 408 

Effect of Complications, 408 

Blood Plaques, 409 

Diagnosis, 409 




Factors of the Blood Changes, 412 

Pyrexial Polycythemia and Post-febrne Anemia, 412 

Coagulation, Fibrin, and Alkalinity, 412 

The Leucocytes, 412 


Occurrence, 413 

Parasitology, 413 

The Filaria Nocturna, 414 

Technic of Examination, 418 

Staining the Filariae, 4*9 

Hemoglobin and Erythrocytes, 419 

Leucocytes, 420 

Diagnosis, 421 



Acute and Chronic Forms, 421 

Hyperchlorhydria, Hypochlorhydria, Gastric Achylia, Gastric 

Dilatation, Gastric Neurasthenia, 422 

Diagnosis, 4 2 3 


Hemoglobin and Erythrocytes, 423 



Effect of Hemorrhage and Emesis, 423 

Leucocytes, 4 2 4 

Diagnosis, 425 



XXXI. GOUT, 426 

Alkalinity, Fibrin, and Uric Acid, 426 

Cellular Elements, 4 2 7 

Perinuclear Basophilia, 4 2 7 


Specific Gravity, 4 2 7 

Bacteriology, 4 2 7 

Alkalinity, 428 

Coagulation, 428 

Hemoglobin and Erythrocytes, 429 

Leucocytes, 4 2 9 

Blood Plaques, 430 


Anemia in Atrophic Cirrhosis, 430 

Effect of Ascites, 431 

Anemia in Hypertrophic Cirrhosis, 431 

Leucocytes in Atrophic and Hypertrophic Cirrhoses, 432 

Diagnosis, 433 




Fibrin, Coagulation, Specific Gravity, and Alkalinity, 434 

Hemoglobin and Erythrocytes, 435 

Leucocytes, 436 

Diagnosis, 436 




Parasites Causing Anemia, 438 

Factors of the Blood Changes, 439 

Hemoglobin and Erythrocytes, 439 

Bothriocephalus Anemia, 439 

Ankylostomiasis Anemia, 440 

Leucocytes, 440 



Parasitology, 44 1 

Leishman-Donovan Bodies, 442 

Hemoglobin and Erythrocytes, 443 

Leucocytes, 44 ^ 

Diagnosis, 443 



Parasitology, 44^ 

Developmental Cycle of the Malarial Parasite in Man, 445 

Developmental Cycle of the Malarial Parasite in the Mosquito. 446 
Varieties of the Malarial Parasite, 447 



The Parasite of Tertian Fever, "448 

Infections with Single and Multiple Groups, 448 

Anticipation of the Paroxysm, 449 

Intracellular Hyaline Forms, 449 

Intracellular Pigmented Forms, 450 

Segmenting Forms, 45 1 

Extracellular Pigmented Forms, 452 

Flagellate Forms, 453 

Degenerate Forms, 454 

The Parasite of Quartan Fever, 454 

Infections with Single and Multiple Groups, 454 

Intracellular Hyaline Forms, 454 

Intracellular Pigmented Forms, 455 

Segmenting Forms, 456 

Extracellular Pigmented Forms, 456 

Flagellate Forms, 45 7 

Degenerate Forms, 45 7 

The Parasite of Estivo-autumnal Fever, 45 7 

Irregularities in Time of Developmental Cycle, 457 

Disc- and Ring-shaped Forms, 458 

Pigmented Forms, 458 

Segmenting Forms, 459 

Erythropyknosis, 459 

Spherical, Ovoid, and Crescentic Forms, 460 

Flagellate Forms, 461 

Degenerate Forms, 461 

Pigmented Leucocytes and Phagocytosis, 462 

Differential Table of the Malarial Parasites, 463 

Technic of Examination, 464 

Artefacts in Fresh Blood Specimens, 465 

Hemoglobin and Erythrocytes, 466 

Causes of Malarial Anemia, 466 

Anemia in the Regularly Intermittent Fevers, 466 

Anemia in Estivo-autumnal Fever, 467 

Anemia in Malarial Cachexia, 468 

Nucleated and Deformed Erythrocytes, 468 

Types of Post-malarial Anemia, 469 

Leucocytes 469 

Differential Changes, : 47 

Blood Plaques, 47 1 

Diagnosis 47 1 


Carcinoma, 47 2 

Fibrin, Coagulation, Specific Gravity, and Alkalinity, 472 

Glycemia, 47 2 

Parasitology, 473 

Hemoglobin and Erythrocytes, 473 

Regeneration of the Blood after Operation, 474 

The Oligocythemia and Polycythemia of Gastric Cancer, . . 474 

Deformed and Nucleated Cells, 475 

Leucocytes, 475 

Frequency of Cancer Leucocytosis, 475 

Causes of Cancer Leucocytosis, 47 

Range of the Leucocytes in Different Forms of Cancer, . . 476 

Digestion Leucocytosis in Gastric Cancer, 477 

Differential Changes, 477 

Sarcoma, 47^ 

General Features of the Blood 47** 

Cytology, 47 8 



Hemoglobin and Erythrocytes, 478 

Leucocytes, 479 

Diagnosis, 479 


Bacteriology, 482 

Hemoglobin and Erythrocytes, 482 

Leucocytes .' 483 

Diagnosis, 483 






Factors of the Blood Changes, 489 

The Edema of Anemia, 489 

Specific Gravity, Fibrin, Coagulation, and Alkalinity, 490 

Bacteriology, 490 

Hemoglobin and Erythrocytes, 490 

Anemia in Acute and Chronic Parenchymatous Nephritis, 490 

Polycythemia, 491 

Anemia in Chronic Interstitial Nephritis, 491 

Leucocytes, 491 

Uremia, 492 

Diagnosis, 492 


Neuritis, Beri-beri, Neuralgia, Akatama, and Erythromelalgia, 492 

Brain Tumor, 493 

Neurasthenia, Hypochondriasis, and Hysteria, 493 

General Paresis, Dementia, Melancholia, and Mania, 494 

Convulsions, Apoplectiform Attacks, and Acute Delirium, . . . 495 
Epilepsy, Chorea, and Tetany, 497 








Serous Pleurisy, 503 

Purulent Pleurisy, 504 

Diagnosis, 505 


General Features of the Blood, 505 

Bacteriology, 505 

Hemoglobin and Erythrocytes, 507 

Leucocytes, 507 

Relation of Leucocytosis to Intensity of Infection, 507 

Frequency and Extent of Leucocytosis, 508 

Effect of Induced Leucocytosis, 509 

Effect of Antipyresis, 510 

Differential Changes, 510 



Blood Plaques, 510 

Diagnosis, 511 




Parasitology, 514 

LowenthaPs Reaction, 516 

Hemoglobin and Erythrocytes, 516 

Leucocytes, 516 

Diagnosis, 517 


Coagulation, Fibrin, and Alkalinity, 518 

Bacteriology, 5 18 

Hemoglobin and Erythrocytes, 519 

Leucocytes, 520 

Diagnosis, 520 


Coagulation, Fibrin, and Specific Gravity, 521 

Bacteriology, 521 

Hemoglobin and Erythrocytes, 522 

Leucocytes, 523 

Blood Plaques, 524 

Diagnosis, 525 


General Features of the Blood, 525 

Serum Reaction, 525 

Bacteriology, 526 

Hemoglobin and Erythrocytes, 527 

Deformed and Nucleated Erythrocytes, 528 

Leucocytes, 529 

Differential Changes, 529 

Diagnosis, 530 



Bacteriology, 532 

Hemoglobin and Erythrocytes, 532 

Syphilitic Chlorosis and Pernicious Anemia, 532 

Effect of Mercury and Iodides on the Blood, 533 

Justus' Test, 533 

Leucocytes, 534 

Diagnosis, 535 

Blood Plaques, 535 





Parasitology, 538 

Cellular Elements, 54* 

Diagnosis, 54 1 


General Features of the Blood, 541 

Bacteriology, 542 

Serum Reaction, 542 



Hemoglobin and Erythrocytes, 544 

Anemia in Pulmonary Tuberculosis, 545 

Anemia in Bone Tuberculosis, 546 

Anemia in Tuberculous Adenitis, Meningitis, Pericarditis, 
Pleurisy, and Peritonitis, and in Genito-urinary Tubercu- 
losis 546 

Leucocytes, 546 

Differential Changes, 547 

Range of the Leucocytes in Pulmonary Tuberculosis, 547 

Range of the Leucocytes in Bone Tuberculosis, 548 

Range of the Leucocytes in Acute Miliary Tuberculosis, Tuber- 
culous Adenitis, Pleurisy, Peritonitis, Pericarditis, and Men- 
ingitis, and in Genito-urinary Tuberculosis, 549 

Effect of Secondary Septic Infections, 549 

Diagnosis, 550 


Parasitology, 550 

Hemoglobin and Erythrocytes, 551 

Leucocytes, 551 

Diagnosis, 551 



Stage of Compensation, 552 

Acute Rupture of Compensation, 552 

Effect of Stasis, 553 



Parasitology, 555 

Hemoglobin and Erythrocytes, 556 

Leucocytes, 556 

Blood Plaques, 557 

Diagnosis, 557 


Fibrin and Coagulation, 558 

Bacteriology, 558 

Hemoglobin and Erythrocytes, 559 

Degenerative Changes, 560 

Leucocytes, 560 

Diagnosis, 560 





The Leucocytes, Frontispiece. 


I. The Erythrocytes, 161 

II. The Leucocytes, 205 

III. Leucocytosis, 228 

IV. Myelogenous Leukemia, ; 306 

V. Lymphatic Leukemia, 317 

VI. The Tertian Malarial Parasite, 449 

VII. The Quartan Malarial Parasite, 455 

VIII. The Estivo-autumnal Malarial Parasite, 459 


I . Pernicious Anemia, 279 

II. Myelogenous Leukemia, 311 

III. Multiple Infections in Malarial Fever, 447 


1 . Blood Lancet, 34 

2. Proper Distribution of Cells in a Blood Film, 36 

3. Zones of Rouleaux and of Isolated Cells in a Fresh Blood Film, 36 

4. Dare's Hemoglobinometer, , 40 

5. Horizontal Section of Dare's Hemoglobinometer, 41 

6. Method of Filling Blood Chamber, 43 

7. Von FleischFs Hemometer, 43 

8. Tinted Wedge of von Fleischl's Hemometer, . : 43 

9. Capillary Pipette of von Fleischl's Hemometer, 43 

10. Method of Using von Fleischl's Hemometer, 46 

1 1 . Light-proof Box for von Fleischl's Hemometer, 48 

12. Method of Using Oliver's Hemoglobinometer, 51 

13. Gowers' Hemoglobinometer, 52 

14. Thoma-Zeiss Hemocytometer, 57 

15. Thoma-Zeiss Counting Chamber, 58 

1 6. Ruled Area of Thoma-Zeiss Counting Chamber, 59 

17. Ruled Area of Zappert's Counting Chamber, 60 

18. Method of Filling the Hemocytometer, 61 

i() . Plan of Counting the Erythrocytes, 63 

20. Ocular Diaphragm, 65 

2 1 . Expelling Contents of the Erythrocytometer, 67 

22. Cross-section of Durham's Blood Pipette, 68 

23. Method of Using Oliver's Hemocytometer, 72 

24. Superimposing the Charged Cover-glass, 77 

25. Drawing Apart the Cover-glasses, 77 

26. The Cover-glasses after Separation, jS 

27. Spreading a Film with Two Glass Slides, 78 

28. Spreading a Film with Cigarette Paper, 78 

29. Oven for Fixing Blood Films, 79 




30. Daland's Hematokrit, 92 

31. Engel's Alkalimeter, 97 

32. Dare's Hemo-alkalimeter, 99 

33. Milian's Coagulation Test, 100 

34. Wright's Coagulometer, 101 

35 . Fontaine's Cryoscope, 105 

36. Sorby-Beck Microspectroscope, 107 

37. Sorby Tubular Cell, 108 

38. Aspirating Tube for Blood Culturing, no 

39. Rouleaux Formation and Fibrin in Normal Blood, 136 

40. Hyperinosis, 137 

41. Mechanism of Toxin-cell Union, 152 

42. Elaboration and Action of Antitoxin, 153 

43. Mechanism of Hemolysis, 155 

44. Mechanism of Antihemolysis, 156 

45. Blood Spectra, 168 

46. Deformities of Size and Shape of the Erythrocytes. 182 

47. Degenerative Changes in the Erythrocytes, 184 

48. Megaloblasts, 190 

49. Atypical Forms of Erythroblasts, 193 

50. Granular Basophilia, 159 

5 1 . Changes in the Erythrocytes in Chlorosis, 269 

52. Changes in the Erythrocytes in Pernicious Anemia, 280 

53. Atypical Myelocytes in Myelogenous Leukemia, 313 

54. Atypical Polynuclear Neutrophiles in Myelogenous Leukemia, 315 

55. Atypical Lymphocytes in Lymphatic Leukemia, 320 

56. Positive Serum Reaction in Enteric Fever, 396 

57. Pseudo-reaction in Enteric Fever, 396 

58. Bacillus Typhi Abdominalis, 397 

59. Filaria Nocturna in Fresh Blood, 414 

60. Filaria Nocturna, showing Granular Degeneration, 416 

61 . Filaria Nocturna, showing Changes in Shape, 418 

62 . Leishman-Donovan Bodies, 442 

63 . Spirilla of Relapsing Fever 480 

64. Trypanosoma Gambiensc, 540 


The rapid growth and development of hematology during 
recent years and the practical application of many of its teachings 
to the diagnosis of various diseases have made this science one 
which no progressive medical man can afford to disregard. Ex- 
amination of the blood gives definite clinical information which 
may be profitable both to the practitioner of internal medicine 
and to the surgeon, and the procedure is capable of throwing light 
upon the diagnosis in so wide a range of pathological conditions 
that it is difficult to single out any disease in which it may not 
be of some utility, either as positive or as negative evidence. 

In the light of our present knowledge of the subject, clinical 
information of two kinds may be derived from hematology, 
namely, findings which are pathognomonic of certain diseases; 
and auxiliary data which, if considered in connection with other 
clinical manifestations, may prove either essential or helpful in 
establishing the precise nature of a disease. 

The list of diseases in which pathognomonic blood findings 
are met with includes leukemia, the malarial fevers, relapsing 
fever, filariasis, trypanosomiasis, and piroplasmiasis. In per- 
nicious anemia a typical picture is also found, if two conditions 
capable of exciting identical. blood changes are excepted, the pro- 
found secondary anemias due to certain intestinal parasites and to 
nitrobenzene poisoning. 

The blood examination affords data which, although not 
pathognomonic, are nevertheless essential for the diagnosis of 
chlorosis, Hodgkin's disease, splenic anemia, chloroma, Osier's 
disease, multiple periostitis, kala-azar, and secondary anemias 
dependent upon various causes. For example, in chlorosis a 
definite group of blood changes must exist in order to justify 
an unconditional diagnosis, although the occurrence of these 
changes, unassociated with other equally definite clinical signs, 
is insufficient evidence of this disease. In Hodgkin's disease, 
a condition indistinguishable from leukemia by an ordinary 
physical examination, the absence of a leukemic state of the 
blood at once excludes the latter disease. In the secondary 
anemias, it is obvious that the blood count alone can give the 


exact clue to the condition, by determining the degree and char- 
acter of the blood impoverishment, and by tracing from time to 
time its progress. In this connection it is important to remember 
that pallor may go hand in hand with a normal hemoglobin per- 
centage and erythrocyte count, and that, on the other hand, a high 
color by no means invariably signifies that the individual is not 
anemic. In addition to the diseases just named, hematology gives 
information which is often of great assistance in, although not 
essential for, the diagnosis of such conditions as enteric fever, 
sepsis, pneumonia, pertussis, appendicitis, diabetes, rabies, syph- 
ilis, gastric ulcer, malignant disease, helminthiasis, the exanthe- 
mata, the hemorrhagic diseases, and suppurative processes. 
Clinical experience has repeatedly illustrated the value of the 
serum reaction in enteric and Malta fevers and in dysentery; of 
Williamson's test in diabetes mellitus; of eosinophilia in trichini- 
asis, in echinococcus disease, and in other forms of helminthiasis ; 
of mononucleosis in variola and in pertussis ; and of leucocytosis 
and iodophilia in sepsis, in suppurative lesions, and in many of 
the acute infections. The forensic use of Bordet's test for the 
identification of blood stains, and the application of cryoscopy to 
the diagnosis of renal disease are also of practical utility. 

Negative results from a blood examination also possess diag- 
nostic value within certain limits, but too great reliance upon 
evidence of this sort more often proves delusive than helpful. In 
a patient whose waxy, yellowish facies suggests with equal force 
pernicious anemia, chronic nephritis, and, perhaps, liver cirrhosis, 
the absence of characteristic blood changes is sufficient to exclude 
the first-named condition. But failure to detect the malarial 
parasite does not necessarily exclude malarial fever; a negative 
serum test does not absolutely rule out enteric fever; and an 
absence of leucocytosis cannot be regarded as an infallible sign 
that a suppurative focus does not exist, nor does it always indicate 
the benignity of a neoplasm. Negative evidence, then, is usually 
to be considered merely suggestive, the real pertinence of the hint 
thus obtained depending upon its correlation with other physical 
signs and symptoms. 

The significance of positive findings in bacteriological investi- 
gations of the blood is patent, and recent improvements in technic 
have made this means of research simple, dependable, and certain. 
Blood cultures furnish conclusive information in general septi- 
cemia, pneumonia, enteric fever, Malta fever, plague, scarlet 
fever, malignant endocarditis, and similar conditions in which 
bacteria invade the blood stream. 

At the present time the most useful information furnished by 


hematology has been derived from study of the cellular elements 
of the blood, but closer familiarity with the chemistry of this tissue, 
still an undeveloped science, will undoubtedly in the near future 
afford not only more tangible clues to the etiology and pathology 
of the blood diseases, but also will bring to light additional facts 
which may be applied to the diagnosis of these and other maladies. 
The study of the coagulation time of the blood is of practical 
utility in the study of purpura, hemophilia, jaundice, and other 
conditions characterized by slow clotting and by a tendency 
toward hemorrhage. 

The technic of blood examinations, such as described in the 
following pages, is neither elaborate nor difficult to master. Nec- 
essarily, it must be rigidly exact, but no more so than any other 
branch of physical diagnosis, if the worker is content only with the 
best results. To acquire a good working knowledge of hema- 
tology takes but a fraction of the time and application that one 
must spend in familiarizing one's self with the most common heart 
murmurs or chest signs, and the time thus spent equips the physi- 
cian with an additional diagnostic agent of the greatest value. If 
the newly graduated physician would provide himself with a 
microscope and a set of blood instruments, and systematically 
study the blood in the various general diseases which he encounters 
in practice, many a slip-shod diagnosis might be avoided, and a 
great stride forward made in popularizing this practical branch 
of clinical diagnosis. 

" L'avenir appartient a I'hematologie. C'est elle qui nous apportera la solu- 
tion des grands problemes nosologiques. Elle doit nous apparaltre comme une 
vaste science puisant ses materiaux dans toutes les branches des connaissances 
biologiques et recueillant les di-verses notions de Vhumorisme ancien pour les 
rajeunir et les cojnplettr a la lumiere des decouvertes modernes en anatomic, en 
physiologic, en chimie biologique el en pathologic" 






P A systematic examination of the blood by 

Sen MA clinical methods of established utility includes 

the following different processes: 
I. Microscopical examination of the fresh blood. 
II. Estimation of the percentage of hemoglobin. 

III. Counting the erythrocytes and the leucocytes. 

IV. Microscopical examination of the stained specimen. 
These four procedures, which should invariably be included in 

every clinical blood report, furnish the most important informa- 
tion to be derived from hematological study, and are sufficient 
for routine clinical work. In certain instances in which more de- 
tailed investigation of special points is sought, it may be thought 
advisable to supplement the above plan by employing one or 
more of these remaining procedures: 

V. Counting the blood plaques. 
VI. Estimation of the relative volumes of corpuscles and 


VII. Estimation of the specific gravity/ 
VIII. Estimation of the alkalinity. 
IX. Determination of the rapidity of coagulation. 
X. Cryoscopical examination. 

XI. Estimation of the resistance of the erythrocytes. 
XII. Spectroscopical examination. 

XIII. Bacteriological examination. 

XIV. Determination of the serum reaction. 
XV. Medico-legal tests for blood. 



The finger-tip or the lobe of the ear is the part 

OBTAINING THE usually selected from which to obtain the blood, 

SPECIMEN. by puncture, for examination. The former site 

is preferable in most instances, owing to its con- 

3 33 


vcnient situation and ease of manipulation; but in nervous indi- 
viduals and in children the ear-lobe may be chosen, because of its 
limited sensibility and on account of the patient's inability to 
watch the operation. 

The puncture may be made with one of the special blood 
lancets devised for this purpose, or, in lieu of such an instrument, 
a Hagedorn or spear-pointed surgical needle or a new sharp- 
pointed steel pen from which one nib has been' twisted off, will 
answer the purpose equally as well. The author is accustomed 
to use a small steel trocar blade, mounted on a metal shaft which 
screws into an outer barrel by means of a thread. By the use 
of a threaded locking-nut, any desired length of the trocar may 
be exposed, so that the depth of the wound may be controlled 
at will, irrespective of the force used to drive the point of the 
instrument through the skin. It is not necessary to sterilize the 
puncture-needle: wiping it with a towel wet with alcohol is all 
that is required in ordinary examinations. Of course, should 
the patient happen to be syphilitic or septic, it is safer to pass 
the blade through an alcohol flame after having used it. 

Having chosen, say, the patient's 
middle or ring-finger, the part is first 
thoroughly cleansed with alcohol or 
ether and then with water, and wiped 
FIG. i. -BLOOD LANCET. perfectly dry with a clean, lint-free 

towel, which may then be folded into 

a pad and slipped behind the finger to isolate it from the 
neighboring digits, and to serve as a cushion for the back 
of the hand. The operator, holding the patient's hand in a 
firm, steady position, makes the puncture \vith a rapid motion 
of the wrist, such as one is accustomed to use in percussing 
the thorax, the depth of the wound being just sufficient to 
cause a free flow of blood in good-sized drops, unaided by 
the slightest pressure on the finger other than that necessary to 
start the initial oozing. The needle should be aimed so as to 
strike a point in the center of the flexor surface of the finger, just 
back of the extreme tip. The blood drop to be used for the ex- 
amination should under no circumstance be squeezed from the 
finger, for blood secured in this manner is certain to be more or 
less highly diluted with lymph from the surrounding tissues a 
condition which will lead to erroneous results, especially to lower 
hemoglobin, specific gravity, and corpuscular estimations than 
actually exist. In severe anemias, especially in those of the 
pernicious type, the bloodless condition of the superficial ves- 
sels is sometimes so marked that it may be impossible to obtain 


enough blood for the examination by an ordinary puncture, even 
from the ear-lobe, which, as a rule, is highly vascular. Relatively 
deep incisions are unavoidable in such instances. On the con- 
trary, in most cases of leukemia, unless the coexisting anemia 
is of striking intensity, the blood usually flows very freely, and 
may even spurt from the wound in a fine jet several inches in 

Most writers on hematology utter an emphatic warning against 
hemophilics, in whom the slightest prick of a needle may cause 
troublesome bleeding. The writer has never had the misfortune 
to meet with this accident, but recognizes the wisdom of observ- 
ing the precaution to question every patient concerning an 
abnormal tendency toward hemorrhage. 

The observer's attention should be directed to the color and 
the density of the blood drop as it flows frpm the puncture, and 
a note taken of the various macroscopical changes which may 
occur, such as the pale, hydremic condition of the blood found 
in severe anemias, the deep blue color in cyanosis, and the milky 
appearance in leukemia and in diabetes. These and other altera- 
tions in the naked-eye appearance of the fresh blood have been 
discussed in another section. 

The first few drops of blood which follow the 

PREPARING puncture are wiped away, and the site of the in- 
THE cision freed from every trace of moisture, after 

SLIDE. which a perfectly clean cover-glass, held edge- 
wise between the thumb and forefinger, is lightly 
touched to the summit of the next drop as it oozes from the punc- 
ture, and is then immediately placed, blood side downward, 
upon the surface of a clean glass slide. If the cover-glass and 
the slide are perfectly clean and dry, and if the drop is of the 
proper size, the blood will at once spread out in a thin film con- 
sisting of a single layer of corpuscles (Fig. 2), surrounded by an 
outer zone in which the cells are heaped up in masses and rouleaux ; 
this thicker area of the specimen is unsuited for examination 
(Fig. 3). Gently heating the slide over an alcohol flame just 
before use will insure a thin, even spread. If prolonged study 
of the specimen is intended, it is advisable to exclude air from 
the film, by ringing the margins of the cover-glass with a thin 
layer of cedar oil or vaselin, but ordinarily this precaution is 
unnecessary. In order to prevent distortion of the corpuscles, 
pressure must be avoided while adjusting the cover-glass. If 
the blood does not spread of itself, without the aid of pressure, 
it is usually owing to the presence of particles of dust or grease 
between the opposed surfaces of the slide and cover-glass. 


Absolute cleanliness of the covers and slides is an essential 
detail to which too great attention cannot be paid, for neglect of 
this precaution is responsible for the majority of failures to secure 
good specimens. Perhaps the most useful cleansing agent is the 


solution popularly known as "acid alcohol" (hydrochloric acid, 
i part; absolute alcohol, 29 parts; water, 70 parts), which quickly 
-and effectually removes all traces of grease and dirt from the glasses, 
so that their preliminary soaking in soap-suds^ or in a strong 
mineral acid, as some 
recommend, may be dis- 
pensed with. The slides 
and covers may be con- 
veniently kept in closed 
glass receptacles contain- 
ing this solution, from 
which they are removed 
as the occasion demands, 
being then dried and 
polished with a bit of 
clean linen or with tissue- 
paper. Ordinary soft 

"toilet-paper" is excellent for this purpose. Oblong cover-glasses, 
measuring X i inches and of "No. i" thickness, are more 
easily handled without forceps than smaller square or circular slips, 
and also have a much larger surface than the latter, which is often 
decidedly advantageous. 



The use of forceps is unnecessary if care is observed to hold 
the cover-glass in the manner already directed, so that only its 
edges come in contact with the thumb and finger. 

The specimen, prepared in the manner just 
MICROSCOPICAL described, is examined under the microscope 
EXAMINATION, with both low and high powers, a ^ or ^ inch 
dry, and a y 1 ^ inch oil-immersion, objective being 
the most satisfactory lenses for the purpose. The substage con- 
denser and diaphragm should be adjusted so that the field is but 
moderately illuminated, rather than flooded with a glare of white 
light. Microscopical examination of the fresh blood film furnishes 
information about the following points: 

Changes Affecting the Erythrocytes. With a little practice one 
soon becomes able to detect, with a tolerable degree of accuracy, 
any conspicuous decrease in the number of erythrocytes, by the 
relatively small number of cells in the field in comparison with 
their number in a similar field of normal blood. With less con- 
fidence it is also possible to decide whether or not the number 
of erythrocytes is much in excess of the normal standard. 

Deficiency in hemoglobin produces unmistakable changes in 
the appearance of the cells, those in which this change is well 
defined appearing as pale, washed-out bodies which stand in 
striking contrast to the darker, yellowish-green color of the normal 

Abnormal viscosity of the erythrocytes, their tendency toward 
rouleaux formation, the presence of deformities of size and of 
shape, and the occurrence of structural degenerative changes 
may also be distinguished in the fresh, unstained blood film. 
Nucleated erythrocytes are not demonstrable in the fresh 

Changes Affecting the Leucocytes. A glance is usually suffi- 
cient to determine whether or not the number of leucocytes is 
markedly in excess of normal, but too great dependence should not 
be placed on such a method of detecting the presence or absence 
of a leucocyte increase, since it is at the best approximate, and 
sometimes erroneous. As will be explained elsewhere, any 
marked decrease in the number of erythrocytes, the leucocytes 
remaining normal, may so increase the ratio of the latter to the 
former, that the leucocytes may be apparently increased. 

Having tentatively determined that an increase in the total 
number of leucocytes is present, it is furthermore possible for 
one familiar with the morphology of the unstained leucocyte to 
make a fairly accurate differential count of these cells, and thus 
to decide whether the increase is due to a pure leucocytosis 


or to some form of leukemia. This distinction is not at all diffi- 
cult in most instances, when one recalls the characteristics of 
the several forms of leucocytes in the fresh blood, viz.: small 
lymphocytes, large lymphocytes, and transitional forms, appear- 
ing as cells having a single spherical or indented nucleus, and a 
clear, shining, non-granular protoplasm ; polynuclear neutrophiles, 
as cells with polymorphous or multiple nuclei, and a protoplasm 
crowded with very fine, moderately refractive granules; eosino- 
philes, as cells with a single polymorphous nucleus or multiple 
nuclei, and a protoplasm containing coarse, spherical, highly re- 
fractive, fat-like granules; and myelocytes, as cells with a single 
spherical or ovoid nucleus, and a protoplasm crowded with very 
fine, moderately refractive granules. It is, of course, obviously 
impossible to distinguish basophile cells in the fresh blood, as 
well as some of the cells containing fine eosinophile granules, 
but the characteristics noted above are sufficiently plain to justify 
at least a provisional diagnosis of either of the conditions in ques- 
tion, which, in every instance, should be verified by a careful 
examination of the stained specimen. 

While most of the degenerative changes which affect the leuco- 
cytes are clearly demonstrable only in the stained specimen, it is 
still possible to recognize some of the grosser examples of such a 
process by a study of the fresh film. Vacuolation of both nucleus 
and protoplasm, extrusion of portions of the cell substance, and 
the various stages of nuclear disintegration and of apparent solu- 
tion of the protoplasm are the alterations most commonly ob- 
served. In certain specimens "fractured" leucocytes are seen 
with more or less frequency, a cell thus affected being drawn out 
into a diffuse, irregularly shaped body with indistinct and ragged 
margins, about which the cell granules, which have escaped from 
the protoplasm, are scattered in the form of a nebulous mass. 
The eosinophile leucocytes seem especially prone to undergo 
this disintegration. The exact significance of this phenomenon 
is not clear, but it probably represents a degenerative change in 
which the cells have become abnormally vulnerable, and thus 
highly susceptible to mechanical injury from the pressure of the 

Ameboid activity of the leucocytes and pigmentation of these 
cells are among the other changes to be observed in a histolog- 
ical examination of the unstained blood film. 

Increase in Fibrin, Blood Plaques, and Hemokonia. The den- 
sity of the fibrin network and the rapidity with which it forms 
may be studied as coagulation of the blood film progresses. Un- 
less the blood plaques are very greatly increased in number, 


they are not usually noticeable in the specimen prepared in the 
ordinary manner. The presence of hemokonia, or "blood dust," 
is at once rendered conspicuous by the rapid and incessant molec- 
ular motion with which these bodies are endowed. 

Blood Parasites. The hematozoa of the malarial fevers, the 
spirilla of relapsing fever, the organisms of trypanosomiasis, 
and the embryonic forms of the parasite of filarial disease should 
be studied in the fresh blood whenever this is possible, rather 
than in the fixed and stained film, since in the latter the charac- 
teristic morphology of these parasites is greatly altered and their 
motility lost. The stained specimen is more useful in studying 
the finer structure of these organisms than for diagnostic examina- 

The distoma of bilharzia disease, although, strictly speaking, 
a blood parasite, is not found in the general 'circulation, since this 
worm resides solely in the portal vein and branches, the vena cava, 
and certain veins of the lower pelvis. Leishman-Donovan 
bodies are obtained by puncture of the spleen, but they do not 
enter the peripheral blood. 

Foreign bodies, such as free fat droplets, collections of extra- 
cellular pigment, and, very rarely, the crystalline bodies of Char- 
cot may also be observed in the fresh specimen during the course 
of certain diseases. 

Microscopical examination of the fresh specimen should form 
the initial step taken in every systematic examination of the 
blood, since it may be the means of determining whether or not 
a more elaborate investigation is necessary. By this simple pro- 
cedure an immediate diagnosis may be made in a number of in- 
stances, while in others the findings, although not pathogno- 
monic, are of distinct clinical value. Close familiarity with the 
normal histology of the blood is, of course, essential for the ap- 
preciation of the various pathological changes which have been 
outlined above. Fuller reference to these changes has been 
made in other parts of this book. (See Sections III and IV.) 


No fewer than half a dozen different hemoglo- 

METHODS. binometers, or instruments for estimating the 

amount of hemoglobin in the blood, are in vogue 

at the present time, of which the most reliable for general clinical 

use are those devised by Dare, by von Fleischl, by Oliver, and by 

Gowers. The hemometer of von Fleischl has been the general 





favorite for a number of years, both in this country and on the 
Continent, but in America, at least, Dare's hemoglobinometer 
is rapidly supplanting it; in England it has been supplanted to 
some extent, first by Gowers' hemoglobinometer, and in recent 
years by the hemoglobinometer lately invented by Oliver. The 
instruments of von Fleischl, Gowers, and Oliver are based upon 
a similar principle, that of measuring the depth of color of the 
diluted blood by a standard color scale of varying intensity, the 
gradations of which correspond to different hemoglobin values; 
that of Dare uses undiluted blood. 

With this instrument a thin film of undiluted 
blood is brought into direct comparison with a 
standard semicircular wedge of tinted glass 
ranging in color from a claret red at the thickest 
part to a pale pink at the thinnest. The in- 
strument consists of the following parts: (i) A capillary blood 

chamber, constructed of 
two rectangular plates of 
polished glass, the op- 
posed surfaces of which 
are so ground that, when 
clamped together in a 
metal bracket, a shallow 
compartment holding a 
thin film of blood is 
formed. One plate is 
made of transparent, the 
other of opaque, glass, 
the latter being next to 
the source of light, in 
order to soften its glare, 
when the instrument is in 
use. The metal bracket 
of the blood chamber is 
adjusted to the stage of 
the instrument in such a 

blood chamber adjusted to stage of instrument, the slip manner that the blood 

of opaque glass, W, being nearest to the source of light; 1__ ._ 

Y, detachable candle-holder; Z, rectangular slot through Him tltS OVCr an aperture 

Sto dfehSS*"" scale indicated n the " m f communicating with a 

camera tube screwed to 

the opposite side of the rubber case. (2) A graduated color standard 
made of a semicircle of glass tinted with Cassius' ''golden pur- 
ple," and thinning out like a wedge with various depths of color 
corresponding to the tints of fresh blood containing different 


R, Milled wheel acting by a friction bearing on the 
rim of the color disc; S, case inclosing color disc, and 
provided with a stage to which the blood chamber is 
fitted; T, movable wing which is swung outward during 
the observation, to serve as a screen for the observer's 
eyes, and which acts as a cover to inclose the color disc 
when the instrument is not in use; U, telescoping camera 
tube, in position for examination; V, aperture admitting 
light for illumination of the color disc; X, capillary 


percentages of hemoglobin. It is mounted upon a disc adjusted 
in the frame of the instrument, so that it may be revolved to 
bring various portions of its surface over an aperture directly 
alongside of the one through which the blood film is visible. 
A scale, read from the outside of the instrument, indicates in 
units the hemoglobin percentages from 10 to 120. (3) A hard 
rubber case incloses the color standard when the instrument is 
in use, the disc upon which the standard is mounted being re- 
volved by turning a small milled wheel acting upon the rim of 
the disc by a friction bearing. To one side of the case a telescopic 
camera tube, fitted with an eye-piece, is attached, while on the 
opposite side a stage furnishes support for the blood chamber, 
back of which a candle, held between a pair of spring clips, is 
adjusted. Two apertures of equal di- 
ameter, placed side by side on the same 
level, transmit the light of the candle 
through the blood film and the color 
standard to the field of vision inclosed 
by the camera tube. By reference to 
the accompanying diagram (Fig. 5) it 
will be seen that the light of the can- 
dle, J, equally illuminates the blood 
film inclosed between the two rect- 
angular glass plates, O and P, and the 
edge of the color standard, L, mounted 
upon the glass disc, K. The differences 
in the two colors are visible through 
the two apertures, M and M', commu- 
nicating with the camera tube, X. 
By revolving the disc the tint of the 
color standard may be altered until it matches that of the blood 

Method 0} Use. The instrument is prepared for use by first 
swinging outward the movable screen which serves as a cover for 
the case. The two apertures overlying the blood film and the 
color scale are thus brought into view, the direct light from the 
candle being shaded from the observer's eyes by the intervening 
screen. The camera tube and the candle-holder are then fitted 
to their attachments on opposite sides of the instrument, and a 
candle adjusted so that the surface of its "wick end" is just on 
a level with the top of the spring clips. 

The blood chamber is filled by touching its edge to the side of 
a rather large drop of blood as the latter flows from the punc- 
ture, so that the blood at once flows into and fills, by capillary 
force, the shallow compartment between the pair of glass plates. 



As soon as this occurs, any excess of fluid which may have adhered 
to the outer surface of the blood chamber is carefully wiped 
away, and the latter is slipped into the tongue, which holds it 
in position on the stage of the instrument. 

The candle having been lighted, the observer holds the instru- 
ment as a field glass, and compares with one eye the colors of 
the blood film and the standard disc which are seen side by side 
in the field of vision limited by the camera tube. The disc is 
made to revolve by making short, quick turns with the milled 
wheel until the two colors are identical, and the hemoglobin per- 
centage indicated by the scale is then noted. 

The color comparisons need not be made in a darkened room, 
although the observer should avoid facing the direct sunlight, 
and, in order to exclude reflected light, should hold the instru- 
ment against a dark surface, such as a black coat-sleeve. When 

the observation is 
completed, the tw r o 
glass plates of the 
blood chamber are 
removed from the 
bracket by loosening 
the screw which holds 
them in position. 
They are then cleaned 
with water and with 
acid alcohol, dried, 
polished, and replaced 
in the bracket. The 
various parts of the instrument, when detached, fit into a small 
leather carrying-case. 

The chief advantage of Dare's instrument lies in the fact that 
dilution of the blood is not required, and therefore errors due to 
incorrect measuring and dilution of the blood, which must be care- 
fully guarded against in the older hemoglobinometers, are entirely 
eliminated. A film of whole blood also gives a relatively deep 
and definite color, which may be judged with greater ease and 
accuracy than the paler and more indefinite tint of a blood solu- 
tion. It is also obvious that errors due to the turbidity of an 
aqueous solution of leukemic blood are avoidable by the use of 
an undiluted film. Coagulation of the film does not occur with 
sufficient rapidity to constitute a source of error, since the test 
may be completed within a few seconds after the blood has been 

Three years' constant use of this hemoglobinometer in the 




Jefferson Medical Clinic has proved it the most accurate, sim- 
ple, and convenient instrument for clinical purposes. Its read- 
ings closely correspond to those of Oliver's hemoglobinometer, 
and average somewhat higher than those of the von';Fleischl 
instrument. The color standard of Dare's apparatus, being 
wedge-shaped and therefore gradually blending the tints, is open 
to the same criticisms w^hich have 
been urged against the scale of 
the hemometer. 

With this in- 
strument the 
color of a fixed 
volume of blood 



in an aqueous 

solution of a definite strength is 
compared with the color of a 
movable glass wedge, tinted with 
Cassius' "golden purple." The 
hemometer consists of the follow- 
ing parts: 

(i) A tinted glass wedge, the 
thickest portion of which is of a 

deep pink color, and the thinnest portion almost colorless, with 
every intermediate color gradation between the two extremes. 
It is mounted in a metal frame provided with a scale, graduated 
at every five degrees from o to 120, the former corresponding to 
the thinnest, and the latter to the thickest, part of the wedge. 
The metal frame is grooved so that it fits beneath (2) a small 




stage, in which it may be moved backward and forward by turn- 
ing a milled wheel. In the center of this stage there is a circu- 
lar opening through which the light of a candle is reflected by a 
disc of calcium sulphate, mounted on the pillar supporting the 
stage, like the mirror of a microscope. Back of this opening there 
is a small oval slot through which the scale of the underlying 


tinted wedge is visible, when the latter is adjusted to the stage. 
(3) A mixing chamber, consisting of a short metal tube closed 
at the bottom by a disc of glass and divided into two equal com- 
partments by a vertical partition, fits accurately over the circular 
opening in the stage. When properly adjusted to the latter, the 
vertical partition exactly coincides with the upper edge of the 
underlying tinted wedge, so that the upper compartment of the 
chamber is illuminated by the dull white light from the reflector, 
while the lower compartment receives the color of the tinted 
wedge. (4) A capillary pipette mounted in a short metal handle, 
used for making the blood dilution. A single pipetteful of normal 
blood mixed with sufficient distilled water to fill exactly one of 
the compartments of the mixing chamber gives a solution which 
matches the color of the tinted wedge opposite the mark 100. 
(5) A small, fine-pointed glass dropper, used for filling the com- 
partments with water. 

Method of Use. As a preliminary step, each compartment of 
the mixing chamber is filled about one-quarter full of distilled 
water by means of the glass dropper, to one end of which a rubber 
cap has been fitted. A puncture having been made, as previously 
directed, a measured volume of blood is collected by bringing one 
end of the capillary pipette lightly in contact with the blood drop 
as it oozes from the wound, so that the tube is instantly filled 
with blood, by capillary force. No difficulty will be experienced 
in quickly filling the tube if it is applied horizontally to the side 
of the blood drop, rather than vertically to its summit, care being 
observed not to immerse the end too deeply. It is needless 
to add that the interior of the tube must be absolutely clean 
and dry, to insure which a very fine needle and thread may be 
passed through it just before using. As soon as the pipette is 
filled, every trace of blood must be removed from its outer surface 
and the precaution taken to see that the column of blood is exactly 
flush with the ends of the tube, being neither bulged out nor de- 
pressed. The blood is then washed into one of the compartments 
of the mixing chamber, by forcing a stream of distilled water 
through the pipette by means of the glass dropper, this rinsing 
being repeated until it is certain that every trace of blood has been 
removed. The preceding steps must be carried out quickly, in 
order to avoid errors arising from coagulation of the blood. The 
blood and water in the compartment are now thoroughly mixed 
by stirring with the handle of the pipette until the color of the 
solution is diffused uniformly, after which water is added, drop 
by drop, to each compartment until they are both filled exactly 
to their brims. In doing this, no water must be spilled on the 


thin edge of the vertical partition, for should this occur, it may 
cause an overflow of the liquid from one compartment to the 
other, and thus alter the strength of the blood solution. If the 
latter should appear turbid or muddy, as it sometimes does with 
leukemic blood, a few drops of a weak aqueous solution of potas- 
sium hydrate may be added to the diluent as a preventive of 
this change. The addition of a little ether will clear the solu- 
tion if the turbidity is due to the presence of fat. 

Having carried out the preceding steps, the mixing chamber is 
adjusted over the circular opening in the stage of the instrument, 
so that the compartment containing the blood solution is upper- 
most, overlying the semicircle illuminated by the clear white 
light ; while the compartment filled with water fits over the semi- 
circle, which receives the tint of the underlying glass wedge. 
The remainder of the test, the comparison of the color of the 
two compartments, must be completed by artificial light, prefer- 
ably by candle-light. Moderately bright illumination is better 
than a strong glare, for the latter interferes seriously with the 
accurate determination of delicate color differences. By means 
of the milled wheel the tinted glass wedge is moved backward 
and forward until its color precisely corresponds to that of the 
diluted blood. When this occurs, the percentage of hemoglobin 
is read off from the scale visible through the oval slot in the stage 
of the instrument. 

While making the color comparison the observer should stand 
facing one end of the glass wedge (not the milled wheel), so that 
the partition between the two compartments of the mixing cham- 
ber is on a line with the vertical axis of his eye's, the distance from 
the latter to the top of the stage of the instrument being about 
ten or twelve inches. Gross errors may be avoided if the observa- 
tion is made with one eye and if the same eye is habitually used, 
since the two eyes may differ radically in their sensitiveness to 
color impressions. It is important to decide the color differences 
as quickly as possible, for prolonged examination rapidly dulls 
one's color perception, and creates uncertainty as to the proper 
reading. It is a good plan first to bring into the field of vision 
the darkest portions of the wedge between the figures 100 and 
120 of the scale, and then, by short, sudden turns of the milled 
wheel, to produce abrupt color contrasts of from 5 to 10 degrees 
at each turn, until the two tints approximately correspond. 1 
When this point is reached, the eye should be rested for a few 

1 It is important to bear in mind the fact that the judgment of color differences 
is much easier if marked contrasts in color value are made, than if a gradual blend- 
ing of the two tints is attempted, by slowly moving the wedge past the visual field. 


moments, and then, by a succession of shorter turns, .the wedge 
is again swept to and fro until the colors appear identical. In 
the average instance an error of about 5 degrees must be antici- 
pated in spite of every precaution to insure accuracy. 

Note that the septum between the two halves of the blood compartment is at right angles to 
the horizontal axis of the observer's eyes. A cylinder of paper may be fitted over the blood com- 

partment, to serve as a camera tube. 

In cases in which low hemoglobin percentages (30 per cent, 
or less) are suspected, it is essential to use two or three pipette- 
fuls of blood in making the dilution, dividing the percentage indi- 
cated by the instrument by two or three, as the case may be. 


This precaution effectually removes the objection which has been 
urged against this instrument on account of its inaccuracies in 
the determination of low hemoglobin percentages. Another 
criticism of the von Fleischl instrument has been made on the 
ground that, since the length of the tinted wedge visible through 
the compartment of the mixing chamber includes a color range 
of 20 per cent., it is impossible for one to select a single point in 
the center of this color for comparison with the even, diffuse tint 
of the blood solution. This objection may be overcome to a 
great extent by using a metal diaphragm, provided with a slit 
one-eighth of an inch in width, which is placed over the glass 
disc at the bottom of the compartments, to limit the field of 
vision. Adjusted so that the slit crosses at right angles the par- 
tition separating the two colors, the use of this device cuts down 
the field of observation to a portion of. the glass wedge corre- 
sponding to about 2.5 degrees on the scale. 

The hemoglobin percentages indicated by this instrument 
appear to be low for the blood of the average healthy American, 
since it is more common to obtain readings of from 90 to 95 than 
of the arbitrary standard 100, in persons in whom there is no 
good reason to suspect subnormal hemoglobin values. In in- 
struments of recent manufacture, however, this fault is largely 

In order to exclude the light of the candle from the field of 
vision while making the color comparison, it is customary to 
use a tube of cardboard or stiff paper, which is slipped over the 
mixing chamber and rests upon the platform of the instrument. 
This sort of a device answers very well when the examination is 
made in a darkened room, as, for example, at a patient's residence. 
In hospital work, however, the inconvenience, sometimes con- 
siderable, of being compelled to carry the diluted blood some 
distance from the bedside to a dark room may be avoided by the 
use of a light-proof box, which may be conveniently carried from 
ward to ward, so that the test may be completed at the bedside 
(Fig. n). A box of this kind should measure sixteen inches 
in height by twelve inches in length and in width, being fitted 
with a hinged door which may be fastened by a simple catch, 
and provided with a circular opening through which the milled 
wheel of the hemometer projects when the door is closed. A 
metal camera tube, flanged at the upper extremity for the obser- 
ver's eye, pierces the top of the box and communicates inside 
with the mixing chamber of the hemometer. The tube fits 
loosely in a circular opening in the top of the box, so that it may 
readily be raised and lowered; its diameter is a trifle greater 


than that of the mixing chamber, around which it should fit 
snugly when lowered into position; and its length is governed 
by a fixed collar outside the box, which prevents it from slipping 
and jarring the instrument. Wooden guides, such as are used 
for securing a microscope in its box, are provided to receive the 

The door of the box is closed and the color comparison made through the camera tube. 

Xote. Reichert, at the suggestion of Miescher, has introduced a modification 
of the original von Fleischl hemometer, designed to increase the accuracy of the 
test, by making it possible, by definite dilution of the blood, to select that part 
of the tinted wedge which is best adapted for the examination of any particular 
sample. This innovation was prompted by the discovery that the intermediate 
portions of the wedge are better adapted for obtaining accurate readings than the 
terminal parts. The principal modification of the new hemometer consists in 
the substitution, for the original capillary blood pipette, of a special mixing pipette, 
similar to a mclangeur, graduated so that the blood may be diluted i : 200, i : 300, 
and i : 400 times. A table supplied with the instrument translates the combined 
results of the dilutions and the figures indicated by the scale on the wedge into 
absolute hemoglobin percentages. The instrument is also supplied with mixing 
chambers of different depths, and with a diaphragm designed to limit the field of 
vision. The writer has had no practical experience with the Miescher-Fleischl hem- 
ometer, but an examination of the instrument justifies the belief that its elaborate- 
ness renders it undesirable for general clinical work. Its cost ($50.00) is also a 
bar to many. 

horseshoe base of the hemometer, holding it firmly in such a 
position that when the camera tube is lowered into place the 
milled wheel of the instrument projects through the opening in 
the closed door. The interior of the tube and of the box is painted 
a dull black. A candle is placed in position on the floor of the 


box, on a line with the "mirror" of the instrument. In using 
this device first the candle within the box is lighted, and the 
hemometer base is slipped into place between the wooden guides. 
The blood dilution having been made in the usual manner, 
the mixing chamber is then set upon the platform of the instru- 
ment, and the camera tube, which has been raised to allow this 
to be done, is lowered until it telescopes around the mixing cham- 
ber and rests firmly upon its collar. The door of the box is 
now closed, and the two compartments are brought into their 
proper positions over the glass wedge by turning the camera 
tube from the outside of the box, the observer meanwhile noting 
the result by looking through the flanged extremity of the tube. 
This accomplished, the projecting wheel of the instrument is 
turned to and fro until the colors of the two compartments are 
the same, when the door is opened and the percentage read off 
from the scale of the hemometer. Care must be observed to 
see that the exterior of the mixing chamber is perfectly dry, 
for if any moisture collects between its outer surface and the 
inner surface of the camera tube, the contents of the compart- 
ments may be disturbed and serious errors result. As the open- 
ing in the door of the box is covered by the hand with which 
the milled wheel is turned, sufficient light to interfere with the 
test cannot leak in at this situation. 

With a light-proof box of this sort it is possible accurately 
to carry on hemoglobin estimations in the brightest daylight, 
which may be entirely excluded from the instrument, while the 
observer's field of vision is limited to the two semicircles illumi- 
nated by the candle burning within the box. 

With this instrument the principles of Lovi- 
OLIVER'S bond's tintometer are applied to the quantitative 
HEMOGLOBIN- estimation of hemoglobin, the color of- a blood 
OMETER. solution of a definite strength being compared, 
by light reflected from a dead white surface, with 
a series of tinted glass standards which constitute a progressive 
color scale. Thus, a series of fixed, definite tints is provided, 
each of which accurately corresponds to the specific color curve 
of progressive dilutions of normal blood, this having been de- 
termined individually by means of the tintometer. Two sets 
of color standards have been devised: one for daylight readings 
and one for observations by candle-light, the latter being prefer- 
able on account of the greater delicacy of its readings. Oliver's 
complete apparatus consists of : (i) A capillary blood measure, 
made of heavy glass tubing, and having a capacity of 5 
The end to be presented to the blood drop, in filling the 


measure, is tapered to a blunt point and highly polished. (2) 
A mixing pipette, provided with a short rubber tube which fits 
over the tapered end of the blood measure, while rinsing out 
the blood from the latter into the (3) standard blood cell, which, 
when filled exactly to the brim with distilled water in which one 
measureful of blood has been dissolved, yields a blood solution 
of approximately one per cent. When filled, the cell is covered 
with a glass slip provided for this purpose. (4) A standard color 
scale, consisting of 12 tinted glass discs, mounted in two series, 
and corresponding to hemoglobin percentages ranging from 10 
to 1 20. (5) A set of riders, or squares of tinted glass, used for 
determining the intermediate degrees of color between the deci- 
mals indicated by the fixed tints of the scale. For ordinary clin- 
ical work two riders are sufficient, which, when laid over the discs 
of the standard scale, read 2.5 and 5 degrees respectively on its 
upper half, but double this amount on the lower half. For physi- 
ological observations requiring readings in units a set of nine 
riders is supplied. (6) A collapsible camera tube through which 
the color comparisons are made. 

Method of Use. In making hemoglobin estimations with 
Oliver's apparatus, first the capillary measure is filled with blood 
by the method directed for filling the pipette of the von Fleischl 
instrument. The rubber nozle of the mixing pipette, previously 
filled with distilled water, is then adjusted over the polished end of 
the blood measure, and the blood washed into the standard cell by 
forcing through the water, drop by drop. As soon as all the 
blood contained in the bore of the measure has been thus washed 
out into the cell, the rubber nozle of the pipette is removed, and 
the handle of the measure used as a stirrer to mix the blood 
solution, more water being added in single drops, from time 
to time, until the cell is accurately filled. The blue cover-glass 
is then adjusted, with the result that, if the cell has not been 
overfilled, a small air-bubble forms on the surface of the liquid. 
The blood cell, filled in this manner with a blood solution of 
definite strength, is now placed by the side of the standard scale, 
opposite the tinted disc to which it corresponds most closely, 
the eye readily recognizing its approximate position. More accu- 
rate matching of the two colors is made with the aid of the 
camera-tube, the cell being moved from disc to disc in an 
endeavor to match exactly the color of the blood solution by one 
of the standard tints of the scale. If this is successful, the hemo- 
globin percentage indicated by the disc is read off, and the obser- 
vation is completed. But if it happens that the tint of the blood 
solution is obviously deeper than a certain disc, but paler than the 


5 1 

one immediately above, the cell is kept alongside the lower of the 
two, over which a rider is adjusted in order to deepen its color, 
while the square of white glass is placed over the cell, so as to 
compensate for the thickness of the rider. If, now, the colors 
correspond, the final reading is ascertained by taking the percent- 
age of the disc plus the value of the superimposed rider. If the 
color of the blood solution is darker than that of any one of the 
standard discs, but paler than the disc plus a rider, the mean 
average of the two is taken as the final reading; similarly, if the 


color of the blood solution is darker than a certain disc plus a 
rider, but paler than the disc immediately above, the values of 
the two must be averaged. An error of two per cent, is un- 
avoidable, even in the hands of a skilful observer. 

During the observation the candle should be placed three or 
four inches from the end of the color scale, being adjusted so that 
the flame is on a line with the opposed sides of the cell and of the 
scale, thus illuminating both with equal intensity. The positions 
of the candle and of the apparatus are shown in the accompanying 
illustration (Fig. 12). Small-sized candles, such as are used 

5 2 


for decorating Christmas trees, furnish a flame of the proper de- 
gree of brilliancy, the candle of ordinary size giving too intense a 
light. Total exclusion of daylight is not necessary, so that the 
observation may be made in the corner of a partly darkened 
room, as, for example, behind a closet door or some other similar 
shield against direct rays of light. 

Oliver's hemoglobinometer is a trial to the patience of one 
who has habitually used the Dare or the von Fleischl, and it 
takes some time to become accustomed to it after having worked 
with the comparatively simple color comparisons of other instru- 
ments. Its accuracy is undeniable, although for clinical work 
a simpler apparatus is to be preferred. 

This instrument, which for 
GOWERS' many years has been popular 
HEMOGLOBIN- in England and is used to 
OMETER. some extent in this country, 
consists essentially of two small 
flattened tubes of equal diameter, which, when 
in use, are fixed upright and parallel to each 
other in a small wooden support furnished for 
this purpose. One tube contains glycerin jelly 
colored with picrocarmin to correspond to the 
tint of a i : 100 solution of normal blood (or 
20 of blood in 2 c.c. of water), this 
being taken as the standard with which the 
blood solution contained in the second tube 
is compared. The second tube is provided 
with a scale graduated in units from 5 to 120, each degree of 
which equals the volume of blood required for the test. Twenty of normal blood, dissolved in sufficient distilled water to 
fill this tube to the 100 mark on the scale, give a solution which 
corresponds to the tint of the standard tube. The special capil- 
lary pipette used for measuring the blood is graduated at 10 and 
at 20, and fitted with a bit of rubber tubing and mouth- 
piece for filling it by suction. 

Method of Use. The technic of hemoglobin estimations with 
Gowers' apparatus is extremely simple. Having made the punc- 
ture in the usual manner, the blood is sucked up the caliber of 
the capillary pipette until the mark 20 is reached, and then im- 
mediately blown out into the graduated tube, into which a few 
drops of distilled water have previously been placed, in order to 
insure instantaneous solution of the measured amount of blood. 
All traces of blood which may have adhered to the bore of the 
capillary pipette are removed by filling it several times with water, 


the rinsings being added to the mixture of blood and water in the 
tube. During the preceding steps the usual precautions must be 
observed to wipe all surplus blood from the outside of the pipette 
before expelling its contents, and to measure the blood rapidly, so 
as to guard against errors arising from rapid clotting. Distilled 
water is now added, drop by drop, to the mixture in the tube 
until the color of the blood solution exactly corresponds to that 
of the picrocarmin standard, the contents of the tube being mixed 
between each addition by rapidly reversing it two or three times, 
w r ith its open end closed by the thumb. The drop or two of 
liquid adhering to the thumb should be wiped off against the wall 
of the tube, so that it may drain back into the liquid. When the 
tints of both tubes are precisely similar, the division of the scale 
to which the diluted blood reaches is read off, to express the per- 
centage of hemoglobin in the specimen under consideration. 

In comparing the colors, which is done by daylight, the tubes 
should be held against a sheet of white paper, or, as suggested 
by Gowers, between the eye and a window, and viewed at such 
an angle that their adjoining edges appear to overlap, thus cutting 
off the vertical streak of white light visible between them should 
this precaution be neglected. Owing to the diagonal position in 
which the two tubes are adjusted in their support, the proper 
angle to produce this effect may readily be determined. 

The chief drawback to the use of this instrument is the likeli- 
hood of overdiluting the blood after it has been mixed in the 
graduated tube, the occurrence of this accident necessitating, of 
course, a repetition of the entire operation. It is not always 
easy to decide just when sufficient water has been added to the 
blood to bring its color down to that of the standard tint, since 
one must depend solely upon a gradual weakening of the tint of 
the blood solution, and this is much more difficult than to com- 
pare a definite blood color with a sliding scale or with a series 
of discs. The instrument may be regarded as accurate within 
two or three per cent, for hemoglobin percentages above ten, 
below which figure it is impossible to distinguish a difference 
between the tints of the two tubes. This source of error, how- 
ever, is too remote a possibility to detract from the instrument's 
practical value. A real source of error in Gowers' instrument is 
the deterioration with age of the picrocarmin color standard, 
with a consequent change in its tint. 1 

1 Haldane (Jour. Physiol., 1901, vol. xxvi, p. 497) has modified Gowers' 
instrument by substituting for the picrocarmin jelly a one per cent, solution of 
carbonic oxid hemoglobin, which remains stable indefinitely. Using Gowers' 
technic, the blood is first partly diluted and then charged with illuminating gas, 


A simple method of approximately determin- 
TALLQUIST'S ing hemoglobin percentages without the aid of 

METHOD. a special instrument has recently been devised 
by Tallquist, 1 the procedure consisting, in brief, 
in allowing a drop of blood to soak into a bit of filter-paper and 
comparing with the naked eye the color strength of the stain 
with a series of printed standard tints of known value. The lat- 
ter are arranged as a scale of ten different colors, corresponding 
to the colors of stains produced by bloods having hemoglobin 
values ranging from 10 to 100 per cent., the latter being regarded 
as the normal. 'A lithographed copy of the color standard ac- 
companies Tallquist's original article. Dealers in laboratory 
supplies also furnish a small book containing the color scale and 
a supply of standard absorbent paper. The test is made in the 
following manner: A drop of blood, large enough to make a 
stain about 5 or 6 mm. in diameter, is caught in the center of a 
piece of white filter-paper, care being taken in collecting it to 
apply the paper to the exuding drop in such a manner that the 
blood soaks in very slowly, and thus produces a stain which is 
evenly colored throughout. Perfectly white filter-paper, having 
a smooth surface and of a thickness corresponding to about 55 
leaves to the centimeter, should be used for the test. The blood 
stain thus made is pressed lightly against a pad of filter-paper, 
and then compared, by direct daylight, with the series of standard 
tints, the figure opposite to the tint which the stain most accurately 
matches being read off, to indicate the percentage of hemoglobin 
in the specimen under examination. The comparison must be 
made immediately after the stain loses its humid gloss, since 
blood soon changes its color after exposure to the air. 

This direct method of hemoglobin testing is, of course, only 
approximate, at the best, and cannot be expected to furnish re- 
sults comparable in point of accuracy with those to be obtained 
by any of the instruments just described. It may, however, be 
employed to excellent advantage when a hemoglobinometer is 

by means of a special fitting to be attached to a gas-burner. The oxyhemoglobin 
of the blood is thus converted into carbonic oxid hemoglobin, the color of which 
is comparable to that of the standard solution. After having been charged with 
gas, the blood is mixed by repeatedly inverting the tube, the open end of which 
is closed by the thumb, after which the dilution is proceeded with until the colors 
match, when the final reading is made. The inventor's claims for the accuracy 
of his instrument have been substantiated by Horder (Lancet, 1903, vol. i, 
p. 1305). In ten estimates by different observers it was determined that the 
possible errors with Haldane's instrument averaged 1.9 per cent., and with the 
hemometer, 4.25 per cent. Haldane's device, despite its accuracy, obviously is 
better adapted to physiological research work than to clinical hematology. 
1 St. Paul Med. Jour., 1900, vol. ii, p. 291. 


not at hand, or in certain cases in which only a rough estimate of 
the amount of coloring matter of the blood is sought. Tall- 
quist, who has tested his method, under the control of the he- 
mometer, in his clinic at Helsingfors, claims that the limit of error 
generally does not exceed ten per cent. 

Here may be mentioned Haig's blood decimal card, devised 
for roughly estimating the color index of the blood. It consists 
of a series of four different colors, scaled 0.80, 0.60, 0.40, and 
0.20, respectively. By matching one of these colors with the 
color of the patient's gums or tongue an approximate idea of the 
individual's color index is obtained. Haig's device may serve 
for a hurried examination, but it is obviously too crude to give 
accurate results. 


Of the various instruments used for count- 
METHODS. ing the blood corpuscles, the hemocytometers de- 
vised by Thoma and by Gowers are most gener- 
ally employed at the present time, the former being used almost 
to the exclusion of the latter everywhere except in England, 
where Gowers' apparatus has many firm adherents. Durham, 
by adapting and modifying a number of the details of the older 
instruments, has succeeded in devising an improved form of hemo- 
cytometer which possesses many advantages over the original 
models, being of simple construction, accurate, and comparatively 
inexpensive. The method of making the estimate, which is es- 
sentially the same with all three of these instruments, consists, 
briefly, in first diluting the fresh blood in definite proportions with 
some indifferent preservative flifid, and in then counting, under 
the microscope, the number of corpuscles in a drop of the diluted 
blood, the latter being contained in a small glass cell on the floor 
of which is ruled a series of micrometer squares of certain dimen- 
sions. The cubic contents of the cell and the degree of the blood 
dilution being known, the number of corpuscles counted in any 
given number of these squares may be taken as a basis for cal- 
culating the total count of corpuscles to the cubic millimeter 
of blood. 

Strong and Seligmann dispense with a special counting cham- 
ber, and enumerate the cells in a measured quantity of blood 
diluted in definite proportions with a diluent-stain and mounted 
as a permanent dry specimen. 

Oliver has devised an instrument with which the number of 


erythrocytes may be estimated by means of their optical effect, 
without the use of the microscope. 

Diluting fluids for use with the hemocytom- 
DILUTING eters of Thoma, Gowers, and Durham should 
FLUIDS. be of such a composition that when mixed with 
the fresh blood they preserve unaltered the form 
of the corpuscles. This requirement being met, the examiner 
may choose from the numerous formulas in current use the one 
which best suits his individual preference. Among the most sat- 
isfactory solutions used for this purpose the following may be 
mentioned : 


Methyl-violet, 5 B 0.025 

Sodium chlorid i .o 

Sodium sulphate 8.0 

Neutral glycerin .* 30.0 

Distilled water 160.0 


Methylene-blue o.i 

Sodium chlorid 1.2 

Neutral potassium oxalate 1.2 

Distilled water 300.0 

For general clinical work no better formulas have ever been 
suggested than the preceding two. Both solutions act as excel- 
lent preservative fluids, and each contains just sufficient quantity 
of a basic anilin dye to stain the leucocytes with great distinct- 
ness, so that they may readily be differentiated from the erythro- 
cytes, which remain uncolored. 


Mercuric chlorid 0.25 

Sodium chlorid 0.5 

Sodium sulphate 2.5 

Distilled water 100.0 

Oliver specifies this solution as the diluent invariably to be 
employed with his instrument, but it may be used also with the 
other forms of hemocytometers, although with less satisfaction 
than the formulas first mentioned. 

Among the simpler diluting fluids, all of which are depend- 
able, are solutions in distilled water of common salt (0.7 per cent.), 
of potassium bichromate (2.5 per cent.), and of sodium sulphate 
(5 per cent.), to any of which about 0.5 per cent, of an alcoholic 
solution of methyl-violet may be added, in order to stain the 
leucocytes, and thus to facilitate the counting. 



An aqueous solution of acetic acid, varying in strength from 
0.3 to 0.5 per cent., which destroys the erythrocytes and at the 
same time renders more conspicuous the leucocytes, has been 
recommended by Thoma as a diluent in counting the latter cells, 
by means of his special pipette. 

Turk recommends the use of a one per cent, solution of acetic 
acid containing one per cent, of gentian-violet, 
in order to recognize the different forms of leu- 
cocytes, as well as to count their total number. 
This method is, however, only approximate, and 
should not replace the examination of the dried 
stained film. (See p. 75.) 

As spores are liable to develop and precipi- 
tates to form in all the above-mentioned solu- 
tions, they should always be filtered before 
using and kept in tightly corked bottles. 

The Thoma-Zeiss hemo- 
THE THOMA- cytometer, which is to-day re- 
ZEISS HEMO- garded as the standard instru- 
CYTOMETER. ment for blood counting, con- 
sists of two graduated capil- 
lary pipettes for diluting and mixing the blood, 
and a counting chamber in which a measured 
volume of diluted blood is placed for the pur- 
pose of counting the corpuscles under the 
microscope. One of the pipettes is intended 
for counting the erythrocytes, and, for con- 
venience sake, may be termed the erythrocy- 
tometer; while the other pipette, which is used 
for counting the leucocytes, may be called the 
leucocytometer. It is not, however, necessary 
to purchase both pipettes, as supplied with the 
complete apparatus, since both erythrocytes and 
leucocytes may be counted accurately with the 

The erythrocytometer consists of a heavy 
glass capillary tube, the lumen of which is ex- 
panded near the upper end into a bulb contain- 
ing a small cubical glass bead, which serves as a stirrer. The 
lower end of the tube is ground to a blunt point, and to the upper 
end is fitted a short bit of rubber tubing capped by a bone mouth- 
piece for filling the tube by suction. A scale is enameled into 
the glass wall of the pipette, the three main divisions of which 
are indicated by the figures 0.5, i, and 101, the first two grada- 


FIG. 14. THE 


tions being below, and the latter one above, the bulb; the lower 
portion of the tube is further graduated in tenths, by cross lines, 
from o.i to i. If blood is drawn up in the pipette to the mark 
i, and the diluent added until the mark 101 is reached, the blood 
is thus diluted one hundred times; or if the blood is drawn up 
only to the mark 0.5, and the diluent added as before, a two- 
hundred-fold dilution is obtained. 

The leucocytometer is a capillary tube similar to the former, 
but having a larger lumen, so that lower dilutions are obtained 
with it. If blood is drawn up to the mark i, and the diluent 
added until the mixture reaches the mark n, the blood is diluted 
ten times; or if the blood column reaches the mark 0.5, with 
the same addition of diluent, the dilution thus made is twenty- 
fold. In the latest model of this pipette the lower end tapers 
to a fine point, the diameter of the lumen thus gradually decreas- 
ing as the extreme tip is approached. The chief object of this 


modification is to prevent accidental leaking out of the column 
of blood when the tube is held vertically, while sucking up the 
diluting fluid an accident difficult to avoid with the old-style 
pipette having a large lumen from tip to bulb. 

The counting chamber (Fig. 15) consists of a heavy glass slide 
in the center of which is cemented a square glass plate provided 
with a circular opening which fits around a ruled disc ; the diameter 
of the latter being slightly less than that of the opening in the 
surrounding plate, a shallow, narrow gutter is thus formed be- 
tween the two. The surface of the ruled disc is exactly y 1 ^ mm. 
below the level of the glass plate by which it is inclosed, so that 
a chamber of this depth is formed when both are superimposed 
by a cover-glass having an absolutely plane surface, two such 
covers being furnished with each instrument. An ordinary cover- 
glass should never be used, for, owing to the unevenness of its 
surface, a deviation from the standard in the depth of the under- 


lying chamber must necessarily result. When an objective having 
an extremely close "working distance" is employed, the special 
hollow cell cover-glass made by Zeiss will prove useful. 

The central part of the disc's surface is divided, by micro- 
scopical diamond-rulings, into 400 small squares, each of which 
has an area of T ^ 7 sq. mm., these small squares being grouped 
into sets of sixteen by a series of vertical and horizontal double 
rulings bisecting each fifth column of squares (Fig. 16). The 
cubic contents of each small square, when the cover-glass is 
adjusted, is j-^nrg-, since they measure individually y 1 ^ 
by -^ by -^ mm. In Zappert's modified ruling of the Thoma- 
Zeiss counting chamber extra lines have been added so as to 


increase the ruled area of the disc to nine times its original size. 
As illustrated in the accompanying diagram (Fig. 17), the 
surface of the disc is thus divided, by heavy cross-rulings, into 
nine large squares, each equal in area to the central group of 400 
small squares, the whole ruled surface therefore equaling an area 
covered by 3600 of the latter. To simplify the counting, the 
peripheral squares are subdivided into four, each of the latter 
being of the same area as 100 of the small central squares. This 
improved form of counting chamber is an invaluable convenience 
in leucocyte counting, and should be chosen in preference to 
the older model. 1 

1 Other counting chambers, ruled so as to provide a larger number of squares 
in the field, have been devised by Turk, Breuer, Elzholz, and others, but none 
of them possesses any real advantages over Zappert's slide. 


If any difficulty should be experienced in distinguishing the 
ruled lines under the microscope, they may be made more con- 
spicuous by blackening them with a little soft lead-pencil dust 
placed on the surface of the disc and thoroughly rubbed in with 
the ball of the finger, the excess being wiped off and the disc 
polished with a bit of lens-paper or a soft handkerchief. 

Counting the Erythrocytes. Having made the puncture, as 
already described, the point of the erythrocytometer is plunged 
into the blood drop as it flows from the wound, and, by making 
gentle, uniform suction, a column of blood is drawn up the capil- 
lary tube exactly to the mark 0.5. The point of the instrument 
is then wiped perfectly dry, and immediately dipped into the dilut- 


ing fluid, which is drawn up the tube in the same manner until 
the mixture of blood and diluent reaches the mark 101 above the 
bulb. While adding the diluent, the pipette should be twisted to 
and fro between the thumb and forefinger, in order that the blood 
and diluent may be mixed, by the whipping about of the glass 
bead, as they fill the bulb; if this precaution is neglected, a por- 
tion of the blood will rise in a distinct layer above the diluent as 
the latter flows into the bulb, and may be drawn, unmixed, into 
the capillary constriction beyond. A more thorough mixture of 
the blood and diluent is now made, the rubber tubing being 
slipped from the instrument, which is then grasped so that its 
ends are closed by the thumb and middle finger, and rapidly 



shaken for about half a minute. By the above steps a mixture 
is made in which the proportion of blood to diluent is i : 200, a 
degree of dilution with which it is most convenient to work in 
the great majority of instances. For two reasons a i : 200, rather 
than a i : 100, dilution is to be preferred in routine work: (i) 
If, as not infrequently happens, the blood column is accidentally 
drawn up -the tube beyond the mark 0.5 in an attempt exactly 
to reach this gradation, it is a simple matter to correct the error 
by gently blowing or shaking the blood column down to the 



proper level; whereas, in attempting to make a i : 100 dilution, 
should the mark i be exceeded, the blood column will almost 
surely escape into the bulb, whence it cannot be blown back again 
into the capillary tube, thus necessitating a repetition of the 
whole operation with a fresh drop after having cleaned and dried 
the erythrocytometer. (2) It is easy to count the corpuscles 
in a i : 200 dilution, since the surface of each ruled square of the 
counting chamber is, as a rule, occupied by not more than half 
a dozen cells; on the contrary, in a i : 100 dilution, except in 
an occasional instance in which there is a striking paucity of cells, 


the field may be so overcrowded with corpuscles that their enu- 
meration is difficult and often inaccurate. 

The next step is to place a drop of the diluted blood in the 
counting chamber, preparatory to counting the corpuscles under 
the microscope. The unmixed diluting fluid in the lower portion 
of the capillary tube is first expelled, by blowing out four or five 
drops, after which the point of the pipette is dried with a soft 
cloth and a small drop of the blood mixture is allowed to fall, 
by force of gravity, exactly in the center of the surface of the 
ruled disc. The cover-glass is then immediately placed in posi- 
tion, and the slide left undisturbed for several minutes, so that 
the corpuscles may settle. The drop placed on the disc should 
be of sufficient size to occupy only its central portion, the object 
being to use just enough of the blood mixture to cover the ruled 
area and exactly to fill in the vertical space between the surfaces 
of the disc and cover-glass when the latter is placed in position. 
If the drop contains air-bubbles, or if it is so large that it over- 
flows into the gutter and perhaps finds its way between the cover- 
glass and the glass plate beneath, errors will result, so that in the 
event of either of these accidents the procedure must be repeated 
with another drop, after having cleaned and dried the cover-glass 
and the counting chamber. Water, and not alcohol or xylol, is 
to be used for this purpose, since the repeated use of chemicals 
will soon dissolve the cement which fixes the disc to the counting 
chamber. In repeating the operation the original technic must 
be rigidly followed i. e., the erythrocytometer must be briskly 
shaken for half a minute or so, and the contents of its capillary 
stem blown out, before placing the new drop in the counting 

In a properly prepared slide concentric rings of color New- 
ton's rings may be seen at the points of contact between the 
cover-glass and the underlying glass plate. If these rings are in- 
visible, or if they do not appear when pressure is made upon the 
cover-glass, it is a sign that the contact between the two glass 
surfaces is not true, this being due to the presence of particles of 
dust or of moisture beneath the cover-glass. Inasmuch as this 
may seriously affect the correctness of the count, it is a safe rule 
invariably to reject a slide in which these color rings are not visible. 

As soon as sufficient time has elapsed for the corpuscles to sink 
to the bottom of the counting chamber about five minutes the 
slide is transferred to the stage of the microscope, which should not 
be inclined, for fear of disturbing the uniform distribution of the 
cells. The field is first brought into focus with a low-power objec- 
tive (a No. 3 objective of Leitz, for example), and the slide moved 


across the stage until the extreme upper left-hand corner of the 
group of small ruled squares is brought into view, when a higher 
power, to be used in counting, is substituted. For this purpose 
the writer is accustomed to use a Leitz No. 6 objective and No. 4 
ocular, which lenses, with a tube length of 155 mm., cut off a 
field occupied by a block of 25 small squares. 

As a basis for the final calculation, the erythrocytes in 400 
small squares, or the entire ruled surface of the old-style disc, 
should be counted, preferably by going over two groups of 200 
squares each in two different drops, rather than by taking the en- 
tire 400 squares in a single specimen. By following this plan the 


The small squares are examined in the order indicated by the arrow, successive blocks of 25 
squares being covered until the required number of cells has been counted. 

count of one drop may be controlled by the count of the other, 
and any discrepancy between the two discovered, for if the dif- 
ference in the counts is striking, a third group of 200 squares 
must be examined in an additional drop, and an average taken 
of the two counts which most closely correspond. 

In order to simplify the process of counting, some routine 
method of examining the ruled area, such as the following, should 
be adopted : Beginning at the upper left-hand corner of the ruled 
disc, the corpuscles in the first 100 small squares are counted, 
the slide being moved from above downward, preferably by the 
aid of a mechanical stage, as the successive groups of squares are 


covered. By employing the magnification to which reference has 
just been made, three shifts of the slide are sufficient to bring 
into the field the requisite number of squares in blocks of 25 
each. Examining each small square in succession, proceed from 
left to right along one row of five, then drop to the next row and 
count from right to left, and continue in the manner illustrated 
by the diagram (Fig. 19) until all the erythrocytes in the first 
group of 100 squares have been counted, the totals of each block 
of 25 squares being noted as they are completed. To avoid repe- 
tition in counting it is necessary to include in the total all the cor- 
puscles which touch the upper and left boundary lines, and to 
disregard those which touch the lower and right boundaries. A 
second group of 100 squares, not immediately adjacent to the 
first, is then inspected in a similar manner, after which the cover- 
glass and counting chamber are washed with water and dried, and 
the operation repeated with a second drop. Thus the 400 squares 
are covered by examining 16 blocks of 25 squares each 8 in the 
first and 8 in the second drop of diluted blood. In a i : 200 
dilution of normal blood this involves the counting of approxi- 
mately from 2400 to 2800 erythrocytes, and gives results which 
are accurate within one and one-half per cent. 

To calculate the number of erythrocytes to the cubic milli- 
meter of blood the following formula is employed: 

A iru* num- 
ber of erythro- 
cytes per c.tnm. 

Number of eryth- ., Degree of dilu- Cubic contents of each Total 

rocytes counted tion (200) square (4000) _ oer O t 

Number of squares counted (400) cytes 

For example, supposing that in the 400 squares of a i : 200 
blood dilution a total of 2500 erythrocytes is counted, the cal- 
culation is made thus : 

.- -. 2500 X 200 X 4000 ., 

(a) -=? = 5,000,000 erythrocytes per 


(6) 2500X2000 = 5,000,000 erythrocytes per 

Counting the Leucocytes. The leucocytes may be counted by 
two different methods: (a) With the erythrocytometer, in the 
same drop of diluted blood in which the erythrocytes are esti- 
mated ; or (6) with the special leucocytometer, as a separate pro- 
cedure. Of the two methods, the former is greatly to be preferred, 
since it is fully as accurate and much more convenient and time- 
saving than the latter. Furthermore, there is an undoubted ad- 
vantage in counting both the red and the white corpuscles in the 
same drop of the blood dilution. 

(a) If the leucocytes are counted with the erythrocytometer, 
the same technic is followed as in determining the number of 


erythrocytes, except that a much larger area of the counting 
chamber must be examined, owing to the comparatively small 
number of leucocytes contained in the i : 200 blood mixture. 
It is necessary, for the sake of accuracy, to count the leucocytes 
in the entire space inclosed by Zappert's ruling, and to repeat the 
count in a second drop, making an area equal to eighteen times 
the ruled space of the old-style counting chamber to be examined. 
If the totals of both counts are approximately the same, their 
combined figures, representing the corpuscles found in a space 
corresponding to 7200 of the small ruled squares, are taken as a 
basis for the final estimate; but if these totals differ widely, a 
third drop is to be examined in the same manner, and, as in 
counting the erythrocytes, an average taken of the two totals 
which are nearest alike. Since in normal blood, in a i : 200 
dilution, each block of 400 small squares contains from 3 to 6 
leucocytes, the examination of the above-mentioned area of the 
counting chamber involves the counting of approximately from 
54 to 108 of these cells an operation which, practically, is not 
nearly so laborious as it appears from the description, being 
easily completed within ten or fifteen minutes in most cases. 1 

As an example of the method of calculating the final estimate, 
supposing that 90 leucocytes have been counted in the area equal 
to 7200 small squares, the blood dilution being i : 200, this for- 
mula is employed: 

90 X 200 X 4000 -s- 7200 = 10,000 leucocytes per 

If the old-style counting chamber is used, the leucocytes in the 
unruled portion of the disc outside of the central block of 400 
squares may be counted with the aid of an eye- 
piece diaphragm, which, when adjusted inside 
the tube of the ocular, cuts off a field exactly 
the size of 100 small squares (Fig. 20). A 
black metal or cardboard disc having a central 
aperture of the proper size will answer just as 
well for this purpose as the more expensive 
and elaborate mechanical eye-piece devised by 
Ehrlich, which is provided with a diaphragm 
having a square opening the size of which is 
regulated by a small lever. Having first counted all the leuco- 
cytes in the 400 small squares, the cells are then counted in 
32 of the diaphragm-fields outside the latter, in order to cover 

Should the leucocytes be decidedly increased, it is unnecessary to cover so 
arce a number of squares. One hundred cells taken as a basis for the calculation 
will give an accurate estimate. 

Fie. 20. OCULAR 


an area of the disc corresponding to the entire ruled surface of the 
Zappert counting chamber. This operation having been repeated 
in a second drop, the totals of both counts are taken as the basis 
for the final calculation, which is made in the manner already 

If one happens to have neither an eye-piece diaphragm nor 
a Zappert counting chamber, the following method of calculat- 
ing the cubic contents of the portions of the disc outside the 
ruled area may be adopted, as advised by Stengel. 1 Using, for 
example, a i-inch objective and a i-inch ocular, the ruled lines 
are brought into focus, and the tube of the microscope drawn 
out until one of the parallel lines of the ruled disc exactly coin- 
cides with either boundary of the field of vision. Assuming that 8 
of these parallel columns, each -^ mm. in width, are included 
in the visual field, the diameter of the latter is therefore T 8 7 , or -| 
mm., and the radius one-half of this figure, T 8 7 , or i mm. The 
area of the field may then be readily determined by multiplying 
the square of its radius by 3.1416. Its cubic contents are 
obtained by also multiplying by y 1 ^ mm., the formula being: 

i X 3 X t o X 3-1416 = 0.0125664, cubic contents of the visual field. 

Having in this manner ascertained the cubic contents of each 
field of vision, the final calculation of the number of leucocytes 
to the of undiluted blood is made by multiplying the total 
number of these cells found in a definite number of fields (for in- 
stance, 50) by the degree of dilution (usually i : 200), and then by 
dividing the cubic contents of each field (0.0125664) multiplied 
by the number of fields examined. The formula for this calcula- 
tion is: 

Total number of leucocytes counted X Degree of dilution -=- 
Cubic contents of visual field X Number of fields examined = 
Total number of leucocytes per 

For example, in a i : 200 blood dilution a total of 30 leucocytes 
is noted in fifty fields, each having a cubic contents of 0.0125664, 
since they individually include 8 parallel columns of the ruled disc : 

(30 X 200) -f- (0.0125664 X 50) = 9550 leucocytes per 

(b) If the special leucocytometer is used for counting the leu- 
cocytes, a 0.3 per cent, aqueous solution of glacial acetic acid 
must be employed as a diluent, in order to render invisible the 
erythrocytes and at the same time to make the leucocytes appear 

1 "Twentieth Century Practice of Medicine," New York, 1896, vol. vii, p. 271. 


more conspicuously in the field. A i : 10 dilution is made by 
drawing the blood up the capillary tube of the instrument until 
the mark i is reached, and by then adding the diluent until the 
mixture reaches the mark n. The leucocytes are then counted 
in an area of the counting chamber equal to 8co of the small 
squares (preferably by examining 400 squares in two separate 
drops), and the calculation made according to the method pre- 
viously described. For instance, if in a given case 130 leucocytes 
were counted in 800 squares, the estimate would be made as 
follows : 

130 X 4000 X 10 -4- 800 = 6500 leucocytes per 

The chief objection to this method of leucocyte counting lies 
in the difficulty in distinguishing the cells, owing to the unavoid- 


By twisting the rubber suction tube into a tight spiral rope the fluid in the bore of the pipette may 
be forcibly expelled in a tine jet. 

able presence in the field of masses of granular debris resulting 
from the action of the acetic acid solution upon the erythrocytes. 
For this reason, if for no other, it seems advisable to dispense 
with the leucocytometer, and to make the count of both red and 
white corpuscles with the erythrocytometer in the same drop. 

Cleaning the Pipette, As soon as the count has been finished, 
the pipette should be carefully cleaned and dried. Having first 
expelled what remains of the blood dilution, the instrument is 
rinsed out, first with distilled water and then with a mixture of 
equal parts of absolute alcohol and ether, the latter being used to 
remove all traces of the dye, in case either Toisson's or Sherring- 
ton's solution has been employed as a diluent, as well as to dry 
the interior of the tube. The pipette, while it may be filled with 
a fluid by suction, should not be emptied by blowing through it, 
for if this is done, a certain amount of moisture from the breath 
unavoidably becomes deposited in its lumen. Its contents may be 


expelled in the form of a fine jet, simply by twisting the rubber suc- 
tion tube into a tight spiral rope, as shown in the illustration (Fig. 
21). When the interior of the instrument is perfectly clean, it 
is dried by forcing through it a current of air by means of a 
rubber atomizer bulb, or an ordinary bicycle-pump, until the glass 
bead no longer clings to the wall of the bulbous expansion, as it 
will as long as the slightest trace of moisture remains. 1 

A new form of hemocytometer has been 
DURHAM'S HEM- recently designed by Durham, who has em- 
OCYTOMETER. bodied in this device the principles of the older 
instruments, together with the substitution of a 
self-measuring pipette designed to overcome the sources of error 
which may occur in making blood dilutions with a suction pipette. 
Durham's instrument, which appears to be a valuable improve- 
ment over other forms of blood-counting apparatus, consists of 
the following parts: 

i. Several capillary pipettes of the Oliver type, each mounted 


T, Glass tube; N, rubber nipple; p, lateral perforation in nipple; c, cork in which a capillary pipette 

is fitted. 

in a glass tube, provided with a rubber nipple having a lateral 
perforation. The capacity of the pipettes is 5 and 10 

2. A number of mixing vessels, each consisting of a small glass 
test-tube, graduated from i and for 0.5 c.c. of fluid. The tubes 
holding i c.c. measure 2f x yV m -> an( ^ those holding 0.5 c.c., 
2f X f in. One or more glass beads are shaken about in the 
tube to mix the blood and the diluting fluid. 

3. A number of graduated pipettes for measuring the diluting 
fluid, of i and 0.5 c.c. capacity, marked at 995 and 990 
and at 495 and 490, respectively. Used with the appro- 
priate capillary pipette, dilutions of i : 200, i : 100, and i : 50 
may be obtained. 

4. A counting chamber of the Thoma-Zeiss pattern. 

Method o] Use. Having placed in one of the mixing vessels 
some of the diluting fluid, the quantity of which is measured with 

1 A o.i per cent, solution of pepsin in one per cent, hydrochloric acid is useful 
for removing any bits of clotted blood which may adhere to the caliber of the 


one of the graduated pipettes according to the dilution desired, the 
capillary pipette is filled with blood by touching it lightly to the 
blood drop as it flows from the puncture. All traces of blood are 
then removed from the outside of the pipette, the contents of which 
are now expelled into the fluid contained in the mixing vessel. 
This is accomplished by inserting the pipette into the latter, keep- 
ing its point about half an inch above the level of the diluting fluid, 
and by then rotating it between the thumb and forefinger so that 
the lateral perforation is brought under the ball of the thumb; 
the nipple is now squeezed gently, and, continuing the pressure, 
the pipette is rotated back so that the perforation is free again. 
In this manner the blood is forced from the pipette but is not 
sucked back. The blood remaining in the pipette is now com- 
pletely washed away by thrusting its point into the diluting fluid, 
this at once filling its caliber, by capillarity. Withdrawing the 
pipette from the fluid, the rotation and pressure of the nipple are 
repeated, the capillary tube being thus rinsed out several times in 
order to remove completely all the blood clinging to its interior. 

The blood and the diluting fluid are now mixed by briskly ro- 
tating the mixing vessel between the opposed hands, so that the 
tumbling about of the glass beads in the vessel may thoroughly 
distribute the cellular elements through the fluid. When the 
mixing is completed, a drop of the fluid is transferred to the 
counting chamber, and the corpuscles counted under the micro- 
scope in the usual manner. 

Durham's device makes it possible for the unskilled to measure 
accurately the desired volumes of blood and diluting fluid, and 
largely eliminates the errors which are likely to occur in sucking 
up the blood and the diluent with either the Thoma-Zeiss or 
the Gowers hemocytometer. The ease and thoroughness with 
which the capillary blood pipette may be cleaned are also ad- 
vantages, this being done by passing through its caliber a piece of 
darning-cotton, dry or soaked in ether, by means of a needle. 
Comparative observations made with the Thoma-Zeiss hemocy- 
tometer have shown that the readings of the two instruments are 
practically identical. 

In this form of hemocytometer the blood and 
GOWERS' HEM- the diluting fluid are each measured in a separate 
OCYTOMETER. pipette, and deposited in a small receptacle, in 
which they are mixed, a small portion of the mix- 
ture then being placed in a counting chamber and the number of 
corpuscles counted under the microscope. Gowers prefers to use 
as a diluent an aqueous solution of sodium sulphate having a 
specific gravity of 1.025, DUt Toisson's solution, or any of the 


other diluting fluids previously mentioned, will prove satisfactory. 
The instrument comprises five working parts, as follows : 

1. A pipette, graduated to hold a volume of 995, for 
measuring the diluting fluid. 

2. A capillary pipette, graduated to hold a volume of 5, 
for measuring the blood. 

3. A small glass mixing jar, in which the dilution of the blood 
is made. 

4. A glass stirring rod, for mixing the blood and the diluent 
in the jar. 

5. A counting chamber, consisting of a glass slide mounted on 
a brass plate, and containing a cell ^ mm. in depth, the floor of 
the cell being divided by cross-rulings into squares the sides of 
which measure y^ mm. When a cover-glass is fitted over this 
cell, being retained in position by means of a pair of clips attached 
to either end of the brass plate, the cubic contents of the space 
overlying each square measure ^-^ 

Method of Use. In using the instrument 995 of the 
diluting solution are first measured by means of the larger pip- 
ette and blown out into the mixing jar. The latter must be 
perfectly clean and absolutely free from moisture before it is 
used, in order to avoid errors in the count. Now, using the 
capillary blood pipette, 5 of blood are secured from the 
puncture, and immediately added to the diluent contained in the 
jar. The blood and the diluent are then thoroughly mixed, by 
rapidly stirring the solution with the glass rod. The dilution thus 
made is in the proportion of i : 200 of blood to diluent. As 
soon as the mixture is completed, a small drop of the solution is 
transferred to the center of the cell in the middle of the counting 
chamber, the small end of the glass rod being used for this pur- 
pose, after which the cover-glass is gently placed in position, and 
the clips adjusted so as to hold it in place. The counting cham- 
ber may then be placed upon the stage of the microscope, and 
the corpuscles overlying the ruled portion of the cell brought 
into focus with a low-power objective. 

It is necessary to use a small drop of the diluted blood, and 
to place it exactly in the center of the block of ruled squares, 
otherwise the fluid may flow toward the walls of the cell, alter- 
ing its volume and making it necessary to reject the specimen 
and to prepare a new drop, after thoroughly cleaning and drying 
the cell, and again stirring the blood solution. 

The corpuscles having settled to the bottom of the cell, their 
number in a given number of squares is noted, and the final cal- 
culation made according to the formula: 


Number of Number of ~, . , 

corpuscles X 200 X 5 oo + squares = To ' al mb * r / """ 
counted counted puscles per 

In counting the erythrocytes at least 20 squares of the count- 
ing chamber should be inspected, in different drops, a procedure 
involving the enumeration of about 1000 cells, in normal blood. 
Except in high leucocytoses, the number of leucocytes is usually 
estimated indirectly, by determining their ratio to the erythro- 
cytes, and basing their actual number upon this figure. This 
plan (the necessity for which is a serious drawback to the use of 
this instrument) is followed so as to dispense with the tedious 
filling and refilling of the counting chamber, in an endeavor to 
find a sufficient number of leucocytes to serve as a basis for the 
calculation, should the latter be direct. Ordinarily, not more 
than two of these cells are contained in an area including 20 
squares. Gowers * claims that the limit of error with his instru- 
ment is less than 3 per cent. 

After use, the different parts of the instrument are to be thor- 
oughly cleaned and dried, in the manner already described. 

For making rapid numerical estimates of the 
OLIVER'S HEM- erythrocytes Oliver has designed an instrument 
OCYTOMETER. based upon the following principles: When a 
candle-flame is viewed through a flat glass test- 
tube filled with water, a bright transverse line is visible, com- 
posed of densely packed, minute images of the flame produced 
by the longitudinal corrugations of the glass. If lor the water a 
mixture of blood and Hayem's solution 2 is substituted, a more or 
less opaque fluid results, so that, in low dilutions, this illuminated 
line is invisible, reappearing only when a definite degree of higher 
dilution is reached, by the gradual addition of the diluent; when 
this point has been obtained, the line is again detected as a bright, 
delicate streak horizontally crossing the tube. Experiments having 
proved that the development of such a line, by gradual dilution 
of the blood with Hayem's fluid, is an accurate gage of the per- 
centage of erythrocytes in the specimen tested, it remained for 
Oliver to devise a hemocytometer consisting of the following 
essential parts: 

1. A capillary pipette for measuring the blood. 

2. A glass dropper, one end of which is capped by a rubber 
nipple, the other by a short rubber nozle which fits over the blunt 
end of the pipette. 

3. A standard graduated tube, in which the blood and the 
diluent are mixed. 

1 Lancet, 1877, vol. ii, p. 797. 2 For formula see page 56. 


The four walls of the tube are flattened so that it is rectangular 
on cross-section, one wall being provided with an etched scale 
indicating units from 10 to 120. Each of these divisions is equiv- 
alent to 50,000 erythrocytes, the point marked ico degrees repre- 
senting the arbitrary normal number, 5,000,000. 

Small-sized wax candles, known as "Christmas candles," are 
to be preferred for the illumination, as they give the small flame 
requisite to obtain a sharply denned line, but the flame from a 
gas-jet turned low may also be used with satisfaction. 

Method o] Use. In making the observation the pipette, which 
has been previously cleaned and dried, is filled with blood in the 
usual manner, and any excess of blood on the outside carefully 

removed. The rubber 
nozle of the dropper, 
filled with Hayem's 
fluid, is then slipped 
over the blunt end of 
the pipette, and the 
blood washed out into 
the graduated tube by 
squeezing the nipple. 
This preliminary dilu- 
tion is continued until 
the column in the tube 
rises to within 10 or 15 
degrees below the fig- 
ure for the hemoglobin 
percentage of the same 
blood, this having been 
previously determined. 
For instance, if the 

hemoglobin percentage was found to be 70, the diluting fluid is added 
in large quantities until the mixture in the tube reaches to about the 
mark 60, after which it is added more cautiously and in smaller 
quantities at a time, careful search for the bright line being made 
after each addition. In cases of chlorosis and of pernicious anemia, 
in which parallelism between the hemoglobin and corpuscular loss 
is lacking, it is, of course, impossible to depend upon the hemo- 
globin percentage as an index to the amount of diluent required, 
so that in instances of this kind the line must be developed more 
slowly, by making a. smaller primary dilution and by adding the 
requisite volume of liquid more deliberately. 

After the first dropperful of diluent has been added to the con- 
tents of the tube, the latter are mixed by inverting the tube a 

Showing manner in which the blood is washed from the capil- 
lary pipette into the tube containing Hayem's solution. 


number of times with the thumb held over its mouth, the precau- 
tion being taken also to remove the thumb by drawing it over the 
mouth of the tube, in order to restore to its contents any liquid 
which may have adhered to the skin. The tube should be inverted 
thus after each addition of the diluting fluid. 

The steps of the observation succeeding the measuring of the 
blood and its primary dilution are to be made in a dark room, 
free from cross lights, the candle being placed about ten feet dis- 
tant from the observer. In order to shut out the diffused light 
of the candle the tube should be held vertically in the concavity 
formed between the thumb and forefinger, being kept close to 
the eye while searching for the bright line. Oliver states that the 
earliest indications of this line are obtained by turning the tube 
on its axis, when it will become visible at the sides. 

Apart from the "personal equation," the serious drawback 
to this test is its failure to indicate the number of leucocytes, 
this fact alone being sufficient to curtail its use for routine clinical 
work. It is also apparent that in cases of marked leucocytosis 
and of leukemia the optical principles of the test must necessarily 
fail because of the enormous number of leucocytes in the blood. 
Furthermore, the instrument gives false results with blood in 
which conspicuous deformities of the erythrocytes exist, for the 
reason that the standard tube is corrected for normally shaped 
corpuscles, so that blood composed largely of microcytes, megalo- 
cytes, or poikilocytes will give different readings from blood in 
which the cells are of unaltered biconcave shape and of normal 
size. The discrepancies between the Oliver and the Thoma- 
Zeiss instruments have been carefully worked out by Emerson l 
and by Baumgarten. 2 

This method, devised by Strong and Selig- 

DRY FILM mann, 3 aims to eliminate errors due to variations 

METHOD. in the depth of the Thoma-Zeiss counting chamber, 
and at the same time to furnish permanent speci- 
mens which may be examined at any subsequent time. It is 
practically a simplification of the method suggested by Einhorn 
and Laporte. 4 The principle involved consists in diluting a 
measured volume of fresh blood with a measured volume of a 
diluent with which a suitable differential stain is combined, a 
definite quantity of the mixture then being spread upon a glass 
slide, allowed to evaporate, mounted in the ordinary manner, and 
examined microscopically. 

1 Johns Hopkins Hosp. Bull., 1903, vol. xiv, p. 9. 

1 Ibid., 1902, vol. xiii, "p. 176. 

8 Brit. Med. Jour., 1903, vol. ii, p. 74. 4 Med. News, 1902, vol. Ixxx, p. 741. 


Two different diluting fluids are used, one for the leucocytes 
and one for the erythrocytes. Tabloids (made by Parke, Davis 
and Company) containing 0.25 gm. of sodium chlorid combined 
with 0.004 gm. of methyl- violet (for leucocytes) and with 0.0025 
gm. of eosin (for erythrocytes) are employed for making the dil- 
uents, instead of stock solutions, which become unstable after a 
short time. To prepare the diluents one of each of these tabloids 
is dissolved in 30 c.c. of distilled water to which 0.5 c.c. of 
formalin is added, the mixture then being filtered. 

Counting the Leucocytes. Five of blood are sucked up 
into a graduated pipette, and then blown out into a small vessel 
containing 495 of the methyl-violet diluent. After stirring, 
5 of this i : 100 mixture are drawn up in a pipette and 
deposited upon the surface of a glass slide so as to form a film 
about 10 or 12 mm. in diameter. The film thus made is allowed 
to evaporate and then mounted with balsam and a cover-glass. 

The count is made by going over the entire area of the film with 
a -inch dry objective, and noting the number of violet-stained 
cells. In order to facilitate this procedure, either a square or an 
oblong ocular diaphragm, made of black paper or of metal, 
should be used, together with a mechanical stage. The blood dilu- 
tion being i : 100 and the volume of this mixture spread upon 
the slide being 5, the number of leucocytes per 
of undiluted blood is therefore 100-^-5, or 20 times the total 
number counted in the entire film. For example, having noted 
400 leucocytes in the latter, the simple formula 400 X 20 = 8000 
leucocytes per, gives the final calculation. An error of 50 
cells in the count alters the final result by only 1000 cells a trivial 

Counting the Erythrocytes. In this instance a i : 20,000 dilu- 
tion is made, by mixing 5 of the above i : 100 blood and 
methyl-violet diluent with 995 of the eosin diluent. After 
allowing the erythrocytes to stain for a few minutes, 5 of 
this mixture are spread over a slide, as described above, similarly 
mounted, and examined microscopically, the erythrocytes being 
recognized by their rose-red color. The number of erythrocytes 
per of undiluted blood is 4000 times the number counted 
in the dry eosined film, since 5 of a i : 20,000 dilution of 
the blood has been used ; thus, 20,000 H- 5 = 4000. For instance, 
having counted 1200 erythrocytes in the 5 film, the formula 
1 200 X 4000 = 4,800,000 erythrocytes per, gives the final 

The originators of the dry film method of blood counting 
insist that its results are more accurate than are possible with a 


hemocytometer. In three Thoma-Zeiss instruments, three years 
old, they found figures consistently ten per cent, higher than those 
obtained by new instruments of the same design and make, an 
error which could be explained by an increase of o.oi mm. in 
the depth of the counting chamber. Counts made by the dry 
method averaged i.i per cent, lower than those in which the hemo- 
cytometer was used. Several sources of error, however, must be 
guarded against. For example, in making the dilutions, unless 
every trace of the measured 5 of whole blood is blown 
from the pipette into the diluting fluid, the dilution will be too 
high ; and if, in making the film, the diluted blood is blown out 
too forcibly, the film will be scattered and ragged and stippled with 
air-bubbles. The erythrocytes may stain violet instead of rose 
should the diluting solution not be perfectly fresh. The leuco- 
cytes, which may stain violet, usually take the color of eosin, with 
the eosin dilution for erythrocytes. Thus, the former are un- 
consciously counted with the latter, but under ordinary circum- 
stances this is an unimportant error, since the number of leuco- 
cytes is comparatively too small appreciably to affect the erythro- 
cyte count. In leukemia, however, the error may be suffi- 
ciently great to need correction. In this disease, therefore, all 
the blood cells, red and white, should be counted, multiplied by 
4000, and from this total is to be subtracted the total leuco- 
cyte estimate previously determined, the result being obviously 
the total number of erythrocytes per Unfortunately, 
the specimens prepared by Strong and Seligmann's technic are 
marred by deposits of salt crystals and by masses of amorphous 
eosin, both at the edges of the film and scattered throughout it. 
This defect, if it does not render an accurate count impossible, 
may at least make it difficult and tedious. In a passably good 
specimen the counts of erythrocytes and leucocytes together should 
not take longer than half an hour. 


The microscopical study of the dried and 
OBJECTS OF stained blood film, which should supplement the 
STAINING. methods of investigation just described, is for 
many reasons the most important step in the 
clinical examination, of the blood. By means of this method 
of "color analysis" it is possible to differentiate easily and with 
absolute certainty the various forms of leucocytes, and, by differ- 
ential counting, to calculate the relative percentages of each 


variety of these cells; to distinguish the several structural de- 
generative changes affecting chiefly the erythrocytes, and to a 
less extent the leucocytes; and to recognize and classify accord- 
ing to their histological character the nucleated forms of the eryth- 
rocytes. To sum up, in the words of Ehrlich, 1 to whom we 
owe this rational means of investigation :" Everything that is 
to be seen in the fresh specimens apart from the quite unim- 
portant rouleaux formation and ameboid movements can be 
seen equally well, and indeed much better, in a stained prepara- 
tion; and there are several important details which are made 
visible only in the latter, and never in wet preparations." 

According to the classification introduced 
THE ANILIN many years ago by Ehrlich, 2 the anilin dyes 
DYES. are divided into three different groups: acid, 
basic, and neutral. Acid dyes, or compounds in 
which the coloring principle acts or exists as an acid, possess a 
special affinity for cell protoplasm, and, therefore, are generally 
employed as plasma stains; in hematological work acid fuchsin, 
eosin, and orange G are the principal dyes used for this purpose. 
Basic dyes, or compounds in which the coloring principle exists 
chemically as a base in combination with a colorless acid, are 
especially useful as nuclear stains, since they exhibit a special 
affinity for chromatin structures; members of this group of 
dyes commonly used in blood staining are methylene-blue, tol- 
uidin-bluej methyl-green, methyl-violet, thionin, and hema- 
toxylin. Neutral dyes are the coloring principles which result 
from the mixture of solutions of an acid and a basic dye; they 
are used for the demonstration of the so-called neutrophile 
granules of the leucocytes, for which they show a selective 

Cover-glass Films. For the preparation of 
PREPARING the dried blood films it is advisable to have at 
THE FILMS, hand at least half a dozen perfectly clean, polished 
cover-glasses, which may be arranged in pairs 
on a sheet of white paper within convenient reach of the examiner. 
After having wiped away the blood which immediately follows the 
puncture, a minute portion of the next drop is collected, by lightly 
touching the center of one of the cover-glasses to its summit, care 
being taken to avoid bringing the polished surface of the glass in 
contact with the skin of the patient's finger. The charged cover- 
glass is then at once dropped, blood side downward, upon the 

1 Ehrlich and Lazarus, "Die Anaemic," Vienna, 1900 (Nothnagel's "Spec. 
Path. u. Therap.," vol. viii, No. 2). 

2 Zeitschr. f. klin. Med., 1880, vol. i, p. 555. 




surface of the second glass (Fig. 24), with the result that the 
blood quickly spreads out in a thin film between the two, and ex- 
tends to their peripheries, provided that the proper quantity of 
blood has been used (Fig. 25). As soon as the film has reached 
the margins of both cover-glasses, they are rapidly drawn apart 
in a horizontal direction, so that the surface of each, when thus 
separated, is covered with a thin layer of blood (Fig. 26), which 
should be rapidly dried, either by blowing briskly upon its sur- 
face or by holding the glass for 
a few seconds high over the 
flame of an alcohol lamp. If 
care is taken to use but a very 
small drop of blood, to avoid 
pressure in opposing the sur- 
faces of the two cover-glasses, 
and to separate them in their 
true horizontal planes, the 
films will consist of a single 
layer of corpuscles, most of 
which will be sufficiently isolated to allow the study of their 
individual morphology and other characteristics. The beginner 
should persistently practise the technic of film-making until he 
is able to obtain a satisfactory percentage of good specimens from 
every batch of spreads. Thick, uneven spreads, in which the 
corpuscles are heaped up and glued together in dense masses, 

are of no value for histologi- 
cal study of the cells, al- 
though they are often useful 
when searching for parasites 
in blood containing very 
few organisms. In such 
instances Ross 1 advises 
smearing the blood thickly 
on the slide, and, after it 
has dried, washing out the 
hemoglobin with water. 
The film thus dehemoglobinized, unfixed, and still wet, is then 
stained with aqueous solutions of eosin and methylene-blue, washed 
with water, and mounted. The parasites stain the color of the basic 
dye, but the erythrocytes, since they contain no hemoglobin, are 
practically colorless. This method, while, of course, unsuitable for 
differential counting, may be of value in certain cases of malarial 
fever, filariasis, and trypanosomiasis. 

1 Thompson Yates and Johnston, Lab. Rep., 1903, vol. v, p. 117. 



The films, after having been dried, may be placed in a pill box 
and labeled, to await fixation and staining at the examiner's con- 
venience. Dried specimens will keep for an indefinite period if 
not exposed to dust or to moisture. Unfixed cover-glass specimens 

of leukemic blood have 
been "triple-stained" by 
the writer, with perfect 
results, more than three 
years after they were 
spread. With the Rom- 
anowsky method, how- 
ever, the fresher the 
specimen, the sharper the 
stained film. 

Many histologists recommend the use of special forceps for 
holding the cover-glasses while making the spreads, claiming thus 
to avoid the injurious effects upon the blood corpuscles which may 
be caused by the moisture of the fingers if they come in con- 



tact with the films. The careful worker need have no fear on this 
score, for if the covers are held in the manner shown in the illus- 
trations, this accident will not occur. A pair of light thumb 
forceps is useful for picking up the cover-glasses 

from a flat surface, but the employment of spe- 
cial spreading forceps is quite unnecessary. 

Glass Slide Films. Some prefer to make the 
spread upon an ordinary glass slide, but this 
method rarely yields as thin and even a film 
as the one just described, though it is easier 
to learn. A small-sized drop of blood is dis- 
tributed along the edge of a glass slide, 1 which is then held 
at an angle of 45 degrees against the surface of another slide, 

1 Waldstein's smearing slip (Med. Rec., 1896, vol. 1, p. 385), made of crown 
glass, with its " spreading edge " ground smooth and round, answers much better 
than an ordinarv slide. 




over which it is rapidly drawn, with moderate pressure, thus 
depositing a thin film of blood upon the surface of the latter 
(Fig. 27). Instead of a glass slide, a leaf of cigarette paper 
or a slip of thin tissue-paper, trimmed to a narrower width 
than that of the slide, may be used as a spreader (Fig. 28). A 
fair spread may also be made by depositing a drop of blood upon 
a slide, near one end, and then distributing it by means of a needle 
or a glass rod, the shaft of which is applied to the drop with even 
pressure. The film thus made is 
immediately dried in air and treated 
as a cover-glass spread. 

As a step pre- 

FIXATION liminary to stain- 
METHODS. ing, the albuminoid 
principles of the 
blood must be fixed, by exposing 
the dried film either to a high de- 
gree of dry heat or to various chemi- 
cal hardening agents, the choice be- 
tween these two methods being de- 
termined by the character of the 
staining solution to be used subse- 

Heat Fixation. This method 
may be employed with any of the 
stains described in the following 
pages, except with Wright's solu- 
tion; it must be used with Ehrlich's triple stain, in preference to 
fixation by chemicals, in order to obtain crisp, clean-cut pictures., 

The author is accustomed to use an oven, such as is illustrated 
above (Fig. 29), consisting of a copper box with a heavy bottom 
and hinged cover, mounted on an ordinary iron burette stand, by 
means of a thumb-screw. A small "baby" Bunsen lamp placed 
underneath the box furnishes the requisite degree of heat, the 
temperature being indicated by a thermometer mounted at one end 
and resting upon the floor of the oven. By sliding the latter up and 
down the vertical rod to which it is attached the desired degree of 
temperature may be obtained at will. The blood films are inclosed 
in the copper box, and the latter fixed at a point eight inches above 
the summit of the burner, after which the gas is lighted and allowed 
to burn until the temperature, as indicated by the thermometer, has 
gradually crept up to 160 C. As soon as this degree of heat has 
been reached the gas is extinguished, the cover of the oven thrown 
back, and, after the temperature has fallen to 30 C., the films 



removed, being now thoroughly fixed and ready for staining. 
Fifteen minutes suffice for the whole operation, from the time the 
gas is lighted until the films have been removed and cooled, for 

A less satisfactory method of heat fixation is by the use of a 
copper plate upon which the films are kept at a temperature of 
from 100 to 110 C. for from one-half to one hour. The ap- 
paratus used for this purpose consists of a rectangular plate of 
sheet copper, about fifteen inches long by four inches wide by 
one-sixth of an inch thick. An alcohol or a Bunsen lamp burns 
under one end of the plate, which is elevated about six inches 
above the flame by four metal legs. After having heated the 
plate for ten or fifteen minutes, until a relatively constant tem- 
perature becomes established, water is dropped upon its surface, 
beginning with the end farthest from the flame, until a point is 
reached at which the water boils. This part of the plate is con- 
sidered to have a temperature of 100 C., and at this point the 
blood films are placed, " spread" side downward, and heated for 
the required time. No one with much blood staining to do will 
choose this method of prolonged heating at a relatively low, ap- 
proximate temperature in preference to brief heating at a high, 
definite temperature in an oven. The use of the latter, aside 
from its convenience as a time-saver, insures constant and cer- 
tain results, for overheating and underheating of the blood film may 
be avoided, since the degree of heat is exactly indicated and easily 
controllable. In triple-stained specimens the blood cells are much 
more brilliantly colored and sharply differentiated when the films 
are fixed at a temperature of 160 C. than at a lower degree. 

Should nothing but a Bunsen or an alcohol lamp be available, 
the cover-glass film, held with a pair of forceps, may be fixed by 
passing it rapidly through the flame thirty or forty times and then 
holding it twelve or fifteen inches above the flame for a minute 
or so. This makeshift method, which is often sufficient for a hur- 
ried clinical examination, usually gives fair, and sometimes very 
good, results, but the fixation is generally uneven, and the speci- 
men is frequently scorched in some places and underfixed in 

Chemical Fixation. Immersion of the dried films in ether, in 
absolute alcohol, or in a mixture of equal parts of the two (Niki- 
foroff's method) gives satisfactory results with specimens stained 
by any of the single basic dyes, or with the simpler double stains, 
such as eosin and methylene-blue or hematoxylin. The time of 
fixation varies from five to fifteen minutes with any of these agents, 
the specimen then being dried without using heat and stained 


without previously washing. If time is an object, the specimens 
may be boiled for one minute in a test-tube containing absolute 
alcohol, as advised by Ehrlich. 1 Some workers employ one 
minute's fixation by a one per cent, alcoholic solution of forma- 
lin (Benario's method), while others prefer to expose the films to 
the vapors of this chemical for the same length of time. Five 
to ten minutes' immersion in a concentrated aqueous solution of 
mercuric chlorid is one of the older, but useful, methods of fixation. 
Solley 2 has recently suggested that the film be flooded with a two per 
cent, aqueous solution of chromic acid, which is washed off after 
exactly thirty seconds, the specimen being then stained while still 
wet ; he recommends this procedure as a substitute for heat in fixing 
specimens for triple staining, but the method, while fairly good, 
cannot be regarded as entirely satisfactory. In the author's hands 
both Merck's methyl alcohol (five minutes) a'nd a two per cent, 
aqueous solution of osmic acid (half a minute) have been found to 
be fair substitutes for heat fixation. 

In hematological as in other histological work 
METHODS OF the choice of a staining method is determined by 

STAINING the character of the investigation to be undertaken. 
For general clinical purposes it is advantageous 
habitually to employ some routine method by means of which 
the greatest possible number of elements may be demonstrated 
in a single blood film, this procedure being known as panoptic 
staining. Thus, by using a solution containing several of the 
anilin dyes, the stroma of the erythrocytes, the cell granules, the 
cell nuclei, and the various blood parasites may be simultaneously 
stained each in a characteristic manner, owing to the selective 
affinity displayed by the different coloring principles of the mix- 
ture toward these several histological elements. The most useful 
solutions which have been devised for this purpose are Wright's 
alkaline eosinate of methylene-blue 3 and Ehrlich's triacid mixture. 4 
Practically all the information that it is possible to derive from 
the study of the stained dry blood film may be obtained with the 
aid of these two solutions. 

Simple combinations of an acid and a basic dye, such as eosin 
and methylene-blue, eosin and hematoxylin, and orange and hema- 
toxylin, are used by many investigators, chiefly for the purpose 
of staining the cell stroma and the nuclear structures; but, as a 
general rule, such mixtures are not adapted for clinical work, 

1 Loc. cit. 

2 Med. and Surg. Reports of the Presbyterian Hospital, New York, 1900, vol. 
iv, p. 169. 

3 Jour. Med. Research, 1902, vol. vii, p. 138. 4 Loc. cit. 



since with none of them is it possible to differentiate the neutro- 
phile granules. Solutions of a single dye are but seldom used 
except for the demonstration of special elements, such, for in- 
stance, as the staining of the malarial parasite by thionin, the 
mast cells by dahlia, and certain bacteria by the basic dyes, such 
as methylene-blue and gentian-violet. Since by this method of 
staining only the particular elements toward which the dye reacts 
are differentiated, the employment of single stains is inadequate 
for the study of the general morphology of the blood cells. 

The following formulas will be found sufficient for all purposes 
of clinical investigation: 

WRIGHT'S STAIN. This excellent stain is an improvement on 
Leishman's modification 1 of the late Louis Jenner's solution, 2 and 
is so prepared that its basic constituent (methylene-blue) acquires 
the polychromatic properties of the Romano wsky stain, 3 so valuable 
in differentiating chromatin and cytoplasm. Wright's stain gives 
sharp pictures of cell protoplasm, chromatin, and nuclei, and is 
especially useful in studying the lymphocytes, the mast cells, the 
blood plaques, and the finer structure of the malarial parasite. 
It is prepared according to the following somewhat complicated 
formula : 

1. To a 0.5 per cent, aqueous solution of sodium bicarbonate 
contained in an Erlenmeyer flask add one per cent, of Griibler's 
methylene-blue (''medicinal"), and steam the mixture in a 
sterilizer for one hour, counting from the time "steam is up."* 
This step not only serves to develop the polychromatic property of 
the alkaline methylene-blue, but increases its power as a nuclear 
and granular stain. 

2. Cool the bicarbonized methylene-blue solution after steam- 
ing, and then, without filtering, add to it, meanwhile stirring with a 
glass rod, sufficient of a i : 1000 aqueous solution of Griibler's 
yellowish eosin ("water soluble") to change the color of the 
solution from blue to purple, with a surface scum of a yellowish, 
metallic luster. About 500 c.c. of the methylene-blue solution to 
100 c.c. of the eosin solution are required to produce this change. 

1 Brit. Med. Jour., 1901, vol. ii, p. 757. 2 Lancet, 1899, vol. i, p. 370. 

8 Centralbl. f. Bakt., 1899, vol. xxv, p. 764. 

4 When a steam sterilizer is not available, polychrome blue may be developed 
by heating the methylene-blue solution with freshly precipitated silver oxid. Dan- 
iels ("Studies in Laboratory Work," London, 1903, p. 63) directs that a solution 
of sodium hydrate be added to one of silver nitrate until no more precipitate forms, 
the precipitate, silver oxid, being then washed until the washings are neutral to 
litmus-paper. This neutral precipitate is added to the methylene-blue solution 
and allowed to stand for twenty-four hours, whan the supernatant fluid is decanted 
off from the sediment and filtered before using. Polychrome blue thus prepared 
is treated according to the above directions for making Wright's solution. 


3. Collect, by filtration, this scum (which consists of a granular 
black precipitate), and dry, without washing. When dry, make 
of it a saturated solution in methyl alcohol. Three-tenths of a 
gram of the dry precipitate saturates 100 c.c. of methyl alcohol. 

4. Filter the alcoholic solution of the precipitate and add to the 
filtrate 25 per cent, of methyl alcohol. For 80 c.c. of the filtrate, the 
amount usually available, 20 c.c. of methyl alcohol are required. 
This alcoholic solution is used for staining. It does not deteriorate 
with age if kept in a well-corked bottle, and is sufficiently dilute 
to prevent precipitation during staining, an accident which was 
the chief drawback to Jenner's original stain. 

Technic of Staining. Owing to the methyl alcohol which it 
contains, Wright's solution fixes and stains the blood film simul- 
taneously, so that preliminary fixation may be omitted. 

The unfixed film is stained for one minute, as much of the 
solution being used as the cover-glass will hold without spilling. 
Next, to the staining fluid upon the specimen are added, drop by 
drop, eight or ten drops of distilled water sufficient to develop 
a reddish tint at the margins of the cover-glass and a semitrans- 
lucency in the stain, with a metallic scum on the surface. The 
stain, thus diluted, is allowed to act for two or three minutes, when 
it is rinsed off with water, showing that the film has become 
stained deep blue or purplish. The final step in the process is the 
decolorization of the overstained specimen and the differentiation 
of its histological elements. This is accomplished by washing 
with water until the color of the film changes to yellowish or pink. 
From one to three minutes' washing, depending upon the intensity 
of the staining, is required to reach the desired tint. The specimen 
is then dried between filter-paper (never by heat), and mounted 
in balsam. 

Wright's stain gives the following results : erythrocytes, orange 
or pink; nuclei of the leucocytes, blue or dark lilac; neutrophile 
granules, lilac; eosinophile granules, pink; fine basophile granules, 
deep blue; and coarse, mast cell granules, deep royal purple. 
The nuclei of the erythroblasts and bacteria stain various shades 
of blue, and the blood plaques purplish, flecked with red. 

The body of the malarial parasite stains blue, and its chromatin 
varies from lilac to red to almost black. 

Aside from its obvious value as a panoptic staining fluid, this 
solution will often prove of great convenience for the reason that 
it does not require special fixation of the blood film. 

EHRLICH'S TRIACID STAIN. This "triple stain," containing 
one basic and two acid dyes (methyl-green, orange G, and acid 
fuchsin), is peculiar in that a chemical combination is formed by 


its acid and basic components, which may be regarded as a neutral 
coloring principle, serving the purpose of selectively staining the 
so-called neutrophile elements for which the primary components 
of the mixture have no affinity. With this stain histological 
structures having an affinity for the acid dyes are stained the color 
of one of its acid constituents, basic structures the color of its 
basic dye, and structures having an equal affinity for acid and 
basic dyes the color of the neutral compound. 

Saturated aqueous solutions of the three dyes are first pre- 
pared, and allowed to stand for several days until they have be- 
come thoroughly cleared. It is essential that the anilin dyes used 
for making these "stock" solutions should be chemically pure, to 
insure which the products of Grubler or of the Berlin Anilin Dye 
Company should invariably be chosen. From these saturated 
solutions the following mixture is made: 

Acid fuchsin solution 6-7 c.c. 

Orange G solution 13-14 c.c. 

Distilled water 15 c.c. 

Absolute alcohol 15 c.c. 

Mix the above thoroughly and add, drop by drop, with con- 
tinuous agitation, in the following order : 

Methyl-green solution 12.5 c.c. 

Absolute alcohol 10.0 c.c. 

Glycerin 10.0 c.c. 

The mixture should under no circumstance be filtered, but 
allowed to stand for about twenty-four hours in order that a slight 
precipitate may form. As soon as this occurs the stain is ready 
for use, the necessary quantity being pipetted from the super- 
natant fluid without disturbing the precipitate. 

Technic of Staining. The heat-fixed film, held preferably 
with a pair of Stewart's staining forceps, is flooded with the stain, 
which is washed off in running water after the lapse of from five 
to eight minutes, the specimen then being dried by gentle heat 
and mounted in xylol balsam or in cedar oil. 

In the specimen thus prepared the stroma of the erythrocytes is 
stained orange, the nuclei of the leucocytes greenish-blue, the neu- 
trophile granules violet or lavender, and the eosinophile granules 
copper red. The nuclei of the erythroblasts react with varying 
degrees of intensity toward the basic component of the mixture, 
those of the normoblasts staining deep purple or black, and those 
of the megaloblasts pale green or greenish-blue. The basophile 
granules remain unstained, appearing as dull white, coarse, 
stippled areas in the cell protoplasm "negative staining." In 
order to stain these granules, as well as the basic protoplasm of 


the lymphocytes, Hewes 1 suggests that the triple-stained film, 
after having been washed, be subjected for from one-half second 
to ten seconds to Loffler's 2 methylene-blue solution, after which it 
is again washed, and mounted as above directed. This modifica- 
tion is of undoubted value, chiefly because it usually enables one 
to differentiate the larger forms of lymphocytes from the large 
mononuclear leucocytes. Malarial and other parasites are also 
distinctly stained by this method. 

Unsatisfactory results with the triple stain, provided that the 
latter is properly made, can almost always be attributed to faulty 
fixation. As already remarked, heat is the only method of fixa- 
tion which will insure faultless differentiation in the specimen 
stained with this mixture. The perfect specimen is of a deep, 
rich orange tint to the naked eye; if underheated, the film reacts 
too strongly toward the acid fuchsin of the mixture, and, conse- 
quently, is the color of this dye; if overheated, the plasma stain, 
orange G, is but feebly displayed, so that the color of the film is 
pale lemon yellow. 3 As a stain for general clinical work Ehrlich's 
is inferior to Wright's. Although a sharper neutrophile stain, it 
reacts feebly toward basophile structures, and requires careful 
and skilful heat fixation of the films. 

PRINCE'S STAIN. This mixture, which consists of an aqueous 
solution of one basic and two acid dyes, is an excellent stain for 
the differentiation of both nuclei and granules, and may be em- 
ployed as a fair substitute for either of the two preceding solutions. 
It should be made in this manner: 

Saturated aqueous solution of toluidin blue 24 c.c. 

Saturated aqueous solution of acid fuchsin i c.c. 

Two per cent, aqueous solution of eosin 2 c.c. 

These solutions are mixed in the order named, and shaken 
briskly for several minutes, so as to secure complete precipitation 
of the basic toluidin blue by the acid dyes. The solution, which 
should not be filtered, is ready for use as soon as made. Only 
the supernatant fluid should be employed, care being taken not 
to disturb the sediment. 

Technic of Staining. If a newly made solution is used, the 
films are stained for from one-half to one minute, after which they 
are rinsed in water, dried in air, and mounted; but if the solution 
has stood for several weeks, its basic constituent becomes less 

1 Boston Med. and Surg. Jour., 1899, vol. cxli, p. 39. 

'Saturated alcoholic solution of methylene-blue, 30 c.c.; i: 10,000 aqueous 
solution of potassium hydrate, 100 c.c. 

* A reliable triacid stain, made according to Ehrlich's formula, is sold by 
Messrs. Shinn and Kirk, Philadelphia. 


active, so that the specimen requires to be stained for from five to 
ten minutes. Either chemical or heat fixation of the blood film 
may be used with this stain, both methods giving equally sharp dif- 
ferentiation. Prince's solution colors the erythrocytes rose-red, the 
nuclei of the leucocytes and erythroblasts blue, the neutrophile 
granules pink, the eosinophile granules maroon, and the fine and 
coarse basophile granules blue. Blood parasites are also stained 
the color of the basic dye. 

Crisp, clear pictures of nuclear and stroma structures, of the ma- 
larial parasites, and of the basophile granules may be obtained 
by the use of these two dyes, and to investigations of this nature 
should this staining method be restricted. It is impossible, for 
example, accurately to distinguish a large lymphocyte from a 
myelocyte in a specimen stained in this manner, so that for differ- 
ential counting a more elaborate stain is essential. In films stained 
by this method the stroma of the erythrocytes and the eosinophile 
granules react toward the acid dye, staining the color of eosin; 
while the nuclei of the leucocytes and erythrocytes, the basophile 
granules, and all blood parasites show an affinity for the basic 
dye, being colored various shades of blue. The protoplasm of 
the polynuclear neutrophiles is either colorless or tinged a deli- 
cate pink, the granules of these cells remaining unstained. 

The author has always found the following simple formula de- 
pendable : 

Eosin (aqueous), to which sufficient water has been added 

for solution 0.5 gm. 

Absolute alcohol 0.5 c.c. 

Saturated aqueous solution of methylene-blue 96.0 c.c. 

Technic of Staining. Films are fixed by immersion for ten 
minutes in absolute alcohol or in equal parts of absolute alcohol 
and ether. The cover-glass is flooded with the stain, gently 
heated for one minute over a Bunsen flame, allowed to stain 
without heat for two or three minutes longer, and then thoroughly 
washed in running water, dried in air, and mounted. 

Another method of staining with eosin and methylene-blue, 
slower than the above, but as a rule giving sharper differentiation, 
is to stain without heat for five minutes with a 0.5 per cent, solu- 
tion of eosin in absolute alcohol to which an equal quantity of 
water is added. Then, after having washed off the eosin solution 
and dried the film in air, the specimen is counterstained for one 
minute or less with a saturated aqueous solution of methylene- 
blue, after which it is rinsed again in water, dried in air, and 


Among the many other methods of staining with eosin and 
methylene-blue those suggested by Chenzinsky, 1 by Plehn, 2 by 
Holmes, 3 by Laporte, 4 and by Hastings 5 will be found the most 

employment of these two dyes the erythrocytes and the eosinophile 
granules are stained the color of eosin, and all nuclei and parasites, 
the color of hematoxylin. This method, which is decidedly in- 
ferior to staining with the eosin and methylene-blue mixtures just 
described, is useful for little else than the study of nuclear struc- 
tures. It should not be used for differential counting, since in 
films stained in this manner the neutrophile granules are invisible. 
Ehrlich 6 recommends this formula: 

Eosin (cryst.) .- 0.5 gm. 

Hematoxylin 2.0 gm. 

Absolute alcohol 100.0 c.c. 

Distilled water 100.0 c.c. 

Glycerin 100.0 c.c. 

Glacial acetic acid 10.0 c.c. 

Alum in excess. 
This mixture must "age" for several weeks before it can be used for staining. 

Technic 0} Staining. Specimens, fixed either chemically or 
by heat, are stained for from one-half hour to two hours, thor- 
oughly washed in water, dried, and mounted. In order to obtain 
the best results, it is advisable to filter the solution before using, 
and to wash the films very thoroughly after staining. 

If time is an object, the following rapid method may be substi- 
tuted for the above : The film is first stained for about five minutes 
with a 0.5 per cent, solution of aqueous eosin in 50 per cent, 
alcohol, washed, and dried in air; it is then counterstained for 
about one-half minute with Delafield's hematoxylin, 7 washed 
a second time, and mounted in the usual manner. 

STAINING WITH THIONIN. Thionin (also known as the "violet 
of Hoyer" and the "violet of Lauth") is an excellent stain for 

1 Zeitschr. f. wiss. Mik., 1894, vol. xi, p. 260. 

2 "Aetiologischeundklinische Malaria Studien," Berlin, 1890. 

3 Jour. Amer. Med. Assoc., 1898, vol. xxx, p. 303. 
* Med. Rec., 1903, vol. Ixiii, p. 1017. 

s Johns Hopkins Hosp. Bull., 1904, vol. xv, p. 122. 8 Loc. cit. 

7 This solution is made by first adding 4 gm. of hematoxylin crystals, dis- 
solved in 25 c.c. of alcohol, to 400 c.c. of a saturated aqueous solution of ammonia- 
alum. The mixture is left exposed to the sunlight and air in an uncorked bottle 
for four days, at the end of which time it is filtered, and mixed with 100 c.c. each 
of methyl alcohol and glycerin. This solution is allowed to stand until it becomes 
dark colored, when it is filtered and placed in a tightly corked bottle to "age" 
for at least two months before it can be used successfully for staining. Owing to 
the complicated manner in which Delafield's hematoxylin must be prepared, it is 
usually preferable to purchase it ready-made, from a dealer in microscopical sup- 
plies, Griibler's make being entirely reliable. 


blood parasites in general, being especially useful for the demon- 
stration of the malarial parasites and the filarial embryos. Thionin 
should not be used as a stain for films in which the general mor- 
phology of the blood cells is to be studied, since the basophile 
granules and the nuclei are the only histological elements for which 
it displays any decided affinity. Structures reacting toward the 
dye are stained violet of varying degrees of intensity. The follow- 
ing recipe, suggested by Futcher and Lazear, 1 will prove satis- 
factory : 

Thionin 0.3 gm. 

Absolute alcohol 10.0 c.c. 

One per cent, solution of carbolic acid q. s. ad 100.0 c.c. 

Technic of Staining. Films which have been fixed either 
chemically or by heat are stained in the above solution for from 
one to three minutes, being then washed in water, dried, and 
mounted as usual. The best results are obtained by using the 
French thionin, made by Cogit et Cie, of Paris. 

solution of methylene-blue and lithium carbonate affords a rapidly 
acting stain, excellent for the demonstration of the finer structure 
of the malarial parasite in every phase of its development. In 
addition to giving crisp, clear-cut pictures of the chromatin of this 
organism, the solution also brings out distinctly the granular 
degeneration of the erythrocytes, the nuclear characteristics of 
the erythroblasts and leucocytes, the basophile granules, and all 
ordinary bacteria. 

Technic of Staining. The films are fixed for fifteen seconds 
in methyl alcohol, rinsed in water, and then stained, unheated, 
for from one to two minutes, after which they are thoroughly 
washed in running water, dried without the use of heat, and 
mounted in balsam. Preliminary staining for ten or fifteen sec- 
onds with a o.i per cent, aqueous solution of eosin, followed by 
washing, gives a picture in which the contrast between the plasma 
and the basic elements of the cells is clearly differentiated. Poly- 
chrome methylene-blue, prepared according to Goldhorn's for- 
mula, 2 is sold by dealers in laboratory supplies, or it may be made 
in this manner: 

Two grams of methylene-blue are dissolved in 300 c.c. of 
warm water and 4 gm. of lithium carbonate are added, with 
constant agitation. The mixture is poured into an uncovered por- 
celain capsule, which is heated over a shallow water-bath for ten or 

1 Johns Hopkins Hosp. Bull., 1899, vol. x, p. 70. 

1 Ibid., 1899, vol. x, p. 70; also N. Y. Univ. Bull, of Med. Sci., 1901, vol. i, 


fifteen minutes, being frequently stirred with a glass rod. After 
removal from the water-bath the fluid is bottled, without filtering, 
and set aside for several days, after which its reaction is cor- 
rected by the cautious addition of a 5 per cent, acetic acid solu- 
tion until the dye is but very faintly alkaline. Should the solution 
become too alkaline after having been kept for some time, its reac- 
tion may be corrected by adding a small quantity of acetic acid, 
as in the preparation of the original mixture. 

A differential count of the leucocytes consists 
DIFFERENTIAL in determining, by microscopical examination of 
COUNTING, the stained specimen, the relative percentages of 
the different varieties of these cells, the estimate 
being based upon a count of several hundred cells, which are 
classified according to the several forms described in a following 
section (Section IV). This procedure, by means of which quali- 
tative changes affecting the leucocytes may be detected, is ob- 
viously a most important step in every blood examination, and 
one which should not be regarded as of secondary importance to 
the numerical estimate with the hemocytometer. 

The technic of differential counting consists simply in exam- 
ining successive microscopical fields until at least 500 leucocytes 
have been counted, the cells in each field of vision being identified 
as they appear, and jotted down on a piece of paper by the ob- 
server under their appropriate class. As soon as the requisite 
number of cells has been counted, the percentages of the different 
forms are calculated, to express the final result. For the exami- 
nation a T Vinch oil-immersion objective is practically indispen- 
sable, for to any but the skilled worker it is difficult, if not some- 
times impossible, to distinguish the various forms of leucocytes 
with a lower magnification than this lens provides. In order to 
be certain that each field is taken in accurate succession to its 
neighbor, the slide should be moved across the visual field by the 
aid of a mechanical stage; systematic examination of any given 
area of the specimen is well-nigh an impossibility if the slide is 
simply laid on, or clipped to, the stage of the microscope, and 
pushed across it with the fingers alone. 

If nucleated erythrocytes are found in the specimen, it is 
equally important to include them also in the differential count, 
classifying them into two histological divisions, normoblasts and 
megaloblasts. In calculating the number of these cells, it is 
obviously impossible to employ any direct method, so that the 
estimate must of necessity be more or less approximate, since it 
is based upon the ratio of erythroblasts to a given number of 
leucocytes. Having first counted the latter with the hemocytom- 


eter, the number of nucleated erythrocytes is noted in an area 
of the stained specimen in which a fixed number of leucocytes 
is contained, and having ascertained these data, the estimate is 
made according to the formula: 

Number of erythroblasts _ Number oj leucocytes 

counted in the stained film per Number of erythro- 

Number of leucocytes counted in the stained film blasts per 

For example, in a case of pernicious anemia in which the leu- 
cocytes number 4000 per, and a total of 35 erythroblasts 
is noted while counting 1000 leucocytes in the stained film, the 
calculation is as follows : 

35 X 4000 -5- 1000 = 140 erythroblasts per 

Whenever erythroblasts are found, it is important to determine 
their number to the of blood, and should normoblasts and 
megaloblasts both occur, to estimate the ratio between these two 
types of cells. 


Determann's method * of indirectly estimating the number of 
plaques to the of blood is both simple and accurate. It 
consists, briefly, in first determining the ratio of these elements to 
the erythrocytes, which are then counted, to furnish the basis for 
the final calculation. 

In obtaining the blood, a drop of the diluting fluid is placed 
upon the patient's finger and the puncture made through it, in 
order that the blood, as it flows from the puncture, will instantly 
mix with the diluent without coming in contact with the air. 
The blood and diluent are then thoroughly mixed for a few mo- 
ments by the aid of a cover-glass, after which a small portion of 
the mixture is transferred to a Thoma-Zeiss counting chamber, 
and the ratio of plaques to erythrocytes determined under the 
microscope. In the healthy adult this ratio, according to Deter- 
mann, ranges from i to 18 to i to 30, averaging about i to 22. 
With another drop of blood the erythrocyte count is then made 
by the usual method, and the actual number of plaques to the of undiluted blood calculated from the figure thus obtained. 
For example, in a given specimen of blood in which the ratio of 
plaques to erythrocytes is found to be i to 25, the count of the 
latter cells being 5,000,000, the actual number of plaques is there- 
fore 200,000 per 

1 Deutsch. Arch. f. klin. Med., 1899, vol. Ixi, p. 365. 


The diluents for which Determann expresses a preference are 
either a 9 per cent, aqueous solution of sodium chlorid to which 
a little methyl- violet has been added, or an aqueous solution con- 
taining one per cent, of sodium chlorid and 5 per cent, of po- 
tassium bichromate. Brodie and Russell 1 recommend equal 
parts of dahlia-glycerin and a two per cent, aqueous solution of 
sodium chlorid. 

Any of the diluting fluids already mentioned are also suit- 
able for the purpose. 


The use of centrifugal force for the purpose of determining the 
relative volumes of blood corpuscles and plasma was first applied 
in a practical manner by Hedin, 2 who embodied the earlier ideas 
of Blix in an instrument known as the hematocrit. More re- 
cently Daland, 3 by improving the mechanical construction of the 
original instrument and by simplifying the technic of using it, 
has made centrifugalization of the blood a method of investigation 
adapted to general clinical work. By the use of the hematocrit 
a pair of capillary glass tubes filled with undiluted blood are ro- 
tated in their horizontal axes at a high rate of speed until, as the 
result of the centrifugal force thus applied, the corpuscular and 
liquid portions of the blood become separated, the former being 
distinguishable in the lumen of the tube as a column the length 
of which is dependent upon the volume which the corpuscles con- 
stitute in relation to the rest of the blood mass. 

This instrument (Fig. 30) is composed of a 

DALAND'S set of cog-wheels inclosed in a metal box and 
HEMATOCRIT. geared in such a manner as to cause 10,000 
revolutions a minute of a vertical spindle, by 
turning a handle at a definite, uniform rate of speed. A metal 
frame, which may be securely fastened to the spindle by a modi- 
fied bayonet-lock, carries a pair of capillary glass tubes, each 
of which fits into two cup-like, rubber- lined depressions, and 
is adjusted and held in place by a spring. Each tube measures 
50 mm. in length with a lumen of 0.5 mm., and has engraved upon 
its outer surface a scale representing 160 equal divisions, the glass 

1 Jour. Physiol., 1897, vol. xxi, p. 390. 

2 Skandinavisch. Arch. f. Physiol., 1890, vol. ii, p. 134. 

3 University Med. Mag., 1891, vol. iv, p. 85; also Edwards' supplement to 
Keating's " Cyclopedia of the Diseases of Children," Philadelphia, 1899, vol. v, 
P- 537- 


immediately above the scale being molded so as to form a lens- 
front, to magnify the column of blood and to facilitate the reading 
of the divisions. A bit of rubber tubing fitted with a mouthpiece 
is used for rilling the capillary tube, in the same manner in which 
the blood is measured with the hemocytometer. While in use, 
the instrument is securely attached to the projecting edge of a 
table or shelf by means of a clamp operated by a thumb-screw. 
Method o] Use. Having cleaned and punctured the patient's 
finger in the usual manner, the beveled end of one of the capil- 
lary tubes is immersed in the drop of blood, which is sucked up 
the lumen of the tube until it is exactly filled. The forefinger, 

smeared with a little vaselin, is then 
applied to the beveled end of the tube, 
while the rubber suction tube is care- 
fully removed by twisting it free not 
by forcibly pulling it off, since this may 
accidentally cause removal of a portion 
of the blood column by suction. The 
tube thus charged with blood is at once 
adjusted to one arm of the frame, and 
the empty tube similarly fixed in the 
other arm, to equalize the balance, this 
step being completed as rapidly as pos- 
sible, in order to anticipate coagulation. 
When the tubes have been- thus ad- 
justed, and the frame securely locked 
in the spindle, the handle of the in- 
strument is turned for three minutes 1 at 
the rate of 77 revolutions a minute, this 
rate of speed securing 10,000 rotations 
a minute of the frame, since the latter 

revolves 130 times with each complete turn of the handle. The 
centrifugalization having been finished, the charged tube is care- 
fully removed from the frame, and held against a piece of dull 
white paper, so that the height of the blood column may be easily 
determined. In order to make the reading with accuracy, it is 
sometimes necessary to use a small magnifying glass, for the 
divisions on the scale of the tube are but 0.5 mm. apart a 
distance too small to judge easily with the naked eye. On ex- 
amination, three distinct divisions of the lumen of the tube 
containing the centrifugalized blood may be distinguished : first, 

1 In a recent personal communication Dr. Daland advises that, in order to 
insure the most accurate results with his instrument, the centrifugalization be con- 
tinued for three, instead of for two, minutes, as he formerly recommended. 



a dark-colored column consisting of erythrocytes, reaching, in 
normal blood, to a point between the divisions marked 50 and 51 ; 
second, a thin layer of leucocytes, showing, in blood in which 
these cells are not largely increased, as an indistinct, milky stratum 
overlying the erythrocytes ; and, third, a layer of clear plasma 
occupying the remainder of the lumen. The normal volume of 
erythrocytes being arbitrarily regarded as 100 per cent., to com- 
pute this result the figure of the scale to which these cells rise is 
multiplied by two. Unless the leucocytes are greatly increased 
in number, the layer formed by these cells is too delicate and too 
dully defined to be read with any degree of accuracy ; but in 
cases of high leucocytosis and of leukemia it is quite possible to es- 
timate roughly the relative proportions of leucocytes to erythro- 

The capillary tube which has been filled with blood should be 
cleaned as soon after use as possible, water, followed by alcohol 
and ether, being used for this purpose. A fine wire should be 
passed through its lumen, to dislodge any obstruction which may 
result from drying of the column of closely packed corpuscles. 

The hematocrit, if its clinical application is limited to the de- 
termination simply of the relative volumes of the blood corpus- 
cles and plasma, may be relied upon to furnish, on the whole, 
dependable information, the necessary errors attending its use 
probably being within two per cent. It is also useful in the study 
of hemoglobinemia, cholemia, and lipemia. If employed in the 
role of a hemocytometer, however, its results must needs be highly 
inaccurate, just, in those instances in which exact methods of 
investigation are all important. It is true that in normal blood, 
in which the size of the corpuscles ranges within the physiologi- 
cal limits, it is correct to consider each percentage volume as 
representing approximately a count of 100,000 erythrocytes per In blood characterized by any considerable deformity 
in the size and shape of these cells, as in high-grade anemia 
or in leukemia, it is perfectly obvious that no such corres- 
pondence between the count and the percentage volume can be 
expected blood in which microcytosis is pronounced is certain 
to show a lower percentage volume of erythrocytes than blood in 
which megalocytosis prevails, or than blood containing normal- 
sized cells, although the counts of all three may be identical. Simi- 
larly, a given number of lymphocytes should indicate a lower per- 
centage volume than an equal number of myelocytes, or even 
polynuclear neutrophiles. On account of these sources of fallacy, 
if for no other reason, the hematocrit estimate should never be 
taken as a basis for calculating the count in pathological condi- 


tions, in lieu of the more accurate, if more laborious, method of 
counting the corpuscles. 

Capps 1 considers that the hematocrit may be used to advan- 
tage in conjunction with the hemocytometer, in determining the 
actual size or volume of the individual erythrocyte, and he regards 
this method as far more reliable than the use of the micrometer, 
since with the latter only the transverse diameter of the cells, and 
not their depth, can be measured. The formula for calculating 
this " volume index " has been given elsewhere. (See p. 173.) 


This method of investigation is used as an indirect means of 
computing the percentage of hemoglobin, owing to the more or 
less constant parallelism maintained between it and the specific 
gravity of the whole blood. The correspondence between the two, 
together with the sources of error inseparable from the test, has 
been pointed out in another section. (See p. 133.) 

Hammerschlag's modification 2 of Roy's 
HAMMER- method 3 of determining the specific gravity of 
SCHLAG'S the blood best serves the purpose of those who 
METHOD. choose this roundabout means of approximating 
the hemoglobin percentage. It consists in first 
making a mixture of benzol and chloroform of such a specific 
gravity that a small drop of blood deposited in the liquid remains 
suspended, after which the specific gravity of the mixture is de- 
termined with a hydrometer, the figure thus obtained representing 
the density of the blood used in the test. The hemoglobin per- 
centage corresponding to this figure is then selected from a table 
giving the various degrees of blood densities and the percentages 
of hemoglobin to which they are equivalent. 

The apparatus required for making the test includes a hy- 
drometer provided with a scale graduated to 1.070, a hydrometer 
jar having a wide, substantial base, a glass capillary tube, and a 
glass stirring rod. An ordinary urinometer may be used instead 
of a special hydrometer, since specific gravities in excess of 1.060 
(the highest gradation on the scale of most urinometers) are not 
often encountered. More accurate results, however, are possible 
with a standardized instrument. Levy 4 has shown that with an 

1 Jour. Amer. Med. Assoc., 1900, vol. xxxvi, p. 464. 

2 Zeitschr. f. klin. Med., 1892, vol. xx, p. 444. 

3 Cited by Devoto, Zeitschr. f. Heilk., 1889, vol. xi, p. 175. 

4 Lancet, 1903, vol. i, p. 1302 


ordinary hydrometer an excessive reading, ranging from 3 to 10 
degrees, always occurs, owing to the disturbing effect upon the in- 
strument of the low surface tension of the chloroform-benzol mix- 
ture. To obtain accurate figures it is necessary to use a hydrome- 
ter which has been standardized to these reagents. Either a 
Thoma-Zeiss leucocytometer or a medicine dropper, the free end 
of which should be heated in a flame and bent into an obtuse 
angle, will serve as a capillary pipette. 

Benzol and chloroform are mixed together in the hydrometer 
jar in such proportions that the specific gravity of the liquid is 
approximately equal to that of normal blood, 1.060. This mix- 
ture having been made and its specific gravity taken, the point of 
the capillary pipette, charged with blood, is plunged beneath the 
surface of the liquid and a small bead of blood gently expelled. 
If the blood drop rises to the surface of the mixture, a few drops 
of benzol are added, while if it sinks to the bottom of the jar, 
chloroform is used, the addition of the appropriate reagent being 
continued until the drop neither rises nor sinks, but remains sta- 
tionary, suspended in the mixture. When this point has been 
determined, the specific gravity of the liquid is taken by means of 
the hydrometer, this figure obviously representing the specific 
gravity of the blood drop itself. To convert the specific gravity 
into its hemoglobin equivalent the figure obtained by the above 
procedure is compared with one of the tables given on page 133. 
After each addition of benzol or of chloroform the contents of the 
jar must be thoroughly mixed by stirring with the glass rod, in 
order to secure uniformity in the density of the liquid. The latter, 
if it is filtered free from blood and preserved in a tightly stop- 
pered bottle, may be used again in subsequent tests. 

In spite of the enthusiasm evinced by certain authors for this 
method of obtaining hemoglobin values, considerable experi- 
ence with the test has convinced the writer that it is both crude 
and untrustworthy it is useful, no doubt, when a hemometer 
cannot be obtained, but in no sense is it an efficient substitute for 
colorimetric methods. The liability of the blood drop to split 
up into numerous fine particles, to adhere to the inside of the 
jar, and to become altered in composition from the influence of the 
reagents, as well as the tedious attempts which must usually be 
made to add just the proper quantities of benzol and chloroform 
to secure a mixture in which the drop neither sinks nor rises, are 
a few of the drawbacks which must make the test unpopular with 
busy clinicians. For a critical review of the clinical value of 
Hammerschlag's test Baumann's article 1 should be consulted. 

1 Brit. Med. Jour., 1904, vol. i, p. 473. 



A convenient clinical method of deter- 
ENGEL'S mining the alkalinity of the blood is by the use 
ALKALIMETER. of Engel's alkalimeter (Fig. 31). By means 
of this instrument a measured quantity of fresh 
blood is diluted with distilled water in the proportion of one to 
ten, and then titrated with a -j-% normal solution of tartaric acid 
until the mixture reacts with lacmoid paper, the total alkalinity 
being calculated from the amount of the tartaric acid used. The 
methods of alkalinity estimation devised by Landois, 1 by Lieb- 
reich, 2 by Haycraft and Williamson, 3 by Wright, 4 and by Kraus 5 
are not well adapted to routine blood work, being either too com- 
plicated and elaborate for such a purpose or too inaccurate. 

The apparatus which Engel has devised consists of the follow- 
ing parts: a diluting and mixing pipette, resembling a large-sized 
Thoma-Zeiss erythrocytometer; a graduated burette by means of 
which the amount of tartaric acid solution used in the test is meas- 
ured; a glass cylinder in which the titration is made; and a glass 
stirring rod. The mixing pipette is graduated in three principal 
divisions marked 0.025, -5> an d 5-o respectively, the first two 
divisions being further scaled in tenths by fine horizontal mark- 
ings; otherwise the instrument is modeled like a blood-counting 
pipette. The burette has a capacity of 5 c.c., and is provided with 
a scale indicating 100 equal divisions; when in use, it is clamped up- 
right, by means of a special attachment, to a vertical brass support 
which screws into a fitting in the box containing the apparatus. 

Method of Use. The technic of using the alkalimeter is simple 
and time saving in comparison with that required by other well- 
known methods of alkalinity testing. Finger-blood, obtained by a 
rather deep puncture, so as to afford a good-sized drop, is sucked 
up in the pipette until it reaches the mark 0.05, immediately after 
which distilled water is similarly drawn up the lumen of the tube 
until the mixture of blood and water fills the bulbous expansion 
and reaches the mark 5.0 in the constricted portion beyond. 
While sucking up the water the pipette should be rapidly twisted 
to and fro between the thumb and forefinger, to insure thorough 
mixing of the blood and water as they together fill the expanded 
portion of the instrument. As soon as the dilution has been 
made, the pipette should be shaken for a minute or so, until the 

1 Real-Encyclop., 1885, vol. iii, p. 161. 

2 Berichte d. deutsch. chem. Gesellsch., 1868, vol. i, p. 48. 

3 Proc. Roy. Soc., Edinburgh, June 18, 1888. 

4 Lancet, 1897, vol. ii, p. 719. 5 Zeitschr. f. Heilk., 1889, vol. x, p. 106. 



mixture becomes of a uniform "laky" tint, which indicates that 
all the hemoglobin has been dissolved from the corpuscular 
stroma. The contents of the pipette are blown out into the glass 
cylinder, which is placed beneath the faucet of the burette, the 
latter having been previously filled to the mark o with a 7 V nor- 
mal solution of tartaric acid. By turning the stop-cock of the 
burette the test solution is now added, drop by drop, stirring be- 
tween each addition, to the measured amount of diluted blood 
in the cylinder. From time to time a drop of the mixture is re- 


moved by means of the glass rod and tested with the lacmoid 
paper, the titration being continued until the reaction, recognized 
as a bright-red halo which forms around the edge of the drop, is 
obtained. The titration is then stopped, and a note made of the 
number of drops of the test solution which have been used. In 
normal blood the writer finds that from 9 to u drops are re- 
quired to give the reaction. The estimate of the total alkalin- 
ity of the blood is made by multiplying by the figure 53.3 the 
number of drops of the tartaric acid solution used, according to 


the formula, 10 : a :: 533.0 :.r, a representing the drops of the 
reagent. 1 The result thus obtained is expressed in milligrams of 
NaOH per 100 c.c. of blood. The following table may be useful 
for reference in determining the various degrees of alkalinity: 

If 6 drops of the solution are used, the alkalinity equals 319 mgm. NaOH 






43 6 


6 39 

After use the pipette should be thoroughly washed out with 
water, alcohol, and ether, and then dried, in the manner already 
directed for cleaning the Thoma-Zeiss instrument. 

This ingenious instrument 2 is used in con- 

DARE'S nection with a spectroscope, its principle depend- 

HEMO-ALKALIM- ing upon the dissipation of the absorption bands 

ETER. of oxyhemoglobin when exact neutralization of 

the blood ensues, after the addition of a 2^V nor ' 

mal solution of tartaric acid. The working parts of the apparatus 

are explained by the accompanying illustration (Fig. 32). The test 

solution to be employed is made up as follows: 

Acid, tartaric. (Merck's reagent) .................... gr. j (0.075 

Alcohol .......................................... Z v (20 c.c.) . 

Aqua destil ................................... q. s. 3 vj (200 c.c.). 

Method of Use. The alkalimeter tube (Fig. 32, A), fitted with 
its blood pipette, B, is held horizontally, and the pipette filled 
with blood by presenting its exposed end to the drop as it flows 
from the puncture. With an ordinary medicine dropper containing 
distilled water and coupled to the blood pipette by a bit of rubber 
tubing, the measured blood is washed into the. alkalimeter tube 
until the point o is reached. Now, with the thumb closing the 
opening, C, in its bulb, the alkalimeter tube is inverted several times, 
in order to mix the blood and water. The reagent pipette, D, is 
filled with the test solution, connected by tubing with the free end 

1 Assuming that 0.5 c.c. of tartaric acid is used to neutralize 0.05 c.c. of blood, 
therefore for every 100 c.c. of blood 1000 c.c., or one liter, of a -^ normal solution of 
tartaric acid are required. As the alkalinity of the blood is not expressed by the 
amount of acid necessary to saturate it, but in milligrams of an alkali, sodium hy- 
drate, the calculation is made thus: as the equivalent weight of tartaric acid is 75, 
and that of sodium hydrate 40, one liter of water dissolving 75 gm. of the former 
saturates 40 gm. of the latter that is, one liter of a 7 V normal tartaric acid solution 
saturates if gm., or, in other words, 533 mgm., of sodium hydrate, this figure 
being taken by Engel as the degree of normal alkalinity of the blood. 

2 Phila. Med. Jour., 1903, vol. xi, p. 137. 



of the blood pipette, and compressed so as to force the reagent 
into the diluted blood within the alkalimeter tube. In doing this 
it is necessary to thumb the opening in the bulb of the latter, to 
avoid the mixture of the blood and reagent. The tube and reagent 
pipette, still connected, are now grasped and inverted, to mix thor- 
oughly the contents of the former, after which a preliminary 
observation is made by adjusting the lower part of the tube 
(below the mark o) to the cleft of a Browning pocket spectroscope. 
If the oxyhemoglobin absorption bands per- 
sist, more of the reagent is added, drop by 
drop, inverting the tube between, and exam- 
ining with the spectroscope after, each drop. 
When the bands disappear, the figure to which 
the neutralized blood solution reaches is noted, 
and compared with the table of equivalents given 
below, to obtain the final result, which is ex- 
pressed in milligrams of NaOH per 100 c.c. 
of blood. The figure for normal blood with 
this instrument is 266. 



REAGENT. 100 c.c. OF BLOOD. 

2-6 345-0 

2.4 3 I 9- 

2.2 292.0 

2 .0 ". 266.O 

1.8 239.0 

1.6 2I2.O 

1 .4 1 76.0 

1.2 I59-O 

1-0 I33- 

0.8 96.0 

0.6 79.0 

0.4 53.0 

0.2 26.6 FIG 

Alkalimeter tube; B, 
automatic blood pi- 
pette; C. opening for 
the admission of air: 
D, reagent pipette. 

While, up to the present time, it cannot be 
claimed that information of any real diagnostic 
pertinence has been obtained from the study of 
the alkalinity of the blood, this procedure should 
prove of value in the systematic investigation of many cases, 
especially those of high-grade anemia. As elsewhere mentioned, 
the degree of normal blood alkalinity varies greatly according to 
the particular method by which this figure is ascertained, so that 
it follows that the results obtained by means of one apparatus 
cannot be compared with those based upon investigation with 
another instrument of different design. 



This procedure is useful in the study of such conditions as 
icterus, purpura, hemophilia, scurvy, various specific infections, 
and other diseases characterized by abnormalities in the clotting 
time of the blood. Knowledge of how long it takes a given blood 
to clot not only adds completeness to the clinical history, but 
allows the physician to judge of the effect of remedies administered 
with a view to promoting coagulation. No careful surgeon to- 
day neglects to test the patient's blood coagulability before operat- 
ing for the relief of obstructive lesions of the biliary passages. No 
physician should omit to make repeated 
coagulation tests in treating one of the 
hemorrhagic diatheses or a primary ane- 

The coagulation time 
GLASS SLIDE may be determined ap- 
L METHOD. proximately by collecting 

I & several individual drops 

^ I of blood of the same size upon the surface 

of a perfectly clean, slightly warmed glass 

slide. At regular intervals of about one 
minute a straw of a whisk-broom is lightly 
trailed through each drop in succession, 
until sooner or later a delicate thread of 
fibrin may be observed clinging to the 
straw. The period which has elapsed 
between the deposit of the blood on the 
slide and the appearance of this indica- 
cinvex drop TC CoAGCLATION - tion of clotting is expressed in minutes, 
to represent the coagulation time of the 

specimen under investigation. Normal blood thus treated coagu- 
lates in from two and one-half to five minutes. 

The method described by Milian 1 is perhaps even simpler. A 
rather large drop of blood is collected upon the center of a clean, 
dry glass slide, which, after the lapse of a minute or so, is carefully 
tilted to a vertical plane. With the slide held in this position the 
profile of the coagulated drop forms a symmetrical, mound-like 
convexity, while that of the incompletely clotted drop is tear- 
shaped (Fig. 33). The time elapsing between the collection of 
the blood drop and the first-named change is considered the coagu- 
lation time. Five minutes is the average period for normal blood. 

1 Presse med., 1904, vol. i, p. 202. 


The latest model of this instrument 1 consists 

WRIGHT'S of a tin reservoir fitted with a removable rack 

COAGULOMETER. holding a thermometer and a set of twelve 

calibrated glass coagulation tubes. For filling 

the latter an aspirator tube with a rubber connection is employed. 

The thermometer indicates degrees of the Centigrade scale, and is 


also graduated at 18.5 and at 37, the temperature of "half blood 
heat" and of "blood heat," respectively. The coagulation tubes, 
made of stout glass, are furnished with tight rubber caps, used to 
seal their blunt ends when immersed; they have an internal di- 
ameter of 0.25 mm., and are marked to indicate a blood column 
of 5 c.c. Each tube is numbered to correspond with its appropriate 
place in the rack of the reservoir. 

1 Lancet, 1902, vol. ii, p. 15. 


Method oj Use. The reservoir is filled with water having a 
temperature of 18.5 C., 1 and the tubes, each fitted with a rubber 
cap, are placed in the rack, sealed end downward, w r here they are 
allowed to remain immersed for a few minutes until they acquire 
approximately the same temperature as that of the water. They 
are then removed, dried, and stripped of their rubber caps. 

Having then pricked the patient's finger, 5 c.c. of blood 
are sucked up each tube, at successive intervals of one min- 
ute, a tube as soon as filled being returned to its appropriate 
place in the reservoir. It is not necessary to replace the caps 
the tubes are simply placed point downward in the water, the 
temperature of which is from time to time readjusted to the stand- 
ard by adding hot water to the reservoir as its contents cool. 
After the lapse of an appropriate interval the tube first filled is re- 
moved from the reservoir and tested by attempting to blow out its 
contents upon the surface of a piece of white filter or blotting paper. 
The other tubes are similarly tested in rotation, at intervals of one 
minute or less, until, after having thus tried a variable number 
one is found from which the blood cannot be expelled. 

Coagulation may then be considered to have occurred, the 
time required for this process being expressed by the number of 
minutes elapsing between the filling of the tube in question and 
the evidence of clotting thus demonstrated. With normal blood 
the coagulation time, as determined by this instrument, gener- 
ally ranges from about three to six minutes. In rare instances 
Wright 2 found that in the healthy male adult clotting may be pro- 
longed for fifteen minutes. 

After use, a fine wire should be forced through the lumen of 
the tubes to dislodge the clots, after which the remaining traces of 
blood are to be removed by thorough washing with distilled water, 
alcohol, and ether, in the order named. 


Cryoscopy is the method' of determining the 
CRYOSCOPY. freezing-point of a given liquid and the com- 
parison of this figure with the freezing-point of 
distilled water. This method of investigation is based upon the 
principle that the freezing-point of any liquid is proportionate to the 
number of its contained molecules. The greater the molecular 
concentration, the lower the point at which the fluid freezes. 

1 In slowly coagulating bloods the temperature should be about that of blood 
heat, 37 C. 

2 Brit. Med. Jour., 1902, vol. ii, p. 1706. 


Cryoscopy of the blood and urine was first applied to clin- 
ical medicine by von Koranyi, 1 who thus was able to determine 
the status of the renal function. The average freezing-point of 
normal blood is 0.57 C., ranging between 0.56 and 0.58 
C. Urine in the healthy individual freezes between 0.9 and 
2 C. Venous blood freezes at a slightly lower temperature 
than arterial. In renal disease with decided insufficiency of the 
kidneys it is found that the freezing-point of the blood is lowered, 
while that of the urine is correspondingly raised, owing to the 
retention in the blood of matter which the crippled kidneys are 
unable to excrete. Excision of one kidney does not disturb the 
normal freezing-points of the blood and urine, but double neph- 
rectomy promptly lowers that of the former and raises that of the 

A freezing-point of 0.6 C. or lower for blood and of i C. 
or higher for urine indicates a sufficient degree of renal impair- 
ment to contraindicate surgical interference in lesions of the kid- 
neys. This dictum was first expressed by Kummel, 2 who bases it 
upon an experience of 170 cases of renal surgery in which cryos- 
copy was practised, including chronic nephritis, nephrolithiasis, 
tuberculosis, cysts, neoplasms, pyonephrosis, hydronephrosis, and 
post-operative anuria. Rumpel's extensive researches 3 in 300 
similar cases showed that in normal individuals and in those 
with unilateral renal lesions the blood freezing-point did not vary 
from the figures given above (: 0.56 to 0.58 C.), while in 
bilateral lesions of the kidneys it ranged between 0.55 and 
0.81 C. In renal disease unaccompanied by uremia, Linde- 
mann 4 found no variations from the normal freezing-point, but 
with the onset of this complication the blood froze at a point as low 
as 0.7 C. 

Koeppe's data 3 in conditions other than those involving the 
kidneys are of interest. He found, in a series of cases including 
functional neuroses, carcinoma, diabetes, pleurisy, and pneumonia, 
a blood freezing-point varying from 0.5 to 0.63 C. Similar 
figures, of theoretical rather than practical interest, have also been 
determined in various general diseases by Ogston, 6 Tinker, 7 Tieken, 3 

1 Zeitschr. f. klin. Med., 1897, vol. xxxiii, p. .45; ibid., 1899, vol. xxxiv, p. i; 
Berlin, klin. Wochenschr. , 1901, vol. xxxviii, p. 424. 

2 Centralbl. f. Chir., 1902, vol. xxix, p. 121. 

3 Munch, med. Wochenschr., 1903, vol. 1, pp. 19, 67, and 117. 

4 Deutsch. Arch. f. klin. Med., 1899, vol. Ixv, p. i. 

5 Cited by Cattell, Internat. Clinics, 1904, vol. i, p. 6. 
8 Lancet, 1901, vol. ii, p. 1253. 

7 Johns Hopkins Hosp. Bull., 1903, vol. xiv, p. 162. 

8 Med. News, 1904, vol. Ixxxiv, p. 416. 


Ceconi, 1 Sollmann, 2 and others. 3 It is of interest to note that the 
pregnant woman's blood freezes at a higher temperature than nor- 
mal, but that the latter figure is reached after delivery, as the 
molecular concentration of her blood rises. In diabetes mellitus 
the freezing-point is low, but in pernicious anemia and in various 
forms of hydremia it is high. Inhalations of oxygen lower the 
figure, but in conditions of cyanosis with an excess of carbonic 
acid in the blood the freezing-point rises. It is of value also to 
observe that cystitis and pyelitis do not affect the blood's freezing- 

Carrara 4 and Revenstorf 5 found that in cases of drowning the 
resulting dilution of the blood causes radical deviations from 
the normal freezing-point, which rises in the case of drowning 
in fresh water and falls after death by submersion in salt water. 
In this connection the general statement applies, that the freezing- 
point of the diluted blood approaches that of the fluid with which 
the body is waterlogged. Whether the deviation be plus or minus, 
it is generally more marked in the blood of the left than of the 
right heart, because of the greater dilution of the venous blood, 
owing to the water drawn into the lungs and entering the pulmon- 
ary capillaries. Exceptionally this difference between the halves 
of the heart is not apparent, as in the case of bodies remaining 
submerged for a long period and in those in which the circulation 
persists for a few moments after the blood dilution, in either of 
which instances the entire blood mass tends to become equally 
diluted. The utility of cryoscopy as a forensic test of death by 
drowning is obvious from these experiments, although its value 
is to some extent restricted by the fact that advanced putrefaction 
and prolonged submersion interfere with its reliability. 

Cryoscopy of the blood and urine is of value in determining 
the condition of renal adequacy, but it should be supplemented by 
other laboratory tests devised for this purpose. Forensically, the 
test may yield reliable evidence not only in cases of death by 
drowning, but possibly also in the identification of blood stains 
from various sources. 

The cryoscope made by Fontaine, of Paris 
FONTAINE'S (Fig. 35), is simply constructed, durable, and 
CRYOSCOPE. thoroughly satisfactory for clinical use. It con- 
sists of a stout glass freezing-jar, A, provided with 
a large test-tube, B, passing to its center and kept in position by a 

1 Rif. med., 1901, vol. iii, p. 109. 2 Amer. Med., 1902, vol. iii, p. 656. 

3 For an excellent resum of cryoscopy in all its phases see Cattell, Proc. Phila. 
Path. Soc., 1904, vol. vi, p. 244. 

4 Arch. Ital. de Biol., 1901, vol. xxxv, p. 349. 

8 Munch, med. Wochenschr., 1902, vol. xlix, p. 1880. 


metal support, C. At the base of the jar there is a drain, D, for 
the liquid which accumulates as the ice-salt mixture melts. A 
small test-tube, E, having a lateral vent, F, fits within the larger 
tube, being adjusted by means of a rubber collar, G, 1 so that be- 
tween the two tubes an air chamber is formed. A thermometer, 
H, encircled by a metal spiral stirrer, I, is let down into the 
smaller test-tube and adjusted so that it touches neither the walls 
nor the bottom of the latter, being kept in this position by means 
of a vertical standard fitted with an adjustable horizontal arm. 
The thermometer registers from 3 C. to +3 C., being gradu- 
ated in Tj-J-g- of a degree, and is provided with a pear-shaped bulb 
at the top, to allow for the expansion of 
the mercury column. The thermometer 
is an extremely delicate and expensive 
bit of apparatus, and must be handled 
with care, for fear of breakage. It 
should be tested with distilled water, 
so that any deviation may be taken into 
account in subsequent observations. 

Method of Use. The freezing- jar, 
with its large test-tube adjusted, is filled 
to the brim with a mixture of cracked 
ice and rock-salt, packed in alternate 
layers, the whole being covered, at the 
level of the mouth of the jar, with a 
layer of salt an inch in depth. The size 
of the bits of ice should be large enough 
to insure gradual thawing, for finely 
crushed ice rapidly turns to slush. Ten 
c.c. of the blood or urine 2 are placed in 
the small test-tube, which is laid against 
a block of ice, to cool, while the freez- 
ing-jar is being packed. By the time 

this is accomplished (about five minutes) the fluid to be tested 
will have cooled sufficiently, and the next step in the test may 

1 In the original model of the instrument this collar, as well as the metal band 
supporting the large test-tube, interferes with the reading of the thermometer 
scale, but this defect may be easily remedied by cutting in each a small window, 
so as to allow a clear view of the mercury column. The apparatus is made by 
G. Fontaine, 16, Rue Monsieur le Prince, Paris; it costs, duty free, 90 francs. The 

'A. H. Thomas Company, Philadelphia, makes an excellent cryoscope of the Fon- 
taine model, the cost of which is considerably less than that of the French instru- 

2 The patient's blood and urine should be collected at the same time: the 
former by aspirating a superficial vein, the latter by catheterizing each ureter if 
the test involves a determination of the integrity of each kidney. 



proceed. Great care must be taken that both the test-tubes 
and the thermometer are absolutely dry, for the slightest trace 
of moisture so alters the freezing-point that gross inaccuracies 
in the final reading may result. The small test-tube is now 
fitted within the larger one, after which the thermometer, with 
the stirrer in place, is carefully lowered into position, rest- 
ing free from contact with the walls and bottom of the tube, 
with its mercury bulb immersed in the test fluid. The ther- 
mometer, when correctly adjusted, is hung from the arm of the 
standard, placed alongside the freezing-jar. The handle of the 
stirrer is now constantly moved up and down, so as to equalize the 
temperature of the test fluid as it congeals, and this mixing is to be 
continued intermittently during the rest of the observation. After 
a wait of about five minutes the column of mercury begins to fall, 
first very slowly, then rapidly, to approximately two degrees 
below zero, at which point it remains for a few moments, and then, 
because of the heat evolved, rises to the true freezing-point, where 
it remains stationary for about two minutes, after which it falls to 
the temperature of the outside mixture of ice and salt. When the 
point of stability is attained, the degree registered by the mercury 
column is noted to obtain the freezing-point of the specimen. In 
making this end-observation the eyes should be on a level with the 
top of the mercury column. It may be hastened somewhat by the 
insertion of a pellet of ice in the vent of the small test-tube just 
before the freezing-point is reached. 


This method of examination is of more than mere theoretical 
value in the study of diseases associated with hemoglobinemia, 
in which conditions it indicates, and with great accuracy, the vul- 
nerability of the erythrocytes, as expressed by their resistance to the 
action of salt solutions of different strengths. Among the patho- 
logical conditions in which the method is useful may be named the 
severe anemias, the specific infections, such as malarial fever, 
yellow fever, and sepsis, all forms of icterus, the different cachectic 
states, paroxysmal hemoglobinemia, and toxemias due to snake 
venom and to other hemolytic agents. 

Hamburger's method, as modified by von Lim- 

HAMBURGER'S beck, 1 requires the use of twelve small glass recep- 

METHOD. tacles, about the size of a Gowers' hemocytometer 

mixing cell, each of which contains a small glass 

1 "Einer klinische Pathologic des Blutes," 26. ed., Jena, 1896 ; also New Sy- 
denham Soc. trans, by Arthur Latham, London, 1901. 



bead. Into these vessels is placed i c.c. of salt solutions of dif- 
ferent strengths, each differing by 0.02 per cent., the minimum 
being 0.3 per cent., the next 0.32 per cent, (or 0.02 per cent, 
stronger), and so on. A drop of blood as it drips from the puncture 
is allowed to fall into each of the vessels, which are then shaken 
briskly for a minute or so, in order to cause defibrination by the 
whipping about of the glass beads. When this is accomplished, 
the blood-charged solutions are allowed to stand for six hours, 
when it will be noted that some of them are tinged with hemoglobin, 
while others remain clear. The first tube showing no solution of 
hemoglobin indicates the isotonicity of the cells under examination. 
For normal blood this index ranges between 0.46 and 0.48 NaCl. 


For clinical work the Sorby-Beck microspectroscope, to be used 
in connection with the microscope, is an excellent instrument, 
being both accurate and, comparatively 
speaking, easy to manipulate. Other 
very perfect instruments for the spec- 
troscopical examination of the blood, 
differing but little from the original 
Sorby model, are also made by Zeiss, 
by Leitz, and by Browning. 

This instrument 
THE (Fig. 36) when in use 

SORBY-BECK fits into the tube of 

MICROSPEC- the microscope like an 
TROSCOPE. ordinary ocular, for 
which it is substituted. 
Its essential part consists of a tube, A , 
in which a series of five prisms, two of 
flint and three of crown glass, is ar- 
ranged in such a manner that the emer- 
gent rays, which are separated by dispersion, leave the prisms in 
practically the same direction as that taken by the entering im- 
mergent ray. At one side of the tube is fixed a right-angle 
reflecting prism, so that the spectrum of a solution of normal 
blood may be thrown alongside that of the specimen under 
investigation, the two spectra thus being comparable. The ad- 
justment of the spectra is effected by means of the two small 
screws, B, B' . The receptacle containing the control solution 
of blood is clamped to the stage, C, by a spring clip, D, light being 
reflected through the liquid and into the rectangular aperture, , 



by the swinging mirror, F. The width of this aperture is controlled 
by the screw, G. The receptacle containing the blood solution to 
be examined is placed upon the stage of the microscope, being 
brought into focus with a low-power (f or i inch) dry objective. 
Beneath the tube inclosing the series of prisms is mounted an 
achromatic ocular, below which a narrow, slit-like diaphragm is 
situated, the vertical size of this opening being regulated by a 
milled screw, not shown in the illustration, and its breadth by the 
two small levers, /, /'. Both ocular and prisms may be moved 
simultaneously toward and away from the diaphragm, by a rack- 
and-pinion mechanism controlled by the wheel, /, so that any part 
of the spectrum may be brought into focus. 

The liquids to be examined should be placed in Sorby's tubular 
cells, and cover-glasses superimposed. These cells (Fig. 37) are 
narrow-lumened glass receptacles made of barometer tubing, both 
ends of which are accurately ground to parallel surfaces, one end 
being cemented to a small polished glass plate. 

Method of Examination. The specimen of blood, obtained in 
the usual manner, by puncture, is first diluted 
with distilled water 100 times by means of the 
Thoma-Zeiss erythrocytometer, and sufficient 
of this laked blood dropped into a Sorby cell 
FIG. 37. SORBY TUBU- to fill it exactly to the brim. A cover-glass is 
then carefully laid over the open end of the 
cell, the precaution being taken to prevent the 
formation of air-bubbles upon the surface of the column of liquid 
thus inclosed. A second cell, to be used as the control, is filled 
with normal blood, similarly diluted, and both are then adjusted 
in their respective positions, as already explained. 

In making the examination a ray of artificial light (that from 
a Welsbach incandescent burner being most suitable) is projected 
by the microscope mirror through the lumen of the cell contain- 
ing the suspected blood, and the surface of the liquid focused 
with an ordinary ocular. The latter is' then removed from the 
microscope tube and replaced by the spectroscope ocular, and 
the second spectrum, that of the normal blood, is brought into 
proper position alongside that of the first, so that any differences 
between the two may be contrasted by the observer. 

The appearance of the spectra of normal and of pathological 
blood, together with the circumstances under which the latter 
occur, has been described in another section. (See p. 168.) 



The demonstration of bacteria in the circulating blood, pro- 
vided that faultless technic is employed, furnishes in some in- 
stances a diagnostic sign of the greatest importance. The patho- 
logical significance of such a finding is much greater than that 
of a similar result obtained postmortem, since with the latter 
there is no means of determining whether the bacterial invasion 
of the blood current took place during the active stages of the 
disease, or whether it occurred as either a preagonal or a postag- 
onal process. 

Cultural methods with blood aspirated directly 

METHODS, from a superficial vein should invariably be used 
whenever such a procedure is practicable, for 
blood obtained simply by pricking the skin is most likely to be 
contaminated with various bacteria which have their normal hab- 
itat in the epidermis and its appendages, notably by the Staphyl- 
ococcus epidermidis albus. Welch, 1 who first drew attention to 
this source of error, emphasizes the fact that no diagnostic sig- 
nificance should be attached to the demonstration of this bac- 
terium in blood obtained by puncture of the skin. 

Direct examination of stained cover-glass specimens prepared 
from finger blood gives either negative or erroneous results in the 
great majority of instances. In certain overwhelming infections, 
notably in some of the severer forms of bubonic plague, it may 
often be possible to detect the specific micro-organism in the 
stained film, but the method must be regarded as too crude -and 
unreliable to furnish accurate findings in the average case. 

Blood Cultures. In order to secure the most reliable informa- 
tion from blood culturing, the systematic observance of three pre- 
cautions is essential. First, contamination by the skin bacteria 
above referred to must be carefully avoided, by the thorough ster- 
ilization of the patient's skin at and adjacent to the site from 
which the blood is aspirated. Second, not less than 0.5 c.c. 
of blood should be used for each culture, since only in rare in- 
stances are bacteria so numerous in the peripheral circulation as 
to be demonstrable in a single drop of blood. Third, -fluid, rather 
than solid, culture media should be used, in sufficiently large 
quantities to dilute the blood freely, about 100 parts of me- 
dium to each part of blood, the object of this precaution being 
to secure attenuation of the bactericidal properties of the blood, 
which otherwise might prove strong enough to prevent all bac- 
terial development. 

'Dennis' "System of Surgery," Philadelphia, 1895, vol. i, p. 251. 


For aspirating the blood the author prefers to use a tube of 
about 10 c.c. capacity, like that illustrated below (Fig. 38). 
One end of the tube is ground to fit a No. 42 hypodermic 
needle, while over the other end (plugged with a small bit of 
cotton) is slipped a piece of rubber tubing for aspirating. The 
apparatus, minus the rubber tubing, is inclosed in a larger glass 
tube, both open ends of which are also plugged with cotton, and 
sterilized by dry heat, the aspirating tube being adjusted at the 
time the blood is to be collected. This instrument is far superior 
to an antitoxin or a hypodermic syringe for the purpose intended, 
being simple, inexpensive, easily sterilized, and readily cleaned 
after use. It is especially well adapted for making cultures at a 
distance from a laboratory, where the sterilization of an ordinary 
piston-syringe is difficult, if not impossible. 

At least one hour before the aspiration of the blood the skin 
of the patient's arm at and for some distance on all sides of the 
bend of the elbow should be scrubbed thoroughly for several 


minutes either with a strong ethereal soap or with tincture of 
green soap, after which the part is well rinsed with hot sterile 
water, and finally washed with alcohol and ether. A moist hot 
i : 500 bichlorid compress is then applied over the site thus cleaned, 
being left in place until the time of the withdrawal of the blood. 
As a preliminary to this operation the dressing is removed, and 
the part freely douched and scrubbed with hot sterile water, in 
order to remove every trace of the bichlorid. A rubber drain- 
age tube, previously sterilized, is twisted tightly around the pa- 
tient's arm above the bend of the elbow, so as to cause disten- 
tion of the superficial veins in this situation, and the point of the 
needle is then thrust obliquely into the most prominent of these 
vessels, with the result that the blood immediately begins to flow 
into the bore of the instrument. If, for any reason, the force of 
t^e blood flow should fail to fill the caliber of the tube, sufficient 
blood may easily be obtained by making gentle suction through 
the rubber tubing. While introducing the needle it should be held 
almost parallel to the long axis of the vein, for should it be simply 


plunged into the vessel at right angles, there is danger that the 
point will pass completely through the vessel from wall to wall 
and penetrate the surrounding tissues an accident which may 
explain the cause of many a "dry tap." The site of the aspira- 
tion may be made anesthetic by preliminary freezing with a spray 
of ethyl chlorid, but to most patients the operation is not painful 
enough to necessitate this. 

Having thus collected, say, 10 c.c. of blood, the con- 
tents of the tube are divided equally among five Pasteur flasks, 
each containing at least 200 c.c. of broth or other suitable 
fluid culture medium. The flasks are . then shaken for a few 
moments, in order to mix the blood and medium and to dilute 
thoroughly the former, after which they are placed in an in- 
cubator. The identity of the growths, should any occur, remains 
to be determined by secondary culturing and microscopical ex- 
amination, for descriptions of which the student should consult 
text-books on bacteriology. Cultures made by this technic, 
suggested by Adami, 1 are much more favorable to the growth 
of any bacteria which may be in the blood stream than the older 
methods of using solid media, except under special circumstances, 
such as the cultivation of the gonococcus, the influenza bacillus, and 
other germs which grow most luxuriantly on special forms of 

Staining Methods. In the limited number of instances to which 
such methods are applicable the technic described below will be 
found useful. 

An attempt should be made to sterilize the skin of the finger 
from which the blood is obtained by thoroughly scrubbing the 
part first with ethereal or green soap, and then with a i : 500 
bichlorid solution, alcohol, and ether, in the order named, this 
being followed by sponging with sterile water. A deep punc- 
ture having been made with a needle which has been sterilized 
by the naked flame, and the first few drops of blood escaping 
from the wound allowed to drip away, one of the succeeding 
drops is transferred by means of a sterile platinum needle to the 
surface of a cover-glass upon which a second cover-glass is at 
once laid, the two being drawn apart, in order to secure a pair 
of spreads. The latter are immediately dried by gentle heat 
and then passed several times through a Bunsen flame. It is 
needless to add that the cover-glasses used for making the films 
must be sterilized by heat, and handled by means of a pair of 
sterile forceps. Films thus prepared may be stained with any of 
the basic anilin dyes (thionin, methylene-blue, and methyl- or 

1 Jour. Amer. Med. Assoc., 1899, vol. xxxiii, p. 1514. 


gentian- violet being most useful for this purpose), after which 
they are washed in water, dried, and mounted in Canada balsam 
or in cedar oil. Should a double-stained specimen be desired, 
one of the eosin and methylene-blue solutions referred to previously 
may be depended upon to give satisfactory results. 

Giinther's method l will be found useful if the object is to de- 
stroy the color of the erythrocytes, so as to leave a freer field of 
vision for any bacteria which may be present in the film. Ac- 
cording to this method, the specimen is first immersed for ten 
seconds in a 5 per cent, aqueous solution of acetic acid, until 
the tint of the hemoglobin has entirely faded away, after which 
the reagent is removed by briskly blowing upon the surface of 
the cover-glass; the latter is then held, face downward, over the 
open mouth of a bottle containing strong ammonia water, so as 
to neutralize all remaining traces of the acid. The film is now 
stained for twenty-four hours with the Ehrlich-Weigert fluid 
(contained in a covered staining dish), at the end of which time 
it will be found to be colored a deep blue. It is then decolorized 
by a few seconds' immersion in a i : 14 aqueous solution of nitric 
acid, until the color fades to a light green; rinsed in alcohol; dried 
in air; and mounted in balsam. 

The Ehrlich-Weigert fluid is prepared by adding from 10 to 
15 drops of anilin oil to 6 c.c. of distilled water, held in a test- 
tube. The fluid is thoroughly mixed by shaking, and then filtered. 
To the filtrate a few drops of a concentrated alcoholic solution 
of methyl- or gentian-violet are added just sufficient of the dye 
to produce a slight turbidity of the liquid, which clears up in a 
few minutes. The mixture prepared in this manner is employed 
as the staining agent. 


In 1894 Pfeiffer 2 noticed that the vibrios of 
WIDAL'S Asiatic cholera, if injected into the peritoneal 
TEST. cavity of a guinea-pig immunized against this 
disease, rapidly lost their characteristic motility, 
and tended to become granular, broken up, and dissolved, while 
in the healthy, non-immune animal they developed normally and 
abundantly, and failed to show any such changes in their mor- 
phology. He claimed that this reaction, known as "Pfeiffer's 

1 Fortschr. d. Med., 1885, vol. iii, p. 775. 

1 Zeitschr. f. Hyg., 1894, vol. xviii, p. i; ibid., 1895, vol. xix, p. 75; also 
Centralbl. f. Bakt. u. Parasitenk., 1896, vol. xix, p. 191; also Deutsch. med. 
Wochenschr., 1896, vol. xxii, p. 97. 


phenomenon," was specific, and emphasized its value as a means 
of laboratory differentiation. Two years later Pfeiffer and Kolle 1 
found that the same changes occurred in experiments with the 
bacillus of Eberth and animals rendered immune to enteric fever, 
and, furthermore, discovered that the test could be conducted in 
vitro, by mixing in a test-tube typhoid cultures and immune serum. 
It is of interest to note that results somewhat analogous to those of 
Pfeiffer had been observed in 1891 by Metschnikoff, 2 and in 1889 
by Bordet, 3 and by Charrin and Roger, 4 although none of these 
workers appeared to recognize the significance of their observa- 

In 1896 Griiber and Durham 5 applied the principles of Pfeif- 
fer's phenomenon to many other motile as well as non-motile 
bacteria, deduced new facts regarding its utility as a means of 
differentiating various species of germs, improved the technic 
of the test, and made the important announcement that aggluti- 
nation and immobility of typhoid bacillus cultures were pro- 
duced by the action of blood serum from a patient having re- 
cently recovered from an attack of enteric fever. It remained, 
however, for Widal, 8 in 1896, first to apply the reaction clinically, 
and to announce that enteric fever could be diagnosed by noting 
the clumping and immobilization of the typhoid bacillus when 
mixed in definite proportions with blood serum from a patient 
suffering from typhoid. This reaction, Widal insisted, was one of 
infection, and was demonstrable not only during convalescence, 
but during the incipiency and the height of the disease. 

The serum reaction is to-day recognized as an important sign 
in the diagnosis not only of enteric fever, but also of Asiatic 
cholera, Malta fever, relapsing fever, paracolon infections, and 
bacillary dysentery, while its value still remains less certainly 
established in many other conditions, such as, for example, lep- 
rosy, tuberculosis, bubonic plague, sepsis, and pneumococcus in- 
fections. The technic of the test and its diagnostic significance 
under various circumstances will be described under the headings" 
of the diseases in which it occurs. (See Section VII, " General 

1 Zeitschr. f. Hyg., 1896, vol. xxi, p. 203; also Deutsch. med. Wochenschr., 1896, 
vol.xxii, p. 735. 

J Annal. de 1'Institut Pasteur, 1891, vol. v, p. 473; ibid., 1894, vol. viii, p. 714; 
ibid., 1895, vol. ix, p. 433. 

3 Ibid., 1895, vol. ix, p. 462; ibid., 1896, vol. x, p. 191. 

4 Compt. rend. Soc. biol., Paris, 1889, vol. i, p. 667. 
6 Munch, med. Wochenschr., 1896, vol. xliii, p. 285. 

' Bull, meld., 1896, vol. x, pp. 618 and 766; Sem. m<?d., 1896, vol. xvi, 
p. 259; ibid., 1897, vol. xvii, p. 69; Lancet, 1896, vol. ii, p. 1371; Munch, 
med. Wochenschr., 1897, vol. xliv, p. 202. 



The examination of suspected blood stains for forensic pur- 
poses includes microscopical search for blood corpuscles, spec- 
troscopy, the hemin and guaiacum tests, and the biological serum 
test of Bordet. The first three of these procedures are used 
merely to determine whether a given stain is or is not composed 
of blood, but they failed conclusively to prove the source of the 
latter. By the biological reaction it is possible to supplement 
these tests by proving the precise origin of the blood, whether 
human or derived from one of the lower animals. It need scarcely 
be added that in medico-legal work it is essential to use all four 
tests in the investigation of the material submitted for study. 

In undertaking the demonstration of eryth- 
MICROSCOPICAL rocytes in a dried clot, the latter must be 
EXAMINATION, treated with an agent which will macerate and 
dissolve the cells without laking them and 
thus removing their hemoglobin and altering their shape. Vir- 
chow's fluid a 30 per cent, aqueous solution of potassium hydrate 
is suitable for this purpose ; or Ranvier's solution a saturated 
aqueous solution of iodin containing 2 per cent, of potassium iodid 
may be used ; this stains various starchy cells which might other- 
wise counterfeit blood cells. The solution having been effected, 
a minute portion of the dissolved stain is either mounted as a wet 
specimen or spread as a dry film between cover-glasses, subse- 
quently stained with appropriate dyes, and mounted in balsam 
as a permanent specimen. For the microscopical examination 
a y^-inch oil-immersion objective is essential, and, if measurements 
are to be attempted, an ocular micrometer. 

If corpuscles resembling those of blood are detected in the 
specimen, especial attention should be directed to their average 
size, their shape, and to the presence or absence of nuclei. The 
.diameter of the normal human erythrocyte (agoff i n O i almost 
equaled by that of the dog's corpuscles, which averages approxi- 
mately -g-jVir in. If the cells are disc-shaped and measure ^in^r 
in. or less in diameter, it is safe to consider them non-human, 
and belonging to some one of the common domestic animals, 
such as the cat, horse, cow, ox, pig, sheep, or goat. If the cor- 
puscles are of oval shape and nucleated, they are certainly not 
mammalian, but are derived from a fowl, a fish, or a reptile. A 
possible, though highly improbable, contradiction to these general 
premises must be recalled, namely, the structural alterations of 
human erythrocytes occurring in anemic bloods particularly 
the tendency toward microcytosis in chlorosis, and toward megalo- 


cytosis, poikilocytosis (often of oval character), and nucleation in 
intense anemia, such as that of the primary pernicious type. 
Leukemic blood, however, is characteristically hall-marked. Here 
may be noted Dresbach's case, 1 unique of its kind, of a healthy 
young mulatto 90 per cent, of whose erythrocytes were of oval 
or elliptical shape. 

In a considerable proportion of cases microscopical examina- 
tion of a dried blood clot avails nothing, owing to the destructive 
changes which have taken place in the cells. Even under the 
most favorable conditions all one can usually accomplish with the 
aid of microscopy is to hazard an opinion that a given specimen 
of blood is either mammalian or derived from a fowl, a fish, or a 
reptile. For the medico-legal aspects of such examinations, 
together with the rigid technical precautions demanded, the 
reader is referred to the standard works on forensic medicine. 2 

This method of examination has already been 
SPECTROS- described. (See p. 107.) It is to be employed 
COPY. whenever a sufficient quantity of a blood solu- 

tion is available, the purpose being to deter- 
mine the presence or absence of blood pigment, and further to 
identify its particular variety. If the blood clot is tolerably fresh, 
it may yield, in aqueous solution, the absorption bands of oxy- 
hemoglobin, which, on the addition of ammonium sulphid, charac- 
teristically change to the spectrum of reduced hemoglobin. Old 
stains, especially those whose hemoglobin has been altered by 
exposure to the air and sunlight, show a spectrum of methemo- 
globin, while blood which has undergone putrefaction, in addition 
to yielding this spectrum, shows that of hematin. Recent stains 
can usually be dissolved in distilled water, but old clots require 
a more active solvent, such as acetic acid or sodium hydroxid. 
When such reagents are used, the spectra either of acid hematin 
or of alkaline hematin result,* according to the reaction of the 
solvent employed. If the stain has been subjected to a high de- 
gree of heat, with the consequent formation of hematoporphyrin, 
it should be dissolved with concentrated sulphuric acid, the re- 
sulting solution producing the spectrum of hematoporphyrin in 
acid solution. 

If positive, Teichmann's hemin test is certain 
TEICHMANN'S proof of blood, although it does not, of course, 
HEMIN TEST, indicate its source. The test is of extreme deli- 
cacy, and may be relied upon to show the 
presence of the slightest trace of blood in the material examined, 

1 Jour. Amer. Med. Assoc., 1904, vol. xlii, p. 837. 

2 Peterson and Haines, "Text-book of Legal Medicine and Toxicology," 
Philadelphia, 1904; Tidy, "Legal Medicine," London, 1882. 


provided that the composition of the blood is not too materially 
altered. Negative results are therefore not conclusive, since the 
reaction may fail in stains exposed to a high temperature or to the 
prolonged action of the* sun's rays, and in those contaminated by 
certain substances, such as naphtha, iron rust, lime, lead, and 
animal charcoal. 

The hemin test is carried out as follows : A small particle 
of the suspected material, reduced to a fine powder, is mixed with a 
drop of normal salt solution upon the surface of a glass slide, the 
mixture then being evaporated slowly by moderate heat until a 
dry film forms. Care should be observed not to use too great heat 
in the evaporation, for fear of spoiling the reaction by causing 
decomposition of the hematin. The dry film is now covered with 
a cover-glass, under which a drop of glacial acetic acid is allowed 
to flow, after which the slide is again heated until minute bubbles 
begin to form. At this instant the heating should cease, and the 
preparation be allowed to cool. Active boiling at this stage of 
the test may drive off all the free hydrochloric acid evolved by the 
addition of acetic acid, and thus prevent the formation of the 
sought-for crystals. When cool, the specimen is examined micro- 
scopically with a low power (^ inch) dry objective, which shows, 
if the material contained blood, distinctive crystals of hemin or 
hematin hydrochlorid, consisting of yellow or brown rtiombo- 
hedral plates, lying singly and arranged as crosses or as stellate 

Van Deen's guaiacum test is sufficiently deli- 
THE GUAIACUM cate to show the presence of blood pigment in 
TEST. a solution as dilute as i : 5000, but unfortunately 

it responds to so many other substances that its 
only value is as a negative sign. A bit of the suspected clot or a 
shred of the stained fabric is moistened with distilled water, in 
order to dissolve the blood pigment, and to this solution are added 
a few drops of freshly prepared tincture of guaiacum. To this a 
drop or two of hydrogen peroxid is added, with the result that a 
blue color immediately develops in the presence of even minute 
traces of blood. Or the test may be carried out just as satis- 
factorily simply by pressing against the dry stain a piece of wet 
filter-paper, and then adding to the moist daub thus made the 
guaiacum and peroxid. According to Peterson and Haines, 1 the 
following substances produce the same reaction with this test as 
given by blood pigment : potato skin, casein, glue, iron and copper 
compounds, the double chlorid of gold and sodium, manganese 
dioxid, potassium permanganate, and indigo; all of which means 

1 Loc. cit. 


that a positive reaction indicates absolutely nothing definite, 
although, on the other hand, a negative result proves with great 
certainty the absence of blood pigment in the material examined. 
Originally Bordet, 1 later Uhlenhuth and 
THE BIOLOG- Tchistovitch, 2 Wassermann and Schutze, 3 and 
ICAL TEST others demonstrated the important fact that the 
FOR BLOOD, blood serum of an animal into which has been 
injected the blood of another animal of different 
species develops the property of agglutinating and dissolving 
erythrocytes similar to those injected, but has no such effect upon 
blood derived from another source. The principle of this biologi- 
cal reaction is well expressed by Valee's law: If an animal, A, be 
inoculated repeatedly with an albuminoid material from an animal 
of a different species, B, the blood serum of A acquires the specific 
property of precipitating in vitro albuminoid fluids derived from 
animals belonging to the species B. Thus, in the blood of the 
inoculated animal are developed antibodies selectively hostile to 
the toxic principles of substances identical with those injected, 
and the serum containing such antibodies is known as an anti- 
serum. Lysins, which dissolve ; precipitins, which precipitate ; and 
agglutinins, which clump, the poisonous substances and antidote 
their toxicity, are examples of the antibodies evolved in this 
aanner. On this principle it is possible to produce antisera not 
only for homologous bloods, but also for different animal albumin- 
ous fluids and cells and for vegetable albumins. For instance, 
such sera have been developed which react specifically with cow's 
milk, with horse-meat, with semen, with various epithelial cells, 
and with a number of other substances homologous to those 

In the case of antisera for various bloods, while the test is 
specific with homologous blood, it is also true that feeble reactions 
may occur with the blood of closely related species. Thus, 
rabbits treated with human blood yield a serum reacting not 
only with the blood of man, but also with that of certain species 
of monkeys. Nuttall and Dinkelspiel 4 and Grimbaum 5 showed 
that it is the blood of the anthropoid apes (Simiada), especially 
the gorilla, the chimpanzee, and the . orang, which reacts most 

1 Annal. de PInstitut Pasteur, 1898, vol. xii, p. 688; ibid, 1899, vol. xiii, 
p. 273. 

* Deutsch. med. Wochenschr., 1901, vol. xxvii, p. 82; ibid., 1902, vol. xxviii, 
pp. 659 and 679. 

s Berlin, klin. Wochenschr., 1901, vol. xxxviii, p. 187; also Wassermann, 
"Immune Sera " (Eng. trans, by Chas. Bolduan), New York, 1904. 

4 Brit. Med. Jour., 1901, vol. i, p. 1141; also Nuttall, "Blood Immunity and 
Blood Relationship," Cambridge, 1904. 

4 Lancet, 1902, vol. i, p. 143. 


decidedly with human antiserum; while the lower orders of 
monkeys, represented by the Hapalida, the Cercopithecida, and 
the CebidcB, give less positive results. Practically, this startling 
biological confirmation of Darwin's views is scarcely a source of 
error, since in high dilutions (500 or more) human antiserum does 
not show a precipitate even with the highest order of primates' 
blood (Layton 1 ). 

Antisera for horses, cattle, sheep, pigs, dogs, birds, and other 
animals and fowls all react typically with homologous bloods, 
but also sometimes, although always atypically, with alien bloods. 
Pig antiserum, for example, reacts faintly with the blood of the 
wild boar; horse antiserum with donkeys' blood; fox antiserum 
with wolves' and dogs' blood; and sheep antiserum with goats' 
blood. It is a notable fact that these are reactions between 
biologically related species of animals, and that they are feeble 
and atypical, in comparison with the reactions occurring between 
homologous antisera and bloods. It is also true that the more 
remote the biological relation of the animal from the one whose 
blood activates the antiserum, the feebler the reaction becomes. 
As explained subsequently, these pseudo-reactions do not occur 
if the blood to be tested is adequately diluted before examination. 

TECHNIC. Preparation 0} the Antiserum. The antiserum is 
prepared by injecting healthy rabbits with from 5 to 10 c.c. of 
human defibrinated blood, at intervals of about four days, until a 
total of between 50 and 80 c.c. has been administered. One or 
two weeks after the last injection the animal is bled, and the serum 
obtained is collected in sterile test-tubes, which are sealed and 
stored for future use. Belgian hares are excellent antiserum 
producers, and are preferable to ordinary rabbits, being more 
resistant to the toxic effects of the injections, and yielding highly 
potent antiserum. Goats, sheep, and dogs have also been used, 
but not with wholly satisfactory results. Ewing 2 obtained ex- 
cellent antiserum from a hen, which, being biologically far re- 
moved from man, should theoretically furnish a highly selective 
human antiserum. 

The blood used for the immunization may be conveniently 
secured from the placenta and umbilical cord. It is collected in a 
sterile flask containing several small glass beads, which defibrinate 
the blood when the flask is agitated for a few minutes. The blood 
thus defibrinated may be injected immediately into the animal, 
or placed in a refrigerator for subsequent use. The fresher the 
blood injected, however, the more powerful the antiserum which it 
produces. The injections may be subcutaneous, intravenous, or 

1 Amer. Med., 1903, vol. v, p. 913. 2 Med. News, 1903, vol. Ixxxiii, p. 925. 


intraperitoneal, the last being the simplest and least dangerous 
to the animal if properly carried out. One-half of the rabbit's 
lower abdomen having been shaved and disinfected with a sub- 
limate solution, the needle of the syringe is firmly thrust through 
the tissues at a point within the prepared area, and .the injection 
made when the abdominal cavity is entered. Experience will 
determine the amount of pressure and manipulation necessary to 
penetrate the abdomen to a sufficient depth and to avoid wounding 
the gut. While making the injection an assistant should hold the 
animal in such a position that its abdominal wall is kept tense. 
Since the injection of human blood not infrequently causes toxic 
symptoms in the treated animals, they should be kept in the best 
possible hygienic surroundings, with ample runways, an abundance 
of air and light, and plenty of food and water. It is a good plan 
temporarily to discontinue the injections in animals which show 
marked loss of weight and other evidences of severe reaction. 

The blood of the animal under treatment should be tested 
from time to time, and when found to be sufficiently potent with 
human blood, the injections are stopped. A week or two later the 
antiserum will be sufficiently powerful, and it is then collected by 
bleeding the rabbit either from an ear vein or from the carotid 
artery. In neither instance is it necessary to exsanguinate the 
animal. The blood is collected, under aseptic precautions, in 
a dish, from which the serum, after coagulation has occurred, 
is pipetted into small sterile test-tubes measuring 10 cm. in 
height by 0.5 cm. in diameter. The filled tubes are then 
plugged with cotton and set upright in a refrigerator until re- 
quired for the test. As a preservative a few drops of chloroform 
may be added to each tubeful of antiserum. If this is done, 
however, the antiserum must be incubated for half an hour before 
being used, in order to remove, by volatilization, all traces of the 
chloroform, which otherwise might cause a pseudo-reaction with 
alien blood. Other preservatives, such as lysol, lysoform, carbolic 
acid, and thymol, may also cause clouding of antiserum, and 
therefore should not be used. Solutions of mercuric chlorid 
strong enough to be antiseptic destroy the antiserum. 

The test antiserum may be preserved in dry form by pouring 
it into Petri dishes and drying in a cool place, the film thus ob- 
tained being powdered and kept in a tightly stoppered bottle until 
required for use. To prepare the antiserum, this powder is simply 
dissolved in normal salt solution in definite proportions. 

Testing the Suspected Stain. In all cases the suspected stain 
must be proved to be blood by the hemin crystal test, the spectro- 
scope, and the other older methods described above. This is 


highly essential in forensic work, because body albumins other than 
blood (sputum, saliva, pus, feces, exudates, and albuminous urine) 
are capable of giving positive reactions with human antiserum. 
Contamination by such materials is, however, readily detectable 
by other tests. 

The stain to be examined is dissolved with a few drops of a 0.6 
per cent, aqueous solution of chemically pure sodium chlorid, 1 and 
the resulting cloudy mixture cleared by nitration through either 
asbestos or Schleicher's "blue ribbon" filter-paper, or by centrif- 
ugalization. The solution of the stain thus rendered perfectly 
clear is then added to the antiserum in the proportion of at least 
100 to i, and the mixture incubated at 37 C. If the test be posi- 
tive, a distinctly flocculent precipitate will form in the test-tube 
within three hours, this reaction being preceded by the formation 
of a more or less marked turbidity immediately or very shortly 
after having mixed the antiserum and the blood solution. Slight 
turbidities do not constitute positive reactions they may occur 
with heterologous bloods in dilutions less than the above. Several 
controls are also to be prepared and incubated simultaneously: 
one of the salt solution used as a solvent; a second of the pure 
dissolved and cleared stain; a third of the pure test antiserum; 
a fourth of human blood and normal rabbit's serum; and a fifth 
of human blood and the test antiserum. All but the last of these 
controls should remain clear. 

Sources of Error. Aside from pseudo-reactions due to under- 
dilution of the antiserum, accidental clouding and precipitation 
may arise from the following group of factors: 

1. Bacterial Growths. Antiserum clouded by bacterial con- 
tamination may be rendered clear by filtration through a Pasteur- 
Chamberland filter or by centrifugalization. Clouding from this 
cause does not occur within the arbitrary three-hour time limit of 
the test. 

2. Preservatives and Solvents. As stated above, various chemi- 
cals used for preserving the antiserum and for dissolving the sus- 
pected stains may lead to wrong inferences. 

3. Hyperacidity and Hyper alkalinity. The solution of the 
stain should be neutral in reaction, feebly acid, or feebly alkaline. 

1 In the case of old, difficultly soluble stains Uhlenhuth uses a one per cent, 
aqueous solution of sodium hydrate, while Ziemke prefers a concentrated solution 
of potassium cyanid, with subsequent neutralization with tartaric acid. These 
chemicals, though active solvents for old blood stains, are not dependable, because 
of their likelihood to cause false reactions with various antisera. Graham-Smith 
and Sanger obtained precipitates with solutions of potassium cyanid (one per cent.), 
tartaric acid (o.i percent.), and sodium hydrate (o.i per cent.). In medico-legal 
work it is safe to be guided by Nuttall's advice, and refuse to submit to the test 
bloods which are insoluble in normal salt solution 


If too highly acid or alkaline, positive reactions may occur with 
unrelated blood. 

4. Contamination of the Specimen. Human albumins and 
various chemicals may be mixed with the suspected blood stain or 
may resemble it, and in forensic work their presence must be ab- 
solutely excluded by appropriate tests. Nuttall 1 found that tannin 
especially causes decided clouding even in i : 1000 dilutions, and 
that solutions contaminated with yellow polished leather acted 
similarly. He found that human blood mixed with shoe polish, 
and blood allowed to dry upon black leather, wall-paper, various 
dress fabrics, rubber, oil-cloth, silver and copper coins, coal, wood, 
and other substances, reacted typically with human antiserum. 
The clouding due to the admixture of earth (referable to the 
presence of lime salts) Nuttall obviates by saturating with carbon 
dioxid and subsequent nitration. 

Old blood stains, except that they develop the precipitate 
slowly, react like fresh blood. Ziemke 2 examined stains known to 
be twenty-five years old and obtained positive findings. Accord- 
ing to J. Meyer, 3 even solutions of the muscular tissues of five- 
thousand-year-old .mummies react positively! Uhlenhuth 4 found 
that specimens could be frozen for two weeks at a temperature of 
ro C. without in any manner affecting the sensitiveness of 
the reaction,, and that blood mixed with soapy water, menstrual 
urine, and other contaminating fluids responds typically and 
promptly. Nuttall and Dinkelspiel 5 demonstrated that human 
blood mixed with the blood of different animals (sheep, oxen, 
horses, and dogs) reacts characteristically with human antiserum. 
Dried blood crusts subjected to an hour's heating at 130 C. do 
not respond to the test. 

Value of the Test. The Bordet reaction has already figured in 
three murder trials in this country, 6 and in these cases it has been 
accepted by the Court as valid evidence. In order to insure in- 
fallibility, faultless technic, bred only of long and intelligent 
experience, is one of the first essentials. This acquired, the ex- 
aminer must exclude every source of error outlined above, and 
consider as positive only those reactions which, with proper dilution, 
afford a turbidity and distinct precipitate within the prescribed 
time limit, the behavior of the several controls being consistent 
with that of the main test. Under these conditions a solution 

1 Loc. cit. 

2 Deutsch. med. Wochenschr., 1901, vol. xxvii, pp. 424 and 731. 

3 Munch, med. Wochenschr., 1904, vol. li, p. 663. 

4 Loc. cit. 6 Brit. Med. Jour., 1901, vol. i, p. 1141- 
8 1901, State of Maine vs. Lambert (Whittier); 1903, State of Del. vs. Collins 

(Robin); 1903, Commonwealth of Penna. vs. Bechtel (Lear). 


proved to be blood which reacts typically with human antiserum 
may surely be pronounced of human origin. 

The agglutination of human erythrocytes after 

HANGING- their admixture with alien serum and the absence 

DROP TEST, of this change on the addition of homologous 

serum may be observed under the microscope. 

This fact was first noted by Marx and Ehrnrooth, 1 who use the 

principle as the basis for the following simple test for identifying 

blood stains. 

The suspected blood stain is dissolved with a small quantity 
of normal salt solution mixed with an equal amount of fresh 
human blood, and of this solution a hanging-drop preparation is 
made and examined under the microscope with a i-inch objective. 
If the stain tested is human blood, the erythrocytes of the added 
fresh blood remain unclumped at the end of fifteen minutes' time, 
while if it is alien blood, they become distinctly agglutinated within 
this period. Monkeys' blood, although it fails to clump the 
erythrocytes, causes them to shrink and to become polygonal in 
shape. The writer can confirm this test, so far as human blood is 
concerned, but its medico-legal value must remain undetermined 
until further studies have been made. The reaction should be 
controlled by Bordet's test, for which it is in no sense a substitute. 


Numerous methods of blood analysis other than those de- 
scribed in this section have also been devised, more especially 
for scientific investigation than for clinical use. Their detailed 
description not being germane to the plan of this book, the reader 
is referred to the original articles for exact data. Among these 
non-clinical methods of research are included the estimations of 
the total volume of blood (Haldane and Smith 2 ); of the amount of 
solids (Stintzing 3 ); of the percentage of blood iron (Jolles 4 ); of the 
quantity of fat and fatty acids in the blood (Engelhardt 5 ) ; of the 
blood viscosity (Hirsch and Beck 8 ) ; of the osmotic tension of the 
plasma (Hamburger 7 ; de Vries 8 ); and of the resistance of the 
erythrocytes to the action of electricity, heat, and mechanical 
injury (Laker 9 ; Maragliano 10 ). 

1 Munch, med. Wochenschr., 1904, vol. li, p. 293. 
1 Jour. Physiol., 1900, vol. xxv, p. 311. 

3 Verhandl. d. XII. Cong. f. inn. Med., 1893. 

4 Arch. f. med. Exp., vol. xiv, p. 73. 

8 Deutsch. Arch. f. klin. Med., 1901, vol. Ixx, p. 182. 

8 Ibid., 1901, vol. Ixix, p. 503. 7 Centralbl. f. Physiol., 1893, vol. vii, p. 656. 
8 Jahrb. f. w. Botanik, 1884, vol. xiv, p. 427. 
* Wien. med. Presse, 1890, vol. xxxi, p. 1375. 
10 Berlin, klin. Wochenschr., 1887, vol. xxiv, p. 797. 





Blood is a tissue consisting of fluid and cor- 
PLASMA, SERUM, puscular elements, the former constituting about 
AND CELLS, three-fifths, and the latter two-fifths, of its total 
volume. It has been approximated that the total 
quantity of blood in the normal individual is from one-twelfth to 
one-fourteenth of the body-weight, the proportion being somewhat 
less in the infant than in the adult. Much lower proportions 
than these were determined by Haldane and Smith, 1 who found 
that in health the bulk of blood ranged from one-thirtieth to one- 
sixteenth of the body- weight (3.34 to 6.27 per cent.), these figures 
having been calculated by a method based upon the capacity of 
the blood to absorb CO 2 The fluid element of the blood, known 
as the plasma or liquor sanguinis, is an alkaline, yellowish liquid, 
of a specific gravity ranging from about 1.026 to 1.030, and 
containing approximately 10 per cent, of solid matter, of 
which three-fourths are proteids, consisting of fibrinogen, serum 
albumin, and serum globulin. Coagulation of the blood results 
in its separation into a densely reticulated, somewhat granular 
substance, fibrin, and into a clear, straw-colored, alkaline fluid, 
serum. Fibrin is a sparingly soluble, highly elastic, proteid body, 
which incloses and imprisons within its multitude of delicate 
fibrils the corpuscular elements, the whole forming the blood clot 
or crassamentum. Serum is a clear, straw-colored, alkaline fluid, 
having a specific gravity of about 1.026 and containing practically 
the same amount of solids and relative proportion of proteids as are 
found in the plasma ; its proteid constituents are fibrin ferment, 
which replaces the fibrinogen of the plasma, serum albumin, 
and serum globulin. 

The corpuscular elements of the blood are free cellular bodies 
suspended in the plasma. They are of two varieties : the eryth- 
rocytes or red corpuscles, and the leucocytes or white corpuscles. 

1 Jour. Physiol., 1900, vol. xxv, p. 33. 


In addition to these cells, two other elements are also found, 
namely, the blood plaques and the hemokonia, although these 
bodies, while they may be conveniently grouped with the red and 
white cells, are not to be regarded as definite corpuscular entities. 

The salts of the blood include sodium chlorid, potassium 
chlorid, sodium carbonate, sodium phosphate, magnesium phos- 
phate, calcium phosphate, and sulphates ; of these salts, sodium 
chlorid is the most abundant, constituting from 60 to 90 per 
cent, of the total amount of mineral matter. 

Certain extractives are also found, among which are urea and 
uric acid, creatin, creatinin, xanthin, hypoxanthin, sugar, fats, 
soaps, and cholesterin. 

The gases of the blood consist of oxygen, nitrogen, and carbon 
dioxid, the oxygen existing chiefly in combination with hem- 
oglobin in the erythrocytes, and the carbon dioxid as carbonates ; 
the nitrogen is held in simple solution. About 60 volumes of gas 
are contained in each 100 volumes of blood. Arterial blood con- 
tains roughly 20 volumes of oxygen and 40 of carbon dioxid, 
while venous blood contains less than 10 volumes of oxygen and 
almost 50 of carbon dioxid ; the quantity of nitrogen in both 
arterial and venous blood is from i to 2 volumes. 


The distinctive color of the blood is due to 
NORMAL the presence of the hemoglobin contained in the 
VARIATIONS, erythrocytes, and alterations in the chemical com- 
position of this pigment produce corresponding 
changes in the color of these cells, and, consequently, in the 
naked-eye appearance of the whole blood. The color of arte- 
rial blood is bright scarlet, inasmuch as it contains a large amount 
of oxygen in chemical combination with the hemoglobin ; while 
venous blood, on the other hand, is of a dark, purplish-blue tint, 
owing to its deficiency in oxygen and to the presence of more or 
less uneliminated carbon dioxid. This difference in color is so 
obvious that a cursory glance suffices to distinguish arterial and 
venous bloods. 

The presence of immense numbers of hemo- 

DENSITY globin-containing elements accounts for the vary- 

AND ing degree of density and opacity which the 

OPACITY. blood possesses, distinguishing it from a mere 

transparent, colored fluid. If, for any reason, the 

hemoglobin escapes from the erythrocytes into the surrounding 

plasma, this characteristic opacity is quickly lost, and the blood 


becomes transparent and of a " laky " color. The cells are laked 
by the influence of heat, water, fat solvents, and hydrogen and 
hydroxyl ions, all of which act upon the outer semipermeable layer 
of the cells' stroma and thus favor hemoglobin dissolution. The 
density and the opacity, and, consequently, the color, of the blood 
increase and diminish according to the fluctuations which occur 
in the relative amounts of plasma and erythrocytes, and also 
according to the cells' richness in hemoglobin, irrespective of their 
numerical variation. 

In anemic conditions the blood is usually 
PATHOLOGICAL pale in color, somewhat transparent, and thin 
VARIATIONS, and watery-looking. This is the case particu- 
larly in primary pernicious anemia, in chlorosis, 
and in leukemia; in [the first- named disease it is sometimes difficult 
to believe that the watery, pale fluid which flows from the puncture 
is anything but pure serum ; in leukemia the blood drop may have 
a peculiar light, mottled, streaked appearance, or a uniform milky- 
white tint may predominate over the normal red hue. In cases 
of dyspnea, arterial blood, because of its inadequate oxygenation, 
may be dark blue, closely resembling blood from the veins. This 
similarity has also been noted in cases of poisoning by sulphur- 
etted hydrogen, in which condition the blood may even be changed 
to a dark greenish tint. In some cases of diabetes mellitus the 
presence of large quantities of free fat in the circulation seem- 
ingly divides the blood drop into two distinct layers an upper, 
light-colored portion, containing supernatant fat droplets, and a 
lower, darker layer of pure blood ; at first glance diabetic blood 
may be somewhat pinkish or salmon-colored. 

In poisoning by anilin, nitrobenzol, hydrocyanic acid, and potas- 
sium chlorate the blood is chocolate- or dun-colored ; and in 
poisoning by carbon monoxid, bright cherry-red. In severe 
icterus a yellowish- red tint of the blood has been observed. 


Owing to the presence of certain volatile fatty acids blood 
possesses a peculiar and characteristic odor or halitus, which may 
be intensified by the addition of concentrated sulphuric acid, and 
which rapidly disappears after the withdrawal of the blood from 
the body. The slippery, greasy feeling of freshly drawn blood is 
quickly lost after its exposure to the atmosphere, and is replaced by 
a viscosity, or stickiness, as coagulation progresses. The viscosity 
of the whole blood is apparently influenced to a large extent by 
the cellular elements, chiefly by the erythrocytes, although the 


viscosity of the serum must also be regarded as a determining 
factor of more or less importance. 

Hirsch and Beck 1 have determined that the "viscosity value," 
as they term it, of human blood is about five times that of distilled 
water i. e., the viscosity of blood having a specific gravity rang- 
ing between 1.045 an d 1-055 is expressed by the figure 5.1, in 
comparison with that of water, which equals i, the temperature 
of both fluids being the same, 38 C. Although no close relation- 
ship can be distinguished between the degree of viscosity and the 
specific gravity of the blood, these experimenters have apparently 
proved that the lower the density of the blood, the less marked 
its adhesiveness. This quality is exaggerated in individuals living 
upon a largely nitrogenous diet, and it is greatly modified by star- 
vation. S. Weir Mitchell 2 has observed that hyperviscosity 
develops when blood is subjected to the direct action of snake 
venom, while Stengel 3 has noted a similar condition resulting 
from contaminating fresh blood with the serum of patients suffer- 
ing from chlorosis, pernicious anemia, and leukemia. Any one 
who has done much blood work is familiar with the marked 
fluidity of the fresh specimen in the high-grade anemias, and 
with the diminished viscosity of the erythrocytes and their dis- 
inclination to form rouleaux under such circumstances. 


Under normal conditions the reaction of the 

REACTION blood is alkaline, owing chiefly to the presence 

IN HEALTH, of sodium carbonate and disodium phosphate. 

Clinically, the degree of alkalinity is determined 

by ascertaining the amount of sodium hydroxid which is exactly 

neutralized by 100 c.c. of blood, the result being usually expressed 

in milligrams of NaOH per 100 c.c. of blood. The figures given 

by different investigators as representing the normal alkalinity 

range within the widest limits, chiefly in consequence of the many 

different methods by which such data were obtained. In view 

of these marked discrepancies the alkalinity figures of different 

workers are in no sense comparable unless they are based upon 

precisely similar methods of investigation pursued with identical 

technic. The following table, compiled from reliable data, illus- 

1 Deutsch. Arch. f. klin. Med., 1901, vol. bdx, p. 503. 

2 Mitchell and Stewart, "A Contribution to the Study of the Effect of the 
Venom of Crotalus adamanteus upon the Blood," Washington, 1898. 

3 "Twentieth Century Practice of Medicine," New York, 1896. 



trates the range of the normal blood alkalinity as estimated by 
various observers: 


Kraus 162-232 mgm. NaOH per 100 c.c. of blood. 

Burmin 182218 

Rumpff 182-218 

Jeffries 200 

Freudberg 200-240 

Lepine 203 

Canard 203-276 

Drouin 206 

Von Limbeck 218 

Zuntz and Lehmann 240 

Orlowsky 240-267 

Von Jaksch 260300 

Schultz-Schultzenstein 260-300 

Dare 266 

Strauss 300-350 

Brandenburg 33~37 

Lowy 449 

Berend 450-500 

Engel 479-533 

Mya and Tassinari 616 

With the titration method, now generally admitted to furnish 
fairly accurate results, appreciably higher figures are obtained 
with laked whole blood than with serum alone, since by the 
former method the alkalinity of all the plasma and cellular ele- 
ments is estimated, while by the latter the influence of the cor- 
puscles is entirely eliminated. 

The alkalinity of the blood is slightly higher, 
PHYSIOLOGICAL as a general rule, in men than in women and 
VARIATIONS, children, and is somewhat influenced by the time 
of day, being at its minimum during the early 
morning hours, gradually rising during the afternoon, and falling 
again during the evening. Some observers maintain that it is 
increased during the period of digestion, but this fact is disputed 
by others. It is temporarily diminished by the effects of mus- 
cular exercise and by a diet deficient in nitrogenous substances; 
on the contrary, richly nitrogenous }ood eaten during the per- 
formance of muscular work overcomes the effect of such exertion 
in lowering the alkalinity. The effects of cold baths are said to 
increase the alkalinity of the blood. Orlowsky, 1 from a study 
of 63 cases of various maladies, concludes that the degree of 
blood alkalinity is proportional to the erythrocyte count, but that 
it bears no relation to the number of leucocytes. 

In health, by the perfect mechanism of the emunctory organs 
of the body, the normal balance of blood alkalinity is constantly 

1 Deutsch. med. Wochenschr., 1903, vol. xxix, p. 601. 


maintained, in spite of the entrance of acids into the blood, 
whether by the ingestion of acid substances or by their produc- 
tion within the system, for the hyperacidity from such causes 
is promptly removed from the blood by the action of the kidneys, 
the skin, and the lungs. Thus, the ingestion of acids is quickly 
followed by increased acidity of the urine and sweat, while at the 
same time an increased quantity of carbon dioxid is given off by 
the lungs. It is probable that the tendency to acidity is partly 
neutralized by the ammonium salts generated from proteid foods, 
and also by the action of the liver. The blood alkalinity may 
be transiently increased by administering an alkali internally or 
by enema, the latter method having the more pronounced effect, 
according to Orlowsky. 1 

Increased alkalinity goes hand in hand with increased antidotal 
action 0} the blood against bacterial infection, as experiments have 
shown that animals whose blood had been artificially rendered 
highly alkaline by the administration of sodium salts, showed 
much greater resistance to the effects of virulent micro-organisms 
than untreated animals. Therefore, it is believed that the power 
of immunity against infections may, to a certain degree, be meas- 
ured by the alkalinity of the blood, for, in animal experimenta- 
tion, the fact is evident that the greatest degree of blood alka- 
linity is found in animals whose immunity is absolute. The rat, 
which is naturally immune to anthrax, has excessively alkaline 
blood and other body fluids. This hyperalkalinity, however, does 
not protect this rodent against plague. An excess of alkali in the 
blood is probably antidotal to invading bacteria, not per se, but 
rather because of its power to dissolve and liberate cell nucleins, 
which, as alexins, are directly bactericidal. 

Unfortunately, the question of alteration in the 

PATHOLOGICAL alkalinity of the blood in various pathological 

VARIATIONS, conditions is at the present time one about which 

the opinions of different observers conflict, so 

that conclusions concerning this subject must be accepted with 

more or less reserve. 

It is of interest, however, to note that most observers agree 
that, as a rule, the alkalinity of the blood is perceptibly lowered 
in those diseases associated with a febrile movement, but no defi- 
nite relation between the intensity of the pyrexia and the degree 
of lessened alkalinity has been established. Auerbach, 2 having 
noted that a temperature of 108 F. renders alkaline culture media 
bactericidal, argues that in acute infections the pyrexia, although 
it diminishes the alkalinity of the blood, at the same time may be 

1 Loc. cit. 2 Abst. in Jour. Amer. Med. Assoc., 1903, vol. xli, p. 1122. 


beneficial, in that it also increases its bactericidal action. Sub- 
normal alkalinity figures have also been met with in the primary 
and secondary anemias, with the exception of chlorosis, in which 
condition the blood alkalinity usually is either normal or perhaps 
slightly increased. Desevres 1 has drawn attention to the fact 
that in the early stages of acute diseases the alkalinity is either 
normal or somewhat increased, and in the majority of instances 
it becomes perceptibly diminished during convalescence. In 
chronic diseases it is usually decreased if the duration of the dis- 
ease has been of long standing. 

Drouin 2 found a lessened alkalinity of the blood in enteric 
fever, pneumonia, malarial fever, diphtheria, rheumatic fever, ery- 
sipelas, appendicitis, and in many other acute infections. Can- 
tani 3 maintains that in the algid stage of Asiatic cholera the 
reaction of the blood during life in some cases may be even 
acid, and that the alkalinity is always markedly reduced. Von 
Jaksch, 4 Peiper, 5 Kraus, 6 and others state that the alkalinity 
is generally diminished in uremia, diabetes, osteomalacia, organic 
diseases of the liver, and in poisoning by carbon monoxid and 
by phosphorus, especially by the latter. Decreased alkalin- 
ity has also been noted in cholemia, Addison's disease, Hodg- 
kin's disease, poisoning by mineral acids, the late stages 
of malignant neoplasms, and in various long-standing cachectic 
conditions. Thomas 7 found the alkalinity reduced in acute 
alcoholism and as the result of chloroform narcosis. TchlenorfT 8 
found a diminished alkalinity in a wide variety of skin diseases, 
among which are named psoriasis, eczema, pemphigus, purpura 
hcsmorrhagica, erythema multiforme, lichen rubra, and elephantiasis. 
Since it was also found that the administration of arsenic failed 
to increase the blood alkalinity, the action of this drug upon 
dermatoses is evidently not due to its influence upon the blood. 

In organic diseases of the heart unassociated with pyrexia and 
in nervous diseases the alkalinity of the blood has been found to 
be increased. In chronic rheumatism and in renal lesions unac- 
companied by uremic symptoms the reaction of the blood is usu- 
ally found to be unaltered. 

1 These de Lyon, 1897-98. 

2 " Hemo-alcalimetrie et Hemo-acidimetrie," These de Paris, 1892, No. 83. 

3 Centralbl. f. d. med. Wissensch., 1884, vol. xxii, p. 785. 

4 Deutsch. med. Wochenschr., 1893, vol. xix, p. 10. 

5 Arch. f. pathol. Anat., 1889, vol. cxvi, p. 337. 
8 Zeitschr. f. Heilk., 1889, vol. x, p. 106. 

7 Arch. f. exper. Pathol. u. Pharm., 1898, vol. xli, p. i. 

8 Russkiy Vrach, 1898, vol. xix, p. 248; abst. in Jour. Cutan. and Genito- 
Urin. Dis., 1898, vol. xvi, p. 544. 



In the majority of healthy male adults the 
NORMAL specific gravity of the blood varies from 1-055 to 
RANGE. 1-065, the average being in the neighborhood of 
i .060. In women the average is somewhat less 
about 1.056; in children it is about 1.051; and in new-born infants 
i. 066 is considered normal. Diurnal variations in the specific 
gravity have been noted, but these fluctuations are slight and 
unimportant. The blood density of habitual dwellers in high alti- 
tudes is distinctly increased. Lloyd Jones 1 and Schmaltz 2 found 
that muscular exercise, if moderate, lowers the blood density, but 
if prolonged and attended by free sweating, distinctly increases it. 
Venous blood is said to be of slightly higher specific gravity than 
arterial. The average specific gravity of the blood of the two 
sexes, as determined by the principal observers, is as follows: 


Askanazy 1060. i 1056.4 

Schmidt 1060.0 1050.0 

Hammerschlag 1061.5 IO 57-5 

Lloyd Jones 1058.5 105 1 .5 

Landois IO 57-5 1056.0 

Becker 1057.0 1056.5 

Schmaltz 105 7.0 1056.0 

Peiper 1055.0 1053.0 

From a clinical standpoint the specific gravity 
PATHOLOGICAL of the blood may be regarded, within certain 
VARIATIONS, limits, as a tolerably accurate index to the 
corpuscular richness of this tissue and to the 
hemoglobin equivalent of the erythrocytes, since fluctuations in 
these constituents immediately give rise to corresponding altera- 
tions in the density of the blood mass. It follows, then, that in 
the various conditions of anemia, characterized by corpuscular 
and hemoglobin losses, low specific gravities are encountered; 
on the other hand, it is also obvious that in conditions of polycy- 
themia the cellular increase and the high hemoglobin equivalent 
are mirrored by the corresponding rise in the density of the 
blood. An increase promptly follows any sudden drain upon 
the fluids of the system sufficient to cause inspissation of the 
blood, such as may result from copious diarrhea, free sweating, 
or hyperemesis; while the density is at once lowered as the re- 
sult of sudden dilution of the blood, following, for example, the 
injection of a large quantity of saline solution or even the inges- 

1 Jour. Physiol., 1887, vol. viii, p. i. 

2 Deutsch. Arch. f. klin. Med., 1890, vol. xlvii, p. 145. 



tion of a large volume of liquid. Fluctuations in the specific 
gravity of the blood under such circumstances, which are purely 
physiological in character, are invariably of transient duration, 
for the normal relation between the relative volumes of corpuscles 
and plasma becomes quickly reestablished by means of the liquid 
interchange between the tissues and the blood vessels. 

Owing to the fact that in most instances a close relationship 
exists between the amount of hemoglobin and the specific gravity, 
some investigators are accustomed to take this parallelism as a 
basis for calculating the percentage of hemoglobin in the blood. 
Thus, by determining the specific gravity and by comparing the 
figure thus obtained with a table giving the hemoglobin equiva- 
lents corresponding to varying degrees of blood density, fairly 
accurate results have been obtained. The following hemoglobin 
equivalents of different specific gravities of the blood have been 
determined by Hammerschlag 1 and by Lichty. 2 


Specific gravity. Hemoglobin equivalent. 


25-30 pe 



I 043-1 045 

















: 95-100 


Specific gravity. Hemoglobin equivalent. 

10 33- I0 35 2 5-3 P er cent - 
1035-1038 3-35 
1038-1040 35-4 
1040-1045 40-45 
1045-1048 45-55 
1048-1050 55-65 
1050-1053 65-70 
io53- I0 55 70-75 
i55- I0 57 75-85 
1057-1060 85-95 

It will be noted that in both these tables the variations in 
density are somewhat greater in high than in low hemoglobin 
percentages. It has been stated by Diabella 3 that, on the aver- 
age, a difference of 10 per cent, in hemoglobin corresponds to 
4.46 parts per thousand in specific gravity, and that differences 
amounting to from 3 to 5 parts per thousand in the specific grav- 
ity may arise from the influence of the stroma of the erythrocytes, 
in blood characterized by a striking disturbance in the parallelism 
which normally exists between these cells and the hemoglobin. 

In the clinical application of this indirect method of computing 
hemoglobin percentages several conditions, in which factors other 
than the presence of hemoglobin in the erythrocytes influence the 
specific gravity, must be excluded. In leukemia, for example, it 
will be found that hemoglobin percentages based on the above 
tables are much higher than actually exist, the cause of this fal- 

1 Loc. cit. * Phila. Med. Jour., 1898, vol. ii, p. 242. 

8 Deutsch. Arch. f. klin. Med., 1896, vol. Ivii, p. 302. 


lacy being the presence of enormous numbers of leucocytes in the 
blood; in pernicious anemia the hemoglobin is frequently higher 
than the specific gravity indicates, for in this disease the individual 
corpuscles are much richer in hemoglobin than normally; and 
in conditions associated with extensive dropsy the hemoglobin 
percentage does not parallel the specific gravity, owing to the 
abnormally high proportion of fluids in the blood mass. 

These three sources of error, aside from the rather trying tech- 
nic by which one must first determine the specific gravity of the 
blood drop (see p. 94), are sufficient to make most workers 
reluctant to adopt this method as a substitute for the hemometer. 


The essential factor of coagulation of the blood is the forma- 
tion of fibrin, a proteid substance, produced in the plasma after 
the withdrawal of the blood from the body, by complex chemical 
changes occurring between the soluble calcium salts and the 
nucleoproteids of the blood, with the consequent production of a 
fibrin ferment. The theories regarding coagulation are numer- 
ous, conflicting, and unsatisfactory, and must necessarily remain 
disputed points until our present uncertain knowledge of the 
chemistry of the blood proteids becomes fuller and more definite. 1 
Coagulation is delayed and imperfect in hemoglobinemia, in as- 
phyxia, in jaundice, in conditions of general dropsy, and in in- 
dividuals who are prone to bleed freely from trivial wounds. A 
similar effect is produced by the administration of alcohol. Lamb 2 
has found that the venom of certain poisonous snakes conspicu- 
ously affects clotting. Cobra venom diminishes coagulation to a 
marked degree, and in some instances may even wholly prevent it. 
Intoxication with daboia poison (the venom of Russell's viper) acts 
variously, according to the quantity injected. In small doses it 
hinders coagulation and causes the formation of soft, loose clots; 
but in large doses this venom so increases the coagulability of 
the blood as to lead to fatal intravascular clotting. Small doses 
of the calcium salts, especially the chlorid, promote coagulation, 
but the opposite effect ensues if the dose is too large. The 
administration of gelatin usually acts in the same manner, be- 
cause, so Gley 3 and others of the French school believe, of the fact 

1 Schafer's "Text Book of Physiology," vol. i, Edinburgh and London, 1898, 
contains a complete exposition of the various theories of coagulation of the blood 
existing up to the present time. 

2 Glasgow Med. Jour., 1903, vol. lix, p. 80. 

3 Sem. m&L, 1903, vol. xxiii, p. 113. 



that it always contains from 2 to 5 per cent, of calcium chlorid. 
Brat 1 explains the therapeutic action of gelatin in promoting in- 
travascular clotting by assuming that it contains substances which 
favor the deposit of plastic material, presumably derived from the 
blood cells, at the site of the clot. The studies of Boggs 2 show that 
gelatin is a much less active clotting agent than calcium chlorid, 
and that it may entirely fail to act in some cases. Blood coagu- 
lability as a factor of intestinal hemorrhage and of venous throm- 
bosis in enteric fever is referred to elsewhere. (See Section VII.) 
Carstairs Douglas 3 found the following coagulation figures in 
healthy women and in normal and complicated pregnancies: 







Albuminurics (16 cases): 

(a) Pregnant 




(b) Puerperal 




Eclampsias (22 cases): 

(a) Pregnant 




(&) Puerperal 




Healthy pregnant women ( 7 cases) .... 




Healthy non-pregnant women (7 cases) 




From these findings Douglas concludes that the thrombi 
found in various organs in fatal cases of eclampsia are not due 
to increased intravascular clotting. 

In the microscopical examination of a slide of fresh blood 
fibrin appears as extremely delicate, straight, filamentous lines 
which cross and recross the field in every direction. It forms 
a network of fine, interlacing, fibrillary bands, in the clear areas 
of the serum intervening between the masses of corpuscles, some 
of the fibrin threads apparently radiating from centers consisting 
of small irregular masses of blood plaques. The relation of these 
islands of blood plaques to coagulation and fibrin formation, if, 
indeed, any exists, is undetermined. 

In normal blood the formation of the fibrin 

HYPERINOSIS network becomes apparent within two or three 

AND minutes after exposure of the blood to the air, 

HYPINOSIS. and the process is completed within about six 

minutes. * In certain pathological conditions, 

however, both the length of time required for its formation 

1 Berlin, klin. Wochenschr., 1902, vol. xxxix, pp. 1146 and 1170. 

1 Med.News, 1904, vol. Ixxxiv., p. 182. 3 Brit. Med. Jour., 1904, vol. i, p. 709. 



and the density of the network vary. An increase in the amount 
of the fibrin network is spoken of as hyperinosis, while a decrease 
in .fibrin is termed hypinosis. 

In general terms it may be stated that fibrin 
PATHOLOGICAL is increased in acute inflammatory and infectious 
VARIATIONS, diseases, especially in those attended by an ac- 
tive febrile movement and by exudative proc- 
esses, the amount of fibrin roughly corresponding to the intensity 
of the process. This statement is made with certain reserva- 
tions, for the rule does not hold true in all such instances, as is 

Showing rouleaux formation and fibrin network. 

noted below. All febrile states do not imply a fibrin increase, 
for none is found in the fevers associated with grave cases of 
chlorosis and of pernicious anemia. Hayem 1 suggests that the 
density of the fibrin network may be taken as an indication 
of the individual's resisting powers against disease, inasmuch 
as it appears to be more marked in the blood of the vigorous than 
of the feeble. In acute inflammations accompanied by serous 
and purulent exudates a dense fibrin reticulum is observed, the 
extent of the exudation being in a degree measured by the density 

1 "Du Sang," Paris, 1889. 



of the network. Fibrin is increased to a slighter extent in par- 
enchymatous inflammations, in inflammations of the mucous 
membranes and skin, and in the febrile stages of chronic sup- 
purations. Among the diseases which are associated with an 
increase in fibrin are the following: abscess, pneumonia, rheumatic 
fever, erysipelas, acute gout, severe angina, bronchitis, influenza, 
diphtheria, pleurisy, peritonitis, pericarditis, hepatitis, meningitis, 
acute gastritis, enteritis, cystitis, vaginitis, pustular stage of variola, 
and suppurating tuberculous cavities. Fibrin is not increased 
in malignant neoplasms, enteric fever, malarial lever, tubercu- 
losis, pernicious anemia, leukemia, chlorosis, and purpura. In 

Showing marked increase in the density of the fibrin network. 

parenchymatous nephritis it is but slightly, if at all, increased, 
while in interstitial nephritis the increase may be notable. 

Pfeiffer 1 declares, as the result of his investigations, that in 
all diseases in which an increase of fibrin exists inflammatory 
leucocytosis is also present, and that he has never been able to 
demonstrate hyperinosis without coexisting increase in the number 
of leucocytes. But leucocytosis does not invariably imply hyper- 
inosis, although the two conditions almost always go hand in 
hand. Leucocytosis may occur in purpura and in malignant 
disease unattended by fibrin increase; on the other hand, in in- 

1 Zeitschr. f. klin. Med., 1897, vol. xxxiii, p. 215. 


fluenza the fibrin network is denser than normal, while the number 
of leucocvtes is not increased. 


The term oligemia signifies a reduction in the total volume of 
the blood, involving a diminution of both the liquid and the cel- 
lular portions. It occurs most conspicuously after hemorrhage, 
and probably after this accident only. Sometimes, the hemorrhage 
having been profuse, the oligemia proves rapidly fatal; but in 
other instances, where the hemorrhage has been less extensive, 
the decreased volume of blood is slowly made up, first by a rapid 
osmosis of serum into the depleted capillaries from the neighbor- 
ing lymph spaces, and later by a slower numerical increase of 
the cellular elements, the products of an actual manufacture of 
erythrocytes by the blood-making organs. 

In some of the advanced cachectic states, in which profound 
adynamia and poor nourishment of the body are prominent clin- 
ical manifestations, there is seemingly good reason for believing 
in the existence of a true oligemia; but in the absence of con- 
firmatory evidence, reduction of the blood volume in this class of 
cases must remain rather a suspicion than an accepted fact. The 
term oligemia is, therefore, in the light of our present understand- 
ing, applicable only to blood losses resulting from hemorrhage. 


The term plethora is currently used to express a condition 
characterized by an actual excess in the total volume of the blood, 
affecting both the liquid and the cellular elements. According 
to the views of many of the older and a few of the modern path- 
ologists, a true polyemia, or an increase in the blood volume, un- 
accompanied by any qualitative changes, is thought to exist in 
certain individuals whose mode of life and luxurious habits are 
supposed to predispose and give rise to excessive blood-forma- 
tion. The compatibility between a real full-bloodedness and 
a "high-liver" was formerly much more generally credited than 
at the present time, and the association of such signs as a rich, 
ruddy complexion, enlargement of the superficial blood vessels, 
and a full, bounding pulse was depended upon for the recognition 
of this condition. Of late, however, the drift of opinion is against 
the probability of any such permanent increase in blood volume, 
but until an accurate method of estimating the total quantity of 


blood in the body has been devised, the presence or absence of a 
real plethora must obviously remain conjectural. 

True plethora may occur as a transitory condition, as the re- 
sult of the direct transfusion of blood, or the mechanical forcing 
back into the general circulation of a quantity of blood from a 
part to be removed from the body, as by the use of an Esmarch 
rubber bandage previous to the amputation of a limb ; in a similar 
manner a new-born infant may become temporarily plethoric by 
a complete emptying of the placenta before tying the umbilical 
cord. Plethora resulting from any of these influences is invari- 
ably of a transient character, for the physiological balance of the 
organism rapidly disposes of the surplus amount of blood by de- 
struction of the excess of cellular elements and by the elimination 
of the liquid portions. 

Serous plethora may be defined as an increase in the volume of 
blood due to excessive quantities of its liquid and saline con- 
stituents, without augmentation in the number of its cellular 
elements. A condition of this sort may be dependent upon the 
ingestion of large amounts of liquids, upon the transfusion of saline 
solutions, or upon vasomotor dilatation, whereby the transfer 
of an unduly large amount of liquids from the tissues to the blood 
vessels is promoted. In organic lesions of the kidneys and of the 
heart, with diminished elimination of water from the system, a 
serous plethora of more or less chronicity may develop. The 
condition, however, is usually of transient duration, as the surplus 
liquids in the circulatory system are quickly disposed of and the 
blood volume reduced to normal by intracapillary transudation. 
J. L. Smith 1 believes that in chlorosis a true excess in the volume 
of blood exists, though Lloyd Jones 2 maintains that the condition is 
one of hydremia (see below) rather than of actual serous plethora. 

Cellular plethora is a term which may appropriately be applied 
to the condition, also known as polycythemia, consisting in an in- 
crease in excess of the normal standard in the number of erythro- 
cytes. The circumstances under which this change occurs will 
be discussed later. (See p. 197.) 


A relative increase in the quantity of the liquid constituents of 
the blood is known as hydremia. This condition must not be 
confused with serous plethora, which is characterized by both a 
relative and an absolute increase in the liquids of the blood. The 

1 Jour. Physiol., 1900, vol. xxv, p. 6. 2 "Chlorosis," London, 1897. 


specific gravity of the blood is observed to fall in relation to the 
degree to which the change develops. 

Hydremia may be produced by any factors which disturb the 
normal relations between the cellular and the liquid elements of 
the blood, so that the latter are unduly increased. In other 
words, the blood is diluted, in consequence of which a given drop 
of such blood shows an apparent decrease in the number of cel- 
lular elements, although the latter are in reality unaffected by the 
change. Hydremia is observed after extensive hemorrhages, in 
which the primary effect of the oligemia is the taking up by the 
capillaries of an excess of tissue fluids to replace the blood loss; 
later, as blood formation gradually makes up for the cellular 
deficiency, the normal ratio between the corpuscles and the plasma 
is reestablished. In the acute febrile infections hydremia may 
develop in consequence of excessive destruction of the blood 
albumins by the pyrexia. Hydremia may also occur as the result 
of the ingestion of large amounts of liquids, after the injection of 
normal saline solution, and as a consequence of vasomotor dila- 
tation. The watery constituents of the blood are relatively in- 
creased in certain of the severe anemias, owing to the deficiency of 
corpuscular elements, which is compensated by fluids derived from 
the tissues. In some dropsical conditions, notably those associated 
with renal and cardiac lesions, hydremia may also be said to 
exist, either with or without anemia. Hydremia, while it does not 
necessarily imply the coexistence of anemia, is naturally often an 
accompaniment of the latter condition. 

Hydremia dependent upon such physiological factors as inges- 
tion of fluids and vasomotor dilatation is a transient condition, 
for the excess of fluid is promptly eliminated, and the normal 
relations restored. In other conditions the duration of the change 
obviously depends upon the nature and permanency of the etiolog- 
ical factor or factors. 


Anhydremia is a condition in which a relative diminution in 
the liquid constituents of the blood occurs, as the result of rapid 
osmosis from the capillaries into the surrounding tissues. Inas- 
much as the cellular elements do not share in this draining-away, 
their number is necessarily increased in a given drop of such con- 
centrated blood. The specific gravity of the blood increases in 
relation to the extent of the fluid drain. 

Conditions which cause the sudden dissipation of large quanti- 
ties of liquids from the body, in consequence of hyperactivity of 


the mucous and serous surfaces, are the most prominent factors 
in producing anhydremia. Thus, after profuse diarrheas, urinary 
crises, free sweating, excessive vomiting, and sudden and exten- 
sive pleural and peritoneal effusions the blood becomes concen- 
trated from a temporary loss of its fluid elements, which pass 
from the vessels into the tissues to replace the liquids lost in 
consequence of the drain. Ewing 1 illustrates this principle of 
anhydremia by his observations on patients in whom a pro- 
longed attack of malarial fever with severe anemia was followed 
by enteric fever or by acute dysentery. In such instances the 
thin, watery blood of the malarial infection promptly became 
thicker and darker in color as the inspissating effects of the com- 
plicating illness supervened. 

Oliver 2 has shown also that a moderate degree of anhydremia 
may arise as the result of various influences ' which cause an in- 
crease in arterial tension and a consequent acceleration in the 
transfer of water from the vessels into the tissues. For example, 
the change has been brought about by the influence of local and 
general exercise, faradism, massage, cold bathing, and the admin- 
istration of suprarenal extract. 

From the nature of the drain, which is rapidly compensated by 
the constant interchange which goes on between the vessel and 
the tissue fluids, anhydremia is a temporary condition. A per- 
fect physiological balance limits its duration to brief periods of 
time. (See " Polycythemia," p. 197.) 


Fat is present in normal blood in the form of an exceedingly 
fine emulsion, the amount varying in man from to 3.25 parts 
per thousand of blood, the mean amount being 1.6, according .to the 
analyses of Becquerel and Rodier. 3 

By the term lipemia is meant the presence of an excess of free 
fat in the circulating blood, a phenomenon which is observed in a 
number of conditions, both physiological and pathological. In 
addition to fat globules, the blood in lipemia may also contain 
fine granular particles failing to respond to the usual tests for 
fat. The nature of these granules is still in dispute, but it is 
generally believed that they are either a proteid substance pre- 
cipitated in the presence of free fat, or that they represent the 

1 "Clinical Pathology of the Blood," ad ed., Philadelphia and New York, 1903. 
2 Croonian Lectures, Lancet, 1896, vol. i, pp. 1541, 1621, 1699, and 1778. 
3 Cited by Futcher, Jour. Amer. Med. Assoc., 1899, vol. xxxiii, p. 1006. 


albuminous envelop surrounding certain of the fat globules. 1 
During the period of digestion, especially after a meal rich in 
fats, the blood may contain a sufficient amount of fat to give rise 
to temporary lipemia; the condition may also be met with in the 
breasl-jed infant, in the pregnant woman, and in the obese. 
Menstrual suppression is also capable of overloading the blood 
with fat. 

The existence of lipemia is of little clinical importance, for it 
has been observed in a number of diseases, so that it cannot be 
considered characteristic of any particular lesion. It has been 
noted in the following conditions: arteriosclerosis, chronic alcohol- 
ism, diabetes mellitus, gout, certain diseases of the liver, heart, and 
pancreas, chronic nephritis, tuberculosis, splenitis, malarial fei'er, 
typhus fever, Asiatic cholera, peritonitis, pneumonia, and poison- 
ing by phosphorus and by carbon monoxid. Lipemia commonly 
occurs as the result of lacerated "wounds of the blood vessels situated 
in fatty tissues, and after fractures of the long bones involving injury 
of the fatty marrow. 

The degree of lipemia may be so marked that the macroscop- 
ical appearance of the fresh blood is altered, the presence of large 
quantities of free fat rendering it salmon-colored, turbid, and 
milky. This is especially conspicuous in the specimen of blood 
serum obtained by centrifugalization, which has a distinct grayish, 
opaque appearance, not unlike that of chyle. 

Macroscopically, the presence of lipemia may be determined 
by mixing with ether in a test-tube a portion of the turbid blood 
serum, the excess of fat promptly dissolving, so that the serum 
becomes clear. Hey I, 2 W. Hale White, 3 and others have been 
able to distinguish lipemic blood in the retinal vessels by means 
of the ophthalmoscope. 

Microscopically, lipemia may be recognized by the presence 
of large numbers of glistening fat droplets, about 0.5 to 2 f* in 
diameter, which lie in the plasma between the groups of cor- 
puscles, often exhibiting very lively Brownian movements. These 
droplets respond to the usual tests for fat, dissolving in ether, 
and staining black with osmic acid and brick-red with Sudan III. 


The occurrence in the circulating blood of minute particles of 
melanin or pigment, derived usually from the hemoglobin of the 

1 See Futcher, loc. cit.; also Cole (cited by White), Lancet, 1903, vol. ii, p. 

2 Trans. Amer. Ophthal. Soc., 1880, p. 54. 3 Lancet, 1903, vol. ii, p. 1007. 


erythrocytes destroyed by blood parasites, is known as melan- 
emia. These melanin particles appear as fine bits of granu- 
lar matter, black or of a reddish-yellow color, either lying free in 
the blood plasma or embedded in the protoplasm of the leuco- 
cytes. In some instances the granules are extremely small-sized 
and few in number, and again the amount may be considerable, 
large numbers of pigment particles being apparently fused into 

Melanemia is frequently present in malarial jever, especially 
of the severer types, both in the form of free pigment and as 
pigmented leucocytes. Particles of pigment in the bodies of the 
leucocytes have also been seen in cases of insolation, in relapsing 
fever, in melanotic sarcoma, and in Addison's disease. 


Glycemia, or the presence of grape-sugar in the blood, occurs 
in perfectly normal blood to a very slight degree, the quantity of 
sugar found under physiological circumstances not exceeding 1.5 
parts per thousand. 1 The presence of sugar in excess of this fig- 
ure, which may be termed hyperglycemia, is met with in diabetes 
mellitus, in which disease as high as 9 parts per thousand have 
been detected. 2 The investigations of Freund 3 and of Trinkler 4 
apparently show that the blood in carcinoma, especially of vis- 
ceral involvement, contains an excess of some reducing agent, to 
all intents and purposes identical with sugar. Tire former author, 
in consequence of this fact, lays stress on the finding as a means 
of differentiating between carcinoma and sarcoma, since no such 
increase has been observed as an accompaniment of the latter 
type of neoplasm. Lepine 5 finds that the sugar content of the 
blood appreciably increases after extirpation of the pancreas, 
a hyperglycemia developing within twenty-four hours after the 
ablation of this organ. He also finds hyperglycemia after liga- 
tion of the duct of Wirsung* 

The most accurate method of detecting small quantities of 
sugar in the blood is by the phenylhydrazin hydrochlorid test, 
conducted by von Jaksch 7 as follows: A small amount of blood, 

1 The term "potential sugar" (Lepine) is applied to the sugar produced in 
normal blood after having been kept outside the body for half an hour, at a tem- 
perature of 58 C. This sugar is believed to be evolved from one or more carbo- 
hydrate molecules of the blood proteids. 

1 Hoppe-Seyler, Virchow's Arch., 1858, vol. xiii, p. 104. 

8 Wien. med. Blatter, 1885, vol. vii, pp. 268 and 873. 

4 Centralbl. f. d. med. Wissensch., 1890, vol. xxviii, p. 498. 

5 Cited by Flexner, Univ. of Penna. Med. Bull., 1902, vol. xiv, p. 391. 

* Sem. me*d., 1903, vol. xxiii, p. 385. 7 Zeitschr. f. klin. Med., 1886, vol. xi, p. 20. 


obtained by wet-cupping, is first freed from proteids by adding 
an equivalent weight of sodium sulphate and then boiling and 
filtering, the filtrate thus obtained being used for the test. A 
solution is now made in a test-tube, by mixing 2 parts of phenyl- 
hydrazin hydrochlorid and 4 parts of sodium acetate with about 
6 c.c. of water, and gently heating the fluid, if necessary, to 
effect solution. Five c.c. of the proteid-free filtrate, while still 
warm, are added to an equal volume of the test solution. This 
mixture is then placed in a test-tube half filled with water, heated 
for half an hour in a water-bath, and allowed to stand until cool. 
When cooling of the mixture has occurred, it shows under the 
microscope the presence of the characteristic yellowish crystals 
of phenyl-glucosazon, either detached or in clusters, together with 
colorless crystals of sodium sulphate. 


The presence in the blood of a demonstrable amount of uric 
acid has been designated as uricacidemia. The blood of the 
normal individual does not contain this substance in amounts 
sufficiently large to be detected by ordinary clinical tests, but it is 
found in appreciable quantities in a number of pathological condi- 
tions. Garrod, 1 many years ago, recognized that excessive accumu- 
lation of uric acid in the blood was associated with gout, and he at- 
tached to this sign much diagnostic significance. Later investiga- 
tions, however, have proved the utter unreliability of this finding 
as a pathognomonic sign of this disease, for in recent years a large 
number of other conditions has been found to be more or less 
constantly accompanied by relatively large amounts of uric acid 
in the circulating blood. Notable examples of such diseases are 
pneumonia, hepatic cirrhosis, acute and chronic nephritis, uremia, 
chronic gastritis, leukemia, severe anemia, and those conditions in 
which deficient blood aeration constitutes a prominent clinical 
symptom, such as organic cardiac disease, exudative pleurisy, and 
emphysema. Uric acid is not found in the blood hi enteric fever nor 
in rheumatic fever. Pyrexia, of itself, evidently has no influence 
in producing uricacidemia, nor is it at all probable that this con- 
dition goes hand in hand with an excessive elimination of uric acid 
in the urine. 

Garrod's test is well adapted clinically for detecting the pres- 
ence of appreciable quantities of uric acid in the blood. Slightly 
modified, it may be applied in the following manner: Two 

1 Med. and Chirurg. Trans., 1854, vol. xxxvii, p. 49; ibid., 1848, vol. xxxi, p. 183. 


and one-half c.c. of blood serum, obtained by blistering, are 
placed in a shallow watch-glass and acidulated by the addi- 
tion of about 4 drops of a 30 per cent, aqueous solution of acetic 
acid. A linen thread is then immersed in the acidulated blood, 
which is slowly evaporated at a temperature not exceeding 70 
F. At the expiration of from tw r enty-four to forty-eight hours, 
if the sample of blood contains uric acid, characteristic crystals 
of this substance are deposited upon the thread, their identity 
being readily detected by microscopical examination and by the 
murexid test. 


The presence in the blood of bile or bile-pigments has been 
termed cholemia, a condition which accompanies various forms 
of icterus. Bilious blood may have, as already stated, a yellow- 
ish-red color, and may yield, on agitation, an abundant foam, 
tinged with yellow. Hypertonicity of the serum and a tendency 
toward hemoglobin dissociation are characteristic of cholemic 
blood. (See " Icterus.") Bilirubin may be detected in the blood 
even when urine tests for this substance have proved negative, 
according to von Jaksch, 1 who employs this procedure to demon- 
strate its presence: About 10 c.c. of blood, obtained by wet 
cupping, are allowed to clot, after which the serum is pipetted 
off, filtered through asbestos, and coagulated at a temperature of 
80 C. Thus treated, the presence of bilirubin is betrayed by a 
greenish discoloration of the serum, which, if bile-free, remains 
a pale straw color. Should a brownish color develop by this 
test, the presence of hemoglobin in the serum is indicated. 


The occurrence in the blood of demonstrable amounts of ace- 
tone and of fatty acids is referred to as acetonemia and lipac- 
idemia, respectively. Acetonemia has been found in associa- 
tion with numerous pathological conditions, chiefly in those 
characterized by pyrexia, while fatty acids in the blood have been 
detected in diabetic coma, in malignant jaundice, in leukemia, 
and in various acute infections. 

For the recognition of acetone Simon 2 recommends Den- 
ige's test, to be applied as follows: About 3 c.c. of blood are 
treated with 30 c.c. of Denige's reagent (20 gin. of concentrated 
sulphuric acid mixed with 100 c.c. of distilled water, to which 5 

1 Loc. cit. 2 "Clinical Diagnosis," 5th ed., Philadelphia, 1904. 



gm. of yellow oxid of mercury are then added), and allowed to 
stand until a dark-brown precipitate has formed, after which the 
supernatant fluid is filtered off and treated with more of the 
reagent, so as to effect complete precipitation. It is then acidi- 
fied by the addition of about 3 c.c. of a 30 per cent, solution of 
sulphuric acid, and boiled for one or two minutes. The appear- 
ance of a white precipitate on boiling indicates the presence 
of acetone. This precipitate may be almost wholly dissolved 
by the addition of hydrochloric acid in excess. 

Fatty acids may be detected by boiling equal parts, by weight, 
of blood and sodium sulphate, filtering, evaporating the filtrate 
to dryness, and then extracting the residue with absolute alcohol. 
Microscopical examination of the residue will reveal crystals of 
fatty acids if lipacidemia exists. 


Bacteriemia, or the presence of bacteria in the 
OCCURRENCE, circulating blood, is a condition associated with 
a number of infectious diseases, in which instances 
it is frequently, but by no means constantly, possible to dis- 
cover the specific micro-organism of the disease in question by 
careful bacteriological examination of the blood. The demon- 
stration in the blood of such bacteria as pyogenic cocci in gen- 
eral septicemia, of the Streptococcus pyogenes and other pyogenic 
organisms in malignant endocarditis, of the bacillus of Eberth 
in enteric fever, of the gonococcus in gonorrheal arthritis, of 
the pneumococcus in pneumonia, and of the Bacillus tuberculosis 
in severe cases of acute miliary tuberculosis, is sufficient proof, 
without citing other instances, of the diagnostic value of bacterio- 
logical blood examinations. Such examinations are warranted 
in every case of severe infection the nature of which appears 
doubtful, since by their aid alone it is often possible to derive 
diagnostic clues of the greatest practical value. 

From the clinician's viewpoint, normal blood 
LATENT is absolutely sterile, since no cultural method has 
INFECTION, yet been devised by which it is possible to demon- 
strate the presence of bacteria in the circulation 
of the healthy individual. From the pathologist's standpoint, 
however, such a statement must be accepted guardedly, in the 
light of recent investigations. Adami, 1 in a comprehensive resume 
of the whole field of bacterial infection, cites a series of apparently 

1 Jour. Amer. Med. Assoc., 1899, vol. xxxiii, pp. 1509 and 1572; also Ford, 
Jour, of Hygiene, 1901, vol. i, p. 277. 


conclusive experiments by his assistants, Nicholls and Ford, who 
found that the kidneys and livers of healthy animals, removed 
aseptically immediately after death and placed in agar-agar kept 
at the temperature of the body, showed, after a few days, a rela- 
tively abundant growth of bacteria. This observer concludes 
that under normal conditions the leucocytes pass out through 
the mucosa on to the free surface of, more especially, the alimen- 
tary tract, some of these cells then undergoing destruction, while 
others, now laden with various foreign matters, including bacteria, 
pass back again into the submucosa and find their way either 
into the lymphatic channels or into the portal venules. In both 
of these sites there exists a decided tendency toward bacterial 
disintegration and destruction. Such isolated bacteria as may 
have escaped leucocytal destruction, or removal by the lymphatic 
glands or by the endothelium of the portal system, may pass 
either through the thoracic duct or through the liver, and enter 
the systemic circulation, from which they are eliminated chiefly 
by the kidneys. Such a condition as this, known as "latent in- 
fection" or "latent microbism," appears to be compatible with 
perfect health, for the number of bacteria which thus gain access 
to the blood stream and tissues is so small that unless their 
virulence is especially striking and the susceptibility of the indi- 
vidual peculiarly marked, the resisting powers of the tissues re- 
main sufficiently strong to prevent bacterial proliferation. It is 
also obvious that the presence in the blood of so limited a num- 
ber of bacteria cannot be demonstrated by culturing. 

If, on the other hand, the conditions are such 
BLOOD that bacteria multiply in the blood to any decided 
CULTURES, extent, then their development in artificial media 
outside the body may be successfully obtained in 
many instances, provided that proper technic is employed. That 
this has not been more successfully accomplished is no doubt 
due to the powerful bactericidal action of the shed blood, whereas 
this influence in the circulating blood is but trifling. As Adami 
remarks, "Because certain observers have failed to discover 
bacteria in the blood from cases of infectious diseases, it by no 
means follows that the blood when shed has been free from bac- 
teria." In modern methods of examination precautions are 
taken to attenuate the bactericidal properties of the shed blood 
by freely diluting it with a large quantity of fluid media, instead 
of using relatively small amounts of solid culture, as has been 
done largely in the past, and as the result of this improved technic 
blood culturing now yields a much higher percentage of positive 


findings, and gives more uniform results than were formerly 
obtained. (See " Bacteriological Examination," p. 109.) 

Among the various bacteria which different 

BACTERIA observers have succeeded in isolating from the 

FOUND IN circulating blood are included many micro-organ- 

THE BLOOD, isms, the identity of which, as etiological factors 

of disease, is generally recognized, and also a 

number to which pathogenicity cannot be convincingly attributed. 

The following list gives the most important examples of the 

former class : 

B. aerogenes capsulatus. B. tetani. 

B. anthracis. B. tuberculosis. 

B. coli communis. B. typhosus. 

B. influenza. Diplococcus intracellularis menin- 

B. leprcB. gitidis. 

B. mallei. Gonococcus. 

B. cedematis maligni. Micrococcus tetragenus. 

B. pestis bubonica. Pneumococcus. 

B. pneumonia. Pyo genie staphylococci. 

B. proteus vulgaris. Pyogenic streptococci. 

In addition to this list, a certain amount of interest attaches to 
the discovery in the blood of certain bacilli (Achalme), micrococci 
(Walker), and diplococci (Triboulet) in rheumatic fever; of 
peculiar bacilli (Afanasiew) in relapsing fever, in addition to the 
specific spirillum of this infection; of diplobacilli (Craig) in 
mumps; and of diplococci (Class) in scarlet fever and in typhus 
fever (Balfour and Potter). The presence of the Bacillus icte- 
roides (Sanarelli) in the blood of yellow fever patients is now 
known to mirror a secondary infection. 

The conditions in which the above-named bacteria occur in the 
blood will be discussed in a later section, under the diseases in 
question. (See " General Hematology," Section VII.) 


In a clinical sense the term anemia refers to 
DEFINITION, any deterioration in the quality of the blood, af- 
fecting the erythrocytes, the hemoglobin, or 
both of these elements. Thus, in pernicious anemia the most 
conspicuous deterioration in the quality of the blood is a diminu- 
tion in the number of erythrocytes, or an oligocythemia; in chlo- 
rosis the most marked change is usually a loss of hemoglobin, 

ANEMIA. 149 

or an oligochromemia; while in many other anemic conditions the 
erythrocytes and hemoglobin are decreased more or less propor- 
tionately. While it is true that, strictly speaking, the word anemia 
may also be used to designate a reduction in the blood volume, 
this condition is better denned by the use of the term oligemia. 
Ischemia is a form of local anemia resulting from some mechani- 
cal interference with the blood supply of the affected area. 

In certain individuals with such decided pallor 
PSEUDO- of the skin and mucous membranes that their 
ANEMIA. appearance at once leads one to infer that they 
are suffering from a well-defined anemia, no signs 
of this condition can be discovered, for even after the most careful 
examination of the blood the number of erythrocytes and the per- 
centage of hemoglobin may be. found to be normal. Such instances 
of apparent blood deterioration have been called pseudo-anemia; 
they are often explained by hereditary peculiarities, by vaso- 
motor disturbances affecting the superficial capillaries, and by 
deficiencies in the pigment and in the development of the capillary 
network of the skin. In pseudo- chlorosis the patient shows the 
typical objective signs of chlorosis, yet her hemoglobin and cor- 
puscular values are normal. (See "Chlorosis.") Vermehren 1 
has described, under the term angiospastic pseudo-anemia, the 
transient periods of almost cadaveric pallor which occur in some 
individuals with normal blood, as the result of vasomotor spasm 
provoked by cold, fatigue, emotion, and like influences. Dwellers 
in tropical countries are especially prone to a spurious form of 
anemia, to which the misnomer tropical anemia is occasionally 
applied. Every medical clinic can furnish patients suffering from 
neurasthenia, tuberculosis, and advanced Bright's disease, whose 
pallid countenances are a striking contrast to their normal blood 
counts. "Prison pallor" suggests lack of fresh air and sunshine 
rather than severe blood deterioration. It does not follow, therefore, 
that pallor of the skin and mucous membranes is invariably an indi- 
cation of anemia, although this sign is not misleading in the ma- 
jority of instances. On the other hand, it should not be forgotten 
that persons of good color and robust appearance sometimes suf- 
fer from decided anemias without the fact becoming evident at 
first glance. In chlorosis florida, for example, red cheeks are not 
incompatible with a low hemoglobin percentage. In view of these 
sources of error, in order to diagnose anemia with absolute accu- 
racy, an examination of the blood is essential, for no matter how 
valuable other clinical signs may appear, the changes in the blood 
are often the real key to the situation. 

1 Sem. med., 1903, vol. xxiii, p. 167. 


An entirely satisfactory classification of the 
CLASSIFI- various forms of anemia still remains to be de- 
CATION. vised, in spite of the numerous attempts which 
not a few eminent authorities have made to 
group these conditions according to sound pathological consid- 
erations. Therefore, largely for the sake of convenience, all 
anemias may be broadly grouped into two theoretical classes: 
primary and secondary. 

According to this tentative classification, primary anemias may 
be considered those in which a lesion of the hematopoietic organs 
is essentially accountable for the production of the disease. In 
anemias of this sort, the etiological factors are either entirely un- 
discoverable, or, if they are to be detected, too trivial to explain the 
intensity of the disease. Here the predominant clinical manifesta- 
tions are to be found in the changes occurring in the composition 
of the blood, the other symptoms being considered secondary to, 
and dependent upon, these alterations. 

Under the term secondary anemia are included those cases 
of anemia which are apparently secondary to, and symptomatic 
of, certain definite pathological lesions not primarily affecting 
the blood-making organs, such as, for example, enteric fever, 
syphilis, septicemia, malignant disease, malarial fever, and hemor- 
rhage. In such anemias the other clinical symptoms are, as a 
rule, much more conspicuous than the blood changes, which 
are thought to be secondary. An exception to this general rule 
must be taken, however, in regard to the anemia caused by the 
presence of the Bothriocephalus latus in the intestinal canal, for 
in this infection the blood picture is by all odds the most striking 
clinical manifestation. It is, furthermore, true that in some 
instances a secondary anemia may apparently merge into one of 
the primary type, should the protracted duration of the former 
in course of time cause such profound systemic effects that finally 
the blood-making organs become exhausted, and refuse ade- 
quately to supply the constant demand for corpuscles, with the 
result that the most prominent clinical signs are now found in 
the blood, and not in the original symptoms of the disease in 
question. The high grade anemia which sometimes follows 
enteric fever, becoming of such intensity that it counterfeits a 
primary anemia, may be cited as an example of this change. 

Until further progress has been made in the study of the physi- 
ology and pathology of the blood-making organs the following 
provisional classification of the anemias may be used for clinical 
purposes : 

I. PRIMARY ANEMIA. Chlorosis, pernicious anemia, splenic 


anemia, lymphatic leukemia, myelogenous leukemia, Hodgkin's 

II. SECONDARY ANEMIA. Dependent upon causes such as 
hemorrhage, intestinal parasites, prolonged lactation, unfavorable 
hygiene, metal poisoning, malignant disease, acute infections, and 
chronic diseases producing long-standing drains on the albumin 
of the blood. 

Excluding the effects of hemorrhage, deficient 
PATHOGENESIS. blood formation, excessive blood destruction, 
and a combination of these two processes are 
generally regarded as the three possible essential factors in the pro- 
duction of anemia. Deficient hemogenesis is to be attributed to a 
large number of different causes, among the most prominent of 
which may be mentioned the influence of unhygienic surroundings 
and insufficient nourishment from improper food and from in- 
adequate powers of assimilation. It is also probable that con- 
genital and acquired failure of the blood-making organs and the 
presence of growths which intercept the material for blood forma- 
tion are to be considered as the origin of defective hemogenesis in 
some instances. 1 Excessive blood destruction may be due to 
acute febrile and infectious conditions, or to the presence in the 
blood of certain toxins which destroy the corpuscles. It is char- 
acterized during life by an excess of urobilin and iron in the urine, 
and by the development of hematogenous jaundice. 


In order to understand the process of hemol- 
EHRLICH'S ysis it is necessary briefly to refer to Ehr- 
SIDE-CHAIN lich's side-chain theory of immunity, 2 the hy- 
THEORY. pothesis which furnishes the best explanation 
of the organism's reaction against bacteria, 
toxins, and other noxious agencies. According to this theory, it 
is assumed that the body cell consists of a central group of mole- 
cules by virtue of which the inherent characteristics of the cell are 
determined and maintained, and a second, subsidiary molecular 
group, which, by means of its unsatisfied affinities, is capable of 
combining with nutrient materials, toxins, and other substances, 
which are thus brought into intimate relationship with the cell 

1 Mackenzie, Lancet, 1891, vol. i, p. 73. 

2 Proc. Roy. Soc., London, 1900, vol. Ixvi, p. 424; Nothnagel's "Spec. Path, 
u. Ther.," 1901, vol. viii, p. i; Klin. Jahrb., 1898, vol. vi, p. 299; also Ehr- 
lich and Morgenroth, Berlin, klin. Wochenschr., 1899, vol. xxxvi, pp. 6 and 481; 
ibid., 1900, vol. xxxvii, pp. 453 and 681; ibid., 1901, vol. xxxviii, pp. 251, 569, 
and 598. 


structure. These subsidiary molecules, known as side-chains or 
receptors, are simply links between assimilable substances and the 
cell, which can neither be nourished nor poisoned except by the 
intermediation of its receptors. Just as receptors possess specific 
affinities for linking with certain food stuffs and with no others, 
so there is believed to be a congenial reaction between receptors 
and toxins, certain varieties of receptors combining with one kind 
of toxin and other varieties with another. The receptor and the 
toxin, therefore, must be homologous before the two can combine. 
The receptors concerned in the process of toxin immunity belong 
to Ehrlich's "first order"; they have but a single uniting bond for 
combining with the toxin, and for this reason are also termed 


~ Tcocophore. 


Body cell. 


A toxin molecule consists of two atomic groups, each with dif- 
ferent affinities and with separate functions. One, termed the 
haptophore group, serves to combine the toxin unit with the cell 
receptor for which it has a selective affinity; the second, known as 
toxophore group, serves to injure the cell when the haptophore- 
receptor combination is formed. The mechanism of this anchoring 
of the toxin to the vulnerable cell may be represented by the above 
diagram (Fig. 41). 

The union of toxin and cell leaves the latter more or less crippled, 
either so severely that it dies, or but slightly, so that it still lives. 
If the latter be the case, the cell commences to elaborate more re- 
ceptors in its combat against the toxin elaborates them in great 
excess of its needs, so that in spite of the fact that many of these 
new-born receptors are promptly seized by other toxin molecules, 


some of them are thrown off by the cell and float off free in the 
blood and other body fluids. These liberated receptors, or hap- 
tins, constitute antitoxin. In their free state they can combine 
with homologous toxin molecules just as readily as when still at- 
tached to the cell, and such a combination obviously renders inert 
the toxin, since its haptophore group is thus satisfied and cannot 
now become anchored to the cell. According to Welch's hypoth- 
esis, these antidotal substances may act not only in their primary 
function as toxin neutralizers, but, under certain conditions, may 
irritate the invading bacteria to elaborate similar substance for 
their own protection. In other words, this theory of reciprocity 
in infection assumes that if bacteria irritate the body cells, the latter 
in turn similarly affect the bacteria. 

Toxins deprived of their toxophore groups, but retaining their 
haptophore group, are designated toxoids, and such bodies, though 
capable of becoming attached to the cells by their haptophore 
group, are inert, because they contain no toxic element. 
Toxoids can also unite with antitoxin (free receptors or hap- 
tins) by means of their haptophore link. Toxins incompletely 
combined with antitoxins, and therefore still capable of causing 
modified poisonous effects, are known as toxones. The following 
diagram illustrates the production of antitoxin and the union of 
toxins and toxoids with the cell and its liberated receptors : 

vi Toxins united 
,/\ with cell. 

Free recep- 
tors; haptin; <- 
antitoxin. './X 

Toxins united with 

Body cell. 

The injection into animals of bacteria, various 

HEMOLYSIS. body cells, and the products of these substances 

in time gives rise to the development of specific 

"antibodies" in the blood serum of the treated animal, as it ac- 


quires immunity. When the latter is complete, the animal's blood 
will be found to have acquired a lytic or destructive action upon 
cells similar to those injected. For example, the injection of ery- 
throcytes gives rise to hemolytic serum; of bacteria, to bacterio- 
lytic serum for the specific organism introduced; of bone marrow 
and lymphatic tissue, to leucolytic serum ; of spermatozoa, hepatic 
cells, and epithelial cells, to spermolytic, hepatolytic, and epi- 
theliolytic sera, respectively. The adaptive changes thus pro- 
duced in the animal injected are absolutely specific the injection 
of erythrocytes produces a serum acting only upon the erythro- 
cytes, not upon the leucocytes, bacteria, or body cells. 1 

Isolysis is a term used to denote a hemolytic action due to the 
injection of one animal with cells from another animal of the same 
species, the active factors of this process being known as isolysins. 
The destructive influence of a normal individual's blood upon the 
erythrocytes of another person illustrates this phase of hemolysis. 
Autolysis should result, theoretically at least, by the immunization 
of an animal against injections of his own cells, with the conse- 
quent evolvement of substances termed autolysins. The occurrence 
of hemolysis in icterus, in Winckel's disease, and after internal 
hemorrhage is suggestive of autolysis. 

The lytic property of the blood Ehrlich attributes to the inter- 
dependent action of two distinct elements of the plasma, the com- 
bined influences of which are essential to produce the change. In 
the process of hemolysis the receptors of the erythrocytes serve 
to connect these cells, by the interposition of an intermediary 
body, or amboceptor, with a complementary toxic body, or com- 
plement, which, when thus united to the cells, exerts its destructive 
influence. The receptors concerned in hemolysis are said to be- 
long to the "third order"; they are provided with two haptophore 
groups, one for combining with the vulnerable cell and the other 
with the complement, which then can act upon the cell attached to 
the first haptophore group. 

The amboceptor is formed within the body, as the result of a 
cellular hyperactivity excited by the organism's adaptation to 
alien blood or to other irritant and toxic material. Amboceptors 
are thought to represent liberated "third order" receptors, evolved 
and cast off by the cells during the process of adaptation. They 
are stable substances, capable of progressive increase, and act as con- 
necting links between the complement and the cell. This function 

1 For a complete account of these reactions see: (i) Vaughan and Novy, 
"Cellular Toxins," Philadelphia, 1902; (2) Welch, Huxley Lecture, Lancet, 1902, 
vol. ii, p. 977; (3) Prudden, Med. Rec., 1903, vol. Ixiii, p. 241; (4) Aschoff, Zeitschr. 
f. allg. Physiol., 1902, vol. i, p. 69; (5) Wassermann, " Immune Sera," Eng. transl. 
by Chas. Bolduan, New York, 1904. 



of anchoring the complement to the cell is performed by means of 
two groups of atoms (hence the term, amboceptor), one with an 
affinity for the corpuscle (cytophilic group), and the other with 
an affinity for the complement (complementophilic group}. The 
complement, which acts upon the erythrocyte partly as an enzyme 
and partly as a toxin, is normally present in the blood, being 
probably derived largely from the leucocytes. It is provided 
with a haptophore group of atoms, which link it to the ambo- 
ceptor, and with a zymophore group (corresponding to a toxin's 
toxophore group) upon the action of which depends the cellular 
destruction. The complement is of unstable nature, being de- 
stroyed at a temperature of 55 C., and possesses many of the 
characteristics of the enzyme. Its toxic action upon the erythro- 
cytes is exhibited only when it is anchored, to them by the inter- 
vention of the amboceptor, so that unless this link is formed, the 
complement can exert no deleterious effect upon these cells. The 
accompanying diagram (Fig. 43) shows the mechanism by which 
an erythrocyte succumbs to the zymotoxic action of the comple- 

Union of comple- 
ment, amboceptor, 
and cell. 


J- Complemento- 

" Cytophile. 



Resistance of the erythrocytes to hemolysis 

ANTI- is attributed to the protective influences of sub- 

HEMOLYSIS. stances known as antihemolysins, which in their 

action correspond to antitoxins. Antihemolysins 

are formed within the blood plasma after inoculation with hem- 

olysins, and are of two kinds: anticomplements and antiamboceptors. 

The former combine with the haptophore group of the complement, 



and the latter unite with the cytophilic group of the amboceptor 
combinations which in either instance break the continuity of the 
complement-amboceptor-erythrocyte chain essential for the cell's 
destruction. The action of these two forms of antihemolysins may 
be illustrated thus: 


Antiamhoceptor. Anticomplement. 


A. Interference of anticomplement with complement-amboceptor union. B. Interference of 
antiamboceptor with amboceptor-cell union. C. Antiamboceptor-amboceptor union. D. Anti- 
complement-complement union. 

Hemolysis, although generally associated with 
AGGLUTINA- more or less agglutination and precipitation of the 
TION AND PRE- erythrocytes, is not always part and parcel of 
CIPITATION. these phenomena. Serum may clump and pre- 
cipitate the cells without exerting the slightest 
hemolytic effect, while, on the other hand, dissolution of the cells 
may progress without their becoming clumped or precipitated. 
The principles of agglutination and precipitation have been applied 
clinically in the diagnosis of enteric fever and other specific in- 
fections, and in the biological test for the detection of human 
blood (q. v.). 

The injection of bacteria, alien erythrocytes, and other foreign 
cells produces in the blood serum of the animal thus treated the 
property of agglutinating and precipitating, in vitro, the bacteria or 
cells injected. These phenomena are attributed to the presence, 
in the animal's serum, of substances known as agglutinins and 
precipitins, developed during the process of adaptation or im- 


munization. According to Ehrlich's theory, agglutinins are 
liberated receptors elaborated by the body cells concerned in 
the process, such receptors having a haptophore group with which 
the bacterium or cell, as it may be, combines, and a zymophore or 
agglutinophore group, which, when this link is formed, exhibits 
its clumping properties. It will be noted that agglutination, unlike 
hemolysis and bacteriolysis, does not involve the action of a com- 
plementary toxic body the receptor at once anchors and clumps 
the homologous cell or bacterium through the combined offices 
of its haptophore and zymophore atomic groups. Receptors 
of this sort belong to Ehrlich's "second order." An agglutinoid 
is the term used to designate a receptor deprived of its zymophore, 
agglutinating group, but retaining its haptophore, combining group. 
It is analogous to a toxoid, and, like this substance, is developed 
by heat. 

Precipitins are elaborated by the injection of albuminous 
body fluids, such as defibrinated blood, into certain animals whose 
blood serum, in consequence, acquires the property of precipitating, 
in vitro, the albumins against which the adaptation is directed. 
The serum of a rabbit adapted to human blood precipitates the 
blood of man but none other, except, in low dilutions, the blood 
of certain monkeys! (See p. 117.) Similarly, the injection of 
milk from one animal develops in the serum of the animal treated 
a precipitin which is specific for the milk used in the adaptation, 
but for milk of no other species. " Antisera," as they are termed, 
can also be prepared for various warm- and cold-blooded animals, 
and such fluids precipitate only the blood of the animals for which 
adaptation is sought, save in the case of closely related species, 
in which a partial, incomplete reaction may occur in a low dilution. 
Other albuminous body fluids (pus, albuminous urine, inflam- 
matory exudates, and saliva) are capable of evolving, after injec- 
tion, sera which precipitate the blood of the species from which 
they were derived. The mechanism of precipitation is analogous 
to that of agglutination, the precipitins consisting of liberated 
"second order" receptors provided with a haptophore, combining 
group, and a zymophore, precipitating group of atoms which 
together act upon the susceptible cell. 




FIG. 1. 




FIG. 3. 



FIG. 4. 




FIG. 5. 


(Figs. 1,3,3, and 4, Triacid Stain ; Fig. 5, Eosin and Methylene-blue. ) 
(E. F. 

(Triacid Stain.) 
Fig. 1. Normal Erythrocytes. 

Fig. 2. Erythroblasts. 

1. Microblast. Note the dense, glistening nucleus, and the scanty, ragged zone of proto- 


2, 3. 4. 5. 6. Normoblasts. The process of partial nuclear textrusion is apparently shown 

in 3 and 5; in the latter cell the basic affinity of the nucleus is singularly slight. The 
cell, 6, while as large as many megaloblasts, retains the nuclear characteristics of the 
normoblast, of which it represents perhaps a hydropic form. Some writers regard 
such erythroblasts as megaloblasts, on account of their large size. 

7, 8, 9, 10, n, 12, 13. Megaloblasts. In 7 the nucleus, while normoblastic in size, is 
megaloblastic in structure and in staining affinity. Note the variation in the size of 
these cells, their delicate nuclear chromatin, and their decided tendency toward 
polychromatophilia. In all the nucleus and protoplasm are separated by a conspicu- 
ous hyaline ring. 

Pig. 3. Erythroblasts with Multiple Nuclei. 

1. Cell with a constricted, convoluted nucleus, apparently undergoing solution in the 


2. Normoblast with three nuclei arranged somewhat in the form of a clover-leaf. 

3. Cell with two large nuclei, each apparently in an early stage of extrusion. Note the 

affinity for fuchsin displayed by the protoplasm and by the upper nucleus, and the 
distinct hyaline zone encircling the lower one. 

4. 5. Normoblasts in karyokinesis. 

Pig. 4. Erythrocytes Deformed in Shape and Size, 
i, 2. Microcytes. 
3. Megalocyte. 

4. 5i 6. 7. 8, 9, 10, ii, 12, 13. Poikilocytes. Many of these cells are highly polychromato- 
philic, especially n, 12, and 13. 

(Eosin and Methylene-blue.) 

Fig. 5. Erythrocytes Showing Degenerative Stroma Changes. 

Granular basophilia is shown by i and 2 ; extreme decolorization by 3, 4, and 5. The other 
cells represent various stages of hemoglobin loss and protoplasmic degeneration. 




Hemoglobin, which occurs in the circulating 
GENERAL blood in chemical union with oxygen as oxyhem- 
PROPERTIES. oglobin, is an extremely complex ferruginous and 
albuminoid substance contained within the stroma 
of the erythrocytes. It constitutes approximately nine-tenths of 
their total bulk, and a trifle less than 14 per cent, of the whole 
blood. Hemoglobin displays a striking avidity for combining 
with oxygen to form a peculiarly unstable, but definite, chemical 
compound, and a similar tendency to yield up to the tissues 
much of its oxygen during its passage through the capillary 
circulation. Under the influence of deoxidizing agents oxyhemo- 
globin may be deprived of its loosely combined oxygen molecule, 
the resulting oxygen- free constituent being known as reduced 
hemoglobin. Rhombic crystals of oxyhemoglobin, scarlet or 
reddish-green in color, are rapidly formed if, for any reason, 
separation of this substance from the corpuscular stroma takes 
place. These crystals may be easily demonstrated by Reichert's 
method l of laking a small quantity of blood with ether and then 
adding from one to five per cent, of ammonium oxalate. Met- 
hemoglobin is an oxygen compound of hemoglobin containing the 
same quantity of combined oxygen as the latter, but differing from 
it in holding its oxygen constituent in. a more intimate union. 
The dingy brown color which develops in a solution of oxyhemo- 
globin after prolonged exposure to the atmosphere evidences the 
production of this variety of blood pigment. (See " Methemo- 
globinemia," p. 167.) 

The amount of iron (in the form of hemochromogen) which 
hemoglobin contains is considerable somewhat in excess of 
four per cent. It has been shown, clinically, by estimates made 
with the ferrometer and the hemometer, that no fixed parallelism 

1 Amer. Jour. Physiol., 1903, vol. ix, p. 97. 
ii 161 


is maintained between the percentage of hemoglobin and the iron 
contained in the blood. 1 

Under the action of acids, strong alkalis, or heat, hemoglobin 
may be readily decomposed into two constituents: hematin, or 
an iron-containing principle; and an albuminous residue of un- 
known character, but somewhat resembling globulin. In combi- 
nation with hydrochloric acid hematin forms a crystalline hydro- 
chlorid of hematin termed hemin, or Teichmanri's crystals. Under 
the microscope these crystals appear as black or dark brown, 
elongated, rhombic prisms belonging to the triclinic system, which 
are insoluble in water, alcohol, ether, chloroform, and dilute 
acids. They may be demonstrated by preparing a slide of blood 
(or of any dried substance containing blood pigment) to which a 
small quantity of common salt has been added; a drop of glacial 
acetic acid is then run beneath the cover-glass, so that it mixes 
with the blood and salt, and the specimen thus prepared is heated 
to just below the boiling point over a Bunsen flame. On cooling, 
Teichmann's crystals may be seen under the microscope with a 
low-power dry objective. (See p. 116.) 

Iron-free hematin, or hematoporphyrin, may be derived from 
blood by the admixture of concentrated sulphuric acid. This 
substance is closely related chemically to urobilin, and occurs oc- 
casionally as a pigment in nature and in normal and pathological 
urines. Hematoidin, which is also free from iron, occurs in the 
form of reddish, rhombohedral crystals, only in old clots resulting 
from blood extravasations, such as cerebral hemorrhages and 
splenic infarcts. It is derived from hematin, and is probably 
identical with bilirubin. 

The chief source of hemoglobin is the iron 

ORIGIN. contained in various food products, about 10 

mgm. daily representing the amount of this 

metal ingested in an ordinary diet, according to the analyses 

of Stockman. 2 In the event of a stoppage of this source of an 

iron supply, the formation of hemoglobin may proceed from the 

supply of iron stored up in the various organs of the body, notably 

in the liver. Bunge 3 has shown that in the young infant, whose 

natural food, milk, contains but a slight trace of iron, this source 

of hemoglobin manufacture is most potent. ' 

The recent experiments of Aporti 4 regarding the origin of 
hemoglobin and the erythrocytes have shown that animals sub- 

1 Rosin and Jellinek, Zeitschr. f. klin. Med., 1900, vol. xxxix, p. 109. 

2 Jour. Physiol., 1897, vol. xxi, p. 55; ibid., 1895, vol. xviii, p. 484. 

3 Zeitschr. f. physiol. Chem., 1892, vol. xvi, p. 177. 

4 Centralbl. f. inn. Med., 1900, vol. xxi, p. 41. 


jected to repeated bleedings and kept on an iron-free diet are able, 
up to a certain point, to utilize the supply of body iron for hemo- 
globin manufacture; but that when such a demand became so 
great that this supply was exhausted, the red corpuscles became 
progressively paler and paler, and the animal finally died. During 
the course of these experiments, if the animal received injections 
of iron, a prompt and striking increase in hemoglobin occurred, 
the gain ranging from 50 to 95 per cent, within a week's time. 
The injection of arsenic, on the contrary, produced no effect upon 
the hemoglobin percentage, although it caused an increase in the 
number of erythrocytes. The investigations of Stockman and 
Charteris 1 show that arsenic does not stimulate hemogenesis in 
the bone marrow, and that its favorable action in malarial fever 
and diseases of this class is probably due to its parasiticidal effect. 
In animals injected with small doses of arsenic the marrow changes 
consisted of hyperemia, decrease in the number of giant cells and 
fat cells, increase in the leucoblasts, and slight, if any, prolifera- 
tion of erythroblasts. 

Baumann, 2 studying the effects of iron carbonate, iron albumin- 
ate, and arsenic upon regeneration of the blood in dogs after hem- 
orrhage, arrived at these conclusions : that the administration of 
either Blaud's pill or an albuminate of iron causes a rapid gain in 
hemoglobin, even to a higher figure than that originally found 
before the blood loss, the effects of each of these preparations of 
iron being similar; that arsenic and iron combined, although 
stimulating hemoglobin formation less than the above drugs, 
are better general hemogenetic agents, so far as regeneration of 
the corpuscles, proteids, and plasma is concerned; and that 
arsenic, when given alone, is but an indifferent blood-builder. 
The practical import of these experiments is patent. 

Diminution in the amount of hemoglobin, as 
VARIATIONS indicated by the hemometer, is known as oligo- 
IN AMOUNT, chromemia, or achroiocythemia. It is a condition 
usually, but not invariably, associated with a cor- 
responding decrease in the number of erythrocytes. An ap- 
parent increase in the hemoglobin percentage may result from the 
concentration of the blood caused by a reduction in the quantity 
of blood plasma consequent to excessive drains upon the liquids 
of the body. By a similar physical mechanism factors produc- 
ing a dilution of the blood are capable of causing an apparent 
diminution in the hemoglobin. Marked oligochromemia is com- 
monly observed in chlorosis, pernicious anemia, and leukemia; 

1 Jour. Path, and Bacteriol., 1903, vol. viii, p. 443. 

2 Jour. Physiol., 1903, vol. xxix, p. 18. 


and in the secondary anemias dependent upon such factors as 
hemorrhage, mineral poisoning, acute and chronic infections, 
malignant neoplasms, and constitutional diseases. The behavior 
of the hemoglobin under such conditions is more fully alluded to 
in connection with the lesions in question. Poggi, 1 from a series 
of experiments upon normal women, has shown that the hemo- 
globin is slightly lowered (10 or 15 per cent.) for a few days before 
menstruation, but with the establishment of the flow the oligo- 
chromemia soon disappears. The primary loss he attributes to 
retarded hemogenesis consequent to the lessened consumption of 
albumin occurring in menstruating women, while the subsequent 
gain he explains by the increased functional activity of the hemato- 
poietic organs. Double oophorectomy in sexually active bitches 
is followed by a distinct oligochromemia and by a less marked 
oligocythemia, persisting for from two to six weeks, but in old 
dogs the operation has no such effect. This experimental evi- 
dence is advanced by Breuer and von Seiller 2 in corroboration 
of the ovarian theory of chlorosis. 

In passing, it may be of interest to compare the degree of 
hemoglobin loss in the various forms of anemia, as illustrated by 
the following averages determined by the writer: 

Average of 50 estimates in pernicious anemia 25.5 per cent. 

" chlorosis 43.2 " 

' leukemia - 39.4 

" secondary anemia 55.2 

Bierfreund's investigations 3 in Mikulicz's clinic have led to the 
current impression among surgeons that it is highly dangerous to 
give a general anesthetic to a patient 'whose hemoglobin percent- 
age is below 30; some operators regard 40 per cent, as the 
lowest limit of safety, and refuse to employ any but a local anes- 
thetic in cases with an oligochromemia exceeding this figure, 
except under circumstances of imperative necessity. Any one, 
however, who has attempted to verify the correctness of this gen- 
eral belief must accept it with a shrug of the shoulders. The 
writer knows of numerous patients whose hemoglobin percentages 
all were below 30 in whom operations under general anesthesia with 
ether were followed by uneventful recovery; in one instance (a 
pan-hysterectomy lasting more than an hour and a half) the 
hemoglobin was but 21 per cent., yet no ill effects were observed. 

Patients with hemoglobin percentages of from 15 to 30 have 

1 Policlin. Roma, 1899, vol. vi, p. i. 

2 Wien. klin. Wochenschr., 1903, vol. xvi, p. 869. 

3 Langenbeck's Arch., 1890-91, vol. xli, p. i. 


been successfully operated by Girvin, Shober, Le Conte, Noble, 
Baldy, J. C. Da Costa and others. 1 

Assuming that in the normal adult 14 gm. 
ABSOLUTE represent the average amount of hemoglobin in 
AMOUNT. 100 gm. of blood, the absolute amount of 
hemoglobin may be readily calculated thus: 

Hemoglobin percentage X 14 -r- 100 = Grams of hemoglobin in 100 gm. of 


For example, in blood in which the percentage of hemoglobin, 
as determined by the hemometer, is found to be 40, the calcula- 
tion (40 X 0.14) gives the absolute amount of hemoglobin as 
5.6 gm. 

The proportionate amount of hemoglobin con- 

COLOR tained in each erythrocyte, or its corpuscular 

INDEX. richness in hemoglobin, is known as the color 

index, or blood quotient, or valeur globulaire. 

In normal blood the color index is theoretically expressed by the 

figure i, although, practically, it varies from 0.95 to 1.05 in men, 

and from 0.9 to i in women. 2 

In those anemias in which the decrease in the amount of 
hemoglobin in the blood is coincident with a proportionate de- 
crease in the number of erythrocytes, the color index remains 
practically at the normal figure. If, however, the cellular de- 
crease happens to be relatively greater than the hemoglobin 
loss, then the index will naturally be found to rise above normal ; 
thus, in pernicious anemia, in which condition the loss of cells 
is proportionately much greater than the loss of hemoglobin, 
high color indices, approaching or even exceeding 1.25, are fre- 
quently observed. On the contrary, if the hemoglobin loss is rela- 
tively more excessive than the corpuscular decrease, the color 
index falls below normal; for example, in chlorosis, in which, as 
a rule, the decrease affects the hemoglobin much more strikingly 
than the erythrocytes, low indices, such as 0.50 or less, are common. 

To calculate the color index, the percentage of hemoglobin is 
divided by the percentage of erythrocytes, the "result being ex- 
pressed in decimals. In order to Simplify this procedure 5,000,- 
ooo erythrocytes per must be arbitrarily considered as 
normal, or 100 per cent. To obtain the percentage of cor- 
puscles the actual number counted in one of blood is 
simply multiplied by two, and two or three decimals pointed off 
from the left, depending upon whether the count is below or 

1 Trans. Coll. of Phys. of Phila. (Sect, on Gynecology), 1902, vol. viii, p. 26; 
also Amer. Jour. Obstet., 1902, vol. xlv, pp. 666 and 701. 

2 Oliver, loc. cit. 


above the normal 5,000,000. The following examples serve to 
illustrate the calculation in several conditions: 
Normal Adult. 

Erythrocytes : 5,000,000 per c. mm. (100 per cent.). 
Hemoglobin: 100 per cent. 

100 -e- loo = i : Color index. 
Secondary Anemia. 

Erythrocytes: 2,650,000 per (53 per cent). 
Hemoglobin: 40 per cent. 
4 ~ 5 ~ 53 -75 : Color index. 
Pernicious Anemia. 

Erythrocytes: 840,000 per (16.8 per cent.). 
Hemoglobin: 18 per cent. 
18 -4- 16.8 = 1.07: Color index. 

Erythrocytes: 4,100,000 per (82 per cent.). 
Hemoglobin: 32 per cent. 
32 -4- 82 = 0.39: Color index. 

These examples, of course, refer only to the usual blood find- 
ings, for the color index is by no means always high in pernicious 
anemia, nor always low in chlorosis. The color index shows 
simply the relative relations of the hemoglobin and the corpus- 
cular percentages. It is only suggestive, not diagnostic, of a 
specific blood disease. 

The term hemoglobinemia is used to designate 

HEMOGLO- a condition in which the hemoglobin is dissolved 

BINEMIA. from the corpuscular stroma as the result of 

some pathological factor, and is held in solution 

by the blood plasma. In extreme instances this condition is 

sooner or later succeeded by hemoglobinuria. 

Among the most potent causal factors of hemoglobinemia are 
certain drugs which act as blood poisons when administered in 
toxic doses, of which the following are examples: arseniuretted 
hydrogen, sulphuretted hydrogen, potassium chlorate, carbolic acid, 
hydrochloric acid, sulphuric acid, pyrogallic acid, nitrobenzol, anti- 
mony sulphid, iodin, naphthol, and many of the coal-tar deriva- 
tives, such as acetanilid, antipyrin, and phenacetin. A similar 
liberation of the hemoglobin may be observed as the result of 
poisoning by certain varieties of mushrooms, by some snake- 
venoms, by the bite of scorpions, and by a number of vegetable 
glucosids. Sunstroke, extensive burns, and exposure to excessive 
cold are also capable of giving rise to hemoglobinemia. Experi- 
mentally, hemoglobinemia may be produced by the transfusion 
oj blood from one animal into the circulation of another belonging 
to a different species. 



Hemoglobinemia is observed with more or less constancy in a 
number of acute infectious diseases, such as grave cases of septice- 
mia, diphtheria, malignant jaundice, syphilis, malarial fever, enteric 
fever, scarlet fever, yellow fever, typhus fever, and variola. It also 
may occur in scurvy and in Raynaud's disease, and is a prominent 
blood rinding in those two obscure conditions known as epidemic 
hemoglobinuria of the new-born and paroxysmal hemoglobinuria. 

Hemoglobinemia may be readily detected by the following 
method, recommended by von Jaksch : 1 A small amount of blood, 
drawn from the patient by means of a cupping-glass, is imme- 
diately placed in a refrigerator, in which it is allowed to remain 
for twenty-four hours. In normal blood the serum which separates 
at the expiration of this period is of a perfectly clear straw-color, 
whereas if hemoglobinemia exists, the serum is colored a beautiful 
ruby-red. If this hemoglobinemic serum is examined with the 
spectroscope, the two characteristic absorption bands of oxy- 
hemoglobin may be observed. If it is coagulated by heat, a 
deep brown color is imparted to the coagulum. 

Methemoglobinemia, or the presence in the 
METHEMOGLO- circulating erythrocytes of methemoglobin, is 

BINEMIA. produced by the action of a number of toxic 
substances, which, if given in sufficiently massive 
doses, may seriously or fatally cripple the oxygenating functions 
of the blood. Among the agencies which cause this conversion 
of oxyhemoglobin into methemoglobin are potassium chlorate, 
anilin, iodin, bromin, ether, turpentine, acetanilid, potassium per- 
manganate, hydrochinon, kairin, thallin, and pyrocatechin. The 
inhalation of amyl nitrite and the intravenous injection of sodium 
nitrite also act in a similar manner. Henri and Mayer 2 found that 
methemoglobinemia could be produced by the influence of radium 

Spectroscopical examination of the blood is essential for the 
detection of methemoglobinemia. The spectrum of methemo- 
globin in alkaline solution shows three absorption bands: one 
well-marked band between C and D of Fraunhofer's lines and 
two others of much less distinct appearance, lying between D 
and E, each immediately adjacent to the lines. In acid and neu- 
tral solutions the spectrum of methemoglobin shows four absorp- 
tion bands: a decided one between C and D, two between D and 
E, and one closely adjacent to F. This spectrum, it is true, is 
identical with that produced by an acid solution of hematin, but 
it may be easily distinguished from the latter by the fact that 

1 "Clinical Diagnosis," 3d ed., London, 1807, p. 75. 
J Sem. med., 1904, vol. xxiv, p. 68. 


the spectrum of methemoglobin, when acted upon by ammonium 
sulphid, changes first to that of oxyhemoglobin, and later to 
that of reduced hemoglobin, while when hematin is thus treated, 
a spectrum which shows two bands between D and E is pro- 

Aside from the bright, cherry-red color of the 

CARBON MON- blood in coal-gas poisoning the presence of car- 

OXID HEMO- ban monoxid hemoglobin may be determined by 

GLOBIN. spectroscopical examination and by a number of 

distinctive chemical reactions. 

Recalling the characteristic spectrum of oxyhemoglobin (two 
distinct absorption bands between D and E, the one nearest D 

Oxyhemoglobin . 

Methemoglobin . 

Reduced Hemo- 

CO Hemoglobin. 


being darker, narrower, and more sharply defined), it is found that 
in the spectrum of carbon monoxid hemoglobin these bands 
are replaced by two others, also between > and E, but nearer 
together, and somewhat closer to the violet end of the spectrum. 
This distinction, which may be so slight as to appear confusing, 
is at once emphasized by the fact that the addition of ammo- 
nium sulphid has absolutely no effect upon the carbon mon- 
oxid spectrum, while it transforms the spectrum of oxyhemoglo- 
bin into that of reduced hemoglobin. 

Carbon monoxid hemoglobin in the blood is also demonstra- 


ble by the following simple test devised by Hoppe-Seyler * : A 
small quantity of blood, removed from the patient by means of 
a wet-cup, is mixed with twice its volume of a 10 per cent, solu- 
tion of potassium hydrate. Thus treated, blood containing car- 
bon monoxid hemoglobin changes the color of the mixture to a 
rich cinnabar-red, while with normal blood the solution turns 


The erythrocytes or red corpuscles are thin, 
APPEARANCE flattened, biconcave discs, of sharply denned, 
IN regular outline and of smooth, even surface. 

FRESH BLOOD. That they are neither bell-shaped nor globular, 
as is also maintained, is obvious on careful 
microscopical examination. In the blood of the normal indi- 
vidual the erythrocyte does not possess a nucleus. When the 
corpuscle is examined microscopically as it rests upon its flat sur- 
face, its central concavity is plainly indicated by a dark central 
area surrounded by a narrower, lighter rim as the periphery of the 
cell is brought into sharp focus, changing to a pale, white center 
encircled by a darker periphery as the objective is brought closer 
to the corpuscle. When viewed in profile, it is shaped somewhat 
like a slim dumb-bell, with regularly rounded poles tapering 
from either end toward a shallow central concavity on either 
surface. The color of the cells, when examined singly under the 
microscope, is a pale greenish-yellow, but when they are col- 
lected in masses, a more or less marked reddish tint becomes ap- 
parent. The erythrocytes possess a peculiar tendency of collect- 
ing and adhering in more or less regularly arranged piles, like 
rolls of coins stacked up face to face, this being known as rouleaux 

After withdrawal of the blood from the body various structural 
changes in the erythrocytes, commonly known as crenation, may 
be observed. In normal blood the .rapidity with which these 
changes progress depends upon the quantity of air which leaks 
in between the slide and the cover-glass, and thus causes de- 
generation of the corpuscular stroma. The development of one 
or more small, bright, highly refractive spots in the body of the 
corpuscle, and a slight indentation of the cell's periphery are the 
the most conspicuous indications of beginning crenation. As the 
process goes on, more and more of these hyaline points develop, 

1 Loc. cit. 


until finally the whole surface of the corpuscle becomes thickly 
studded with glistening, bead-like spines. As the stroma be- 
comes drier and drier, its typical biconcavity and sharply-cut out- 
line are lost, contracting strands of the stroma are seen to extend 
from point to point among the beaded projections, the periphery 
of the cell changes to a cogged rim, and finally the cell becomes 
shrunken and shriveled up into a small, many-starred asterisk. 
Some of the erythrocytes become fragmented, and small bits of 
their stroma are observed to break off and float through the plasma. 
Others become progressively paler and paler as the hemoglobin 
is dissolved out, until complete decoloration occurs. Still others 
become distorted into designs of every conceivable shape, so 
that their resemblance to the normal cell becomes most remote. 
These changes, which never occur in normal blood until the cells 
have been exposed to prolonged atmospheric influence, must 
not be confused with similar alterations in the structure of the 
erythrocytes occurring as the result of pathological states of the 
blood. The latter changes are described more fully in another 
place. (See p. 184.) 

The finer structure of the erythrocyte is still 
HISTOLOGICAL a moot point among different histologists, the 
STRUCTURE, view most generally accepted regarding it as a 
homogeneous cell composed of an insoluble 
spongy network, the stroma of Rollet, in the interstices or trabec- 
ulas of which is embedded a soluble, finely granular substance, 
the hemoglobin, existing probably as a compound with some un- 
known constituent of the cell. In lieu of a distinct limiting 
membrane the portions of the stroma nearest to the surface of 
the corpuscle are condensed, to protect it from injury during its 
movement through the blood stream. This outer layer, accord- 
ing to Peskind, 1 is composed of a proteid substance, lecithin, 
and cholesterin, but contains no hemoglobin. The corpuscles are 
highly elastic and contractile, to permit of the rapid and marked 
temporary distortions of shape which they constantly undergo in 
the circulating blood. 

Other authorities, notably Schafer, 2 disagree with Rollet's view, 
inclining rather to consider the erythrocytes as vesicular masses, 
consisting of an external envelop inclosing a fluid contents. 
Thus, Schafer believes that the cell consists of two distinct por- 
tions, a colored and a colorless, the former being a solution of 
hemoglobin, while the latter, or so-called stroma, consists chiefly 
of lecithin and cholesterin, together with a small amount of cell 

1 Amer. Jour. Physiol., 1903, vol. viii, p. 404. 

2 Quain's "Anatomy," Philadelphia, 1891, pt. 2, p. 210. 


globulin. Without attempting to discuss the correctness of either 
of these two views, a single tangible reason for regarding the 
corpuscle according to Rollet's opinion may be stated, viz.: the 
fact that exposure of blood to destructive temperatures results in 
fragmentation of the corpuscles into numerous minute portions, 
each one of which consists of a bit of hemoglobin-containing 
stroma. This obviously seems to disprove the existence of a 
limiting membrane, without further investigation. 

In the human body an active manufacture of 

ORIGIN AND erythrocytes constantly goes on during health, 

LIFE HISTORY, in order to compensate for the continuous 

drain on their number by the destruction of those 

cells which have become incapable of function and useless, their 

life cycle being run. That this reproduction is the direct answer 

to a call for new cells is proved by the prompt and rapid increase 

of corpuscles following the loss of blood from hemorrhage. That 

such a manufacture is attempted in severe pathological conditions, 

although the attempts are sometimes abortive, is evinced by the 

large numbers of immature and misshapen erythrocytes which 

appear in the blood in certain of the grave anemias. 

In the adult it is generally conceded that the erythrocytes 
are reproduced in the red bone marrow, being developed from 
their direct antecedents, the nucleated erythrocytes or erythro- 
blasts, which exist in this tissue in large numbers. The erythro- 
blasts appear to multiply in the thin-walled capillaries and veins 
of the red marrow, and, having lost their nuclei, become trans- 
formed into normally developed erythrocytes, which pass from the 
blood channels of the marrow into the general circulation. Some 
authorities have attributed to the spleen and lymphatic glands a 
share in the formation of the red cells, while others have main- 
tained that they may be transformed from the leucocytes in the 
circulating blood, but none of these theories has been associated 
with convincing evidence, so that it is fair to consider the red 
bone marrow the chief, if not the only, seat of production, in 
the light of our present knowledge of. the subject. Hayem's 
ingenious theory, that the erythrocytes arise from the hematoblasts, 
does not enjoy the confidence of modern investigators. 

When finally the erythrocyte, after having executed its function 
for a certain length of time, becomes useless in its primary office 
as an oxygen carrier, its death ensues, the destruction of the 
cell probably taking place largely in the liver and to a less degree 
in the spleen. Bain 1 has shown that this hemolytic power re- 
sides in both of these viscera, the liver acting chiefly upon the 

1 Jour. Physiol., 1903, vol. xxix, p. 352. 


erythrocytes and the spleen affecting mainly the leucocytes. 
After the passage of blood through the liver it was proved that 
the hemoglobin-deficient cells were most prone to destruction, 
that the hemoglobin content of the invulnerable cells was dis- 
tinctly increased, and that the perfusion appreciably augmented 
the quantity of iron in the liver and was accompanied by a consid- 
erable output of highly pigmented bile. The perfused spleen was 
also found to contain an increase in the amount of iron, indicative 
of a relatively slighter destruction of erythrocytes. Splenectomy 
in animals does not to any great extent interfere with erythro- 
cytic destruction (Lapicque). 1 Warthin's 2 studies show that de- 
struction of the erythrocytes also occurs in the splenolymph 
glands, minute vascular sinuses situated chiefly in the retroperi- 
toneal and mediastinal tissues, and in the thyroid and thymus 
regions. The possibility that certain of the partly degenerate 
cells also undergo a certain form of repair, first in the spleen and 
then in the liver, rendering them still capable of function, is an 
interesting but obviously unproved conjecture. 

The average diameter of the erythrocyte is 
SIZE. about 7.5 //, 3 its average thickness being about 

1.8 p. According to Gram, 4 the diameter appears 
to vary somewhat with the geographical and climatic conditions 
surrounding the individual, being considerably larger in inhabi- 
tants of northern countries than in southerners, as the following 
average measurements of this observer attest : 


Italy 7 to 7.5 p 

France , 7.5 to 7.6 /* 

Germany 7.8 fi 

Norway 8.5 p 

Hayem 5 distinguishes three different sizes: large, averaging 
8-5// in diameter; medium, averaging 7.5," in diameter; and 
small, averaging 6.5 // in diameter. Of these three classes, ap- 
proximately 75 per cent, are of the medium size, while 12.5 per 
cent, each are large and small. The diameter varies within some- 
what wider limits in the infant and in the young child than in the 
adult. It is, however, not materially influenced by sex. The 
pathological increase and decrease in the diameter of the erythro- 
cytes occurring in certain anemias are discussed in another place. 

1 Med. News, 1903, vol. Ixxxii, p. 311. 

2 Jour. Boston Soc. Med. Sci., 1901, vol. v, p. 414. 

3 The Greek letter fi is used to represent a micromillimeter, or T ^Vu f a 
millimeter, which is a standard unit of measurement used in microscopy. 

4 Fortschr. d. Med., 1884, vol. ii, p. 33. 5 Loc. cit. 


The normal number of erythrocytes in the 

NORMAL healthy male adult may be approximated at 

NUMBER. 5,000,000 to the of blood. Higher counts 

than this are frequently observed, however, 

especially in healthy, well-developed men, so that this figure 

should be taken to represent a rather low average, subject to an 

upward fluctuation of half a million cells, and occasionally even 

more. In females a count of about 4,500,000 erythrocytes per may be regarded as normal. 

Arterial and venous blood contain practically the same number 
of corpuscles, the apparent slight increase in favor of the latter, 
mentioned by some observers, being within the limits of technical 
error. For a like reason, under normal conditions, peripheral 
blood may be taken as representative of the blood of the entire 
body. Blood derived from dependent parts of the body contains 
a diminished proportion of corpuscular elements. Oliver's 1 
studies of this question have shown that blood from the finger 
invariably gives a higher count of erythrocytes than blood from 
the toe, this disparity being explained by the fact that the larger 
quantity of lymph gravitating to the more dependent parts of the 
body causes a dilution of the blood in these areas. 

This term has been applied by Capps 2 to the 

VOLUME figure representing the percentage volume of the 
INDEX. individual erythrocyte, in contradistinction to the 
color index, which expresses the amount of hemo- 
globin in the single cell. It is calculated by dividing the percent- 
age volume of the erythrocytes as a whole, obtained by centrifu- 
galization of the blood, by the percentage number of erythrocytes, 
as determined by the actual count with the hemocytbmeter, the 
normal volume index being taken as For example, the 
erythrocyte column, after centrifugalization with the hematokrit, 
reaches to the mark 40 on the capillary tube, indicating a total 
volume of 80 per cent.; while the count with the hemocytometer 
gives 3,000,000 cells per, or 60 per cent, of the normal num- 
ber. Then, 80-^-60, or 1.33, equals the volume index, a figure 
which in this instance shows an increase of 33 per cent, in the 
volume of each corpuscle. As a general rule, it may be. stated that 
the volume index and the color index, rise and fall together, 
although the parallelism between the two is not always closely 
maintained. The volume index is generally lowered in chlorosis, 
in leukemia, and in most of the secondary anemias, while in 
pernicious anemia it tends to rise above the normal standard. 

1 Loc. cit. 

7 Jour. Amer. Med. Assoc., 1900, vol. xxxvi, p. 464; also Jour. Med. Re- 
search, 1903, vol. v, p. 367. 


The cell's volume is influenced mainly by factors affecting cell 
degeneration, and is altered but slightly by osmotic influences. 
During blood regeneration the volume of the cell is restored before 
the color becomes normal. 


Polycythemia, associated with a proportion- 
AGE AND SEX. ately high percentage of hemoglobin, is found 
in the blood of the new-born infant immediately 
after birth, the maximum counts being observed some time during 
the first twenty-four hours of life, after which period they progres- 
sively diminish until, at the end of eight or ten days, about 
1,000,000 cells per have been lost. Each period of nursing 
is generally followed by a prompt temporary decrease in the 
count, and a similar change has been observed as the. effect' of 
premature ligation of the cord at birth. Hayem 1 found an aver- 
age of 5,368,000 erythrocytes per in 17 infants at birth, 
the highest count being 6,362,000, and the lowest, 4,340,000. 
Fehrsen 2 regards 6,000,000 per as the average count at 
birth. The cause of this polycythemia is attributed to concentra- 
tion of the blood from the abstraction of water by the tissues to 
replace the fluids of the body lost during the first few days of life. 
As soon as this loss is made up by the i-ngestion of a sufficient 
amount of liquids by the child, the normal relation between the 
liquid and the solid portions of the blood is reestablished, so 
that the polycythemia disappears. 

During the growth of the adult the average number of erythro- 
cytes continues to rise, until the maximum number is attained at 
some time between the third and fifth decades, after which a de- 
crease is observed, usually becoming more marked as the decline 
of life progresses. Schwinge 3 and others have shown that during 
the period of sexual activity the counts in females are generally 
lower than in males, but that after the climacteric the number of 
cells in the two sexes is practically identical. 

The influence of age and sex upon the number of erythrocytes 
is well illustrated in the following table prepared by Sorensen 4 : 


SJtoJS days 5,769,500 i to 14 days . . . .5,560,800 

5jyears 4,950,000 2 to 20 years. . . .5,120,000 

19.5 to 22 years 5,600,000 15 to 28 years .4,820,000 

25 to 30 years 5,340,000 41 to 61 years 5,010,000 

50 to 52 years 5,137,000 

82 years 4,174,700 

1 Loc. cit. 2 Jour. Physiol., 1903, vol. xxx, p. 322. 

3 Pfluger's Arch., 1898, vol. Ixxiii, p. 299. 4 Cited by von Limbeck, loc. cit. 


There are no conspicuous changes in the num- 

PREGNANCY, her of erythrocytes in any of these conditions. 

MENSTRUATION, In primiparae there is often a slight decrease in 

AND the number of corpuscles, particularly in the later 

LACTATION, months of pregnancy, but in multiparae this 
change is rarely observed. Distinct anemia dur- 
ing gestation is invariably pathological, although a moderate de- 
gree of hydremia may be physiological. J. Henderson 1 found 
an average hemoglobin percentage of 68.2 in 38 cases at term, 
and an average erythrocyte count of 3,975,348; by the eleventh 
day after delivery these values increased to 74 and 4,020,0x30, 
respectively. Postpartum oligocythemia commonly develops, the 
intensity of which depends largely upon the amount of blood 
lost and upon the general health of the woman; this loss of cells 
is gradually made up, and unless convalescence is delayed, reaches 
the normal by the second or third week after delivery. Bar and 
Daunay 2 have determined that the density of the blood progres- 
sively declines toward the end of pregnancy, but rapidly increases 
after delivery. 

During menstruation there may be a trifling reduction, caused 
by the physiological hemorrhage of the phenomenon, but the 
loss is rapidly made up in a few days' time. Sfameni 3 found that 
a transient polycythemia usually occurs shortly before the estab- 
lishment of the menstrual flow, and that the average loss of hemo- 
globin and corpuscles during the flow does not exceed 4.5 per cent., 
the decrease being in direct proportion to the actual volume of 
blood lost. 

In healthy, robust women lactation is accompanied by a 
normal count, but in weak, young girl-mothers, particularly 
those of the "chlorotic age," a moderate reduction is sometimes 

Here may be noted O. Schaffer's observation, 4 thus far uncon- 
firmed, that pregnancy can be detected by an increased affinity 
of the erythrocytes for iodin in blood obtained by puncture of the 
cervix. Using a reagent containing i gm. of iodin, 2 gin. of potas- 
sium iodid, and 300 c.c. of water, Schaffer found that the number 
of iodin-stained erythrocytes began to increase immediately after 
conception, until just before delivery they became at least twenty 
times more numerous than the unstained cells, which in the non- 
pregnant woman they outnumber only between two and five to one. 

1 Amer. Jour. Obstet., 1902, vol. xlv, p. 745. 

2 Sem. med., 1904, vol. xxiv, p. 28. 

8 Centralbl. f. Gynakol., 1899, vol. xxiii, p. 1311. 
4 Ibid., igoi, vol. xxv, p. 1375. 


Well-developed, robust individuals average a 

CONSTITUTION larger percentage of erythrocytes than the poorly 

AND nourished and weakly. In the former, counts 

NUTRITION, much in excess of 5,000,000, and in the latter 

counts of less than 5,000,000, are the rule. 
Fasting, inasmuch as it causes a drain upon the liquid elements 
of the vascular system, may rapidly bring about an apparent 
polycythemia due to concentration of the blood, this increase in 
cells being in direct relation to the length of abstinence from food. 
Hayem 1 states that a twenty-four hours' fast will cause a gain of 
between 400,000 and 500,000 cells; while the experiments of Reyne 2 
on a dog, starved to death after a twenty-four days' fast, showed 
an increase of 2,500,000 corpuscles at the expiration of this period. 
Active muscular exercise (gymnasium work, 
MUSCULAR walking, running, swimming, and so forth) pro- 
EXERCISE. vokes transient increase in the peripheral eryth- 
rocyte count, due primarily to increased blood 
pressure, which not only inspissates the blood, but also disseminates 
peripherally many cells which hitherto lay inactive in the deeper 
circulation. From the studies of Willebrand 3 and of Zuntz and 
Schumberg 4 it seems that the duration of this increase stands in 
inverse ratio to the length of the period of exercise. Hawke 5 
found that the average gain in erythrocytes was 21 per cent, after 
swimming, 16.6 per cent, after sprinting, 12.8 per cent, after walk- 
ing, and 12 per cent, after bicycling. 

Physical labor prolonged to the point of fatigue 
FATIGUE. appreciably diminishes the number of erythro- 
cytes. Cadet's 8 investigations of the blood of a 
number of peasants, examined after two months of hard field 
labor during the summer, showed a moderate oligocythemia in 
one instance a loss of over 1,000,000 cells, and in the others dimi- 
nutions averaging about one-half of this figure. Cadet believes 
that this anemia is referable to a true blood destruction, and notes 
as a rather mythical support of this view that the blood plaques 
were increased in these cases. 

Among other factors increasing the erythrocyte count may be 
noted cold tubbing (Winternitz ; 7 Thayer 8 ), warm baths (Knopfel- 
macher 9 ), and general massage (J. K. Mitchell 10 ). 

1 Loc. tit. 3 Cited by Hayem, loc. cit. 

3 Skandin. Arch. f. Physiol., 1903, vol. xiv, p. 176. 

4 " Studien zu einer Physiologic des Marsches," Berlin, 1901. 

5 Amer. Jour. Physiol., 1904, vol. x, p. 384. 

* Cited by Hayem, loc. cit. 1 Centralbl. f. klin. Med., 1893, vol. xiv, p. 177. 
8 Johns Hopkins Hosp. Bull., 1893, vol. iv, p. 37. 

* Wien. klin. Wochenschr., 1893, vol. vi, p. 810. 
10 Amer. Jour. Med. Sci., 1894, vol. cvii, p. 502. 


Within an hour after a meal there is a slight, 
DIGESTION, transitory increase in the number of erythro- 
FOOD. cytes, amounting on the average to a gain of one- 
quarter of a million cells to the This acme 
is soon followed by a gradual decline, corresponding to the period 
of digestion, the normal i standard being regained within two or 
three hours after the preliminary rise began. Corresponding fluc- 
tuations in the hemoglobin and in the density of the blood also 

Oliver * states that these variations are not affected by the taking 
of liquids with meals, for he has noticed that they were quite 
as pronounced when water was withheld. The same observer has 
shown 2 that these fluctuations correspond accurately to what he 
terms the "digestive lymph wave," or the to-and-fro intermediary 
circulation between the capillaries and the lymph spaces excited by 
the ingestion of food. Immediately after a meal this wave begins 
to rise, and coincidentally with this leakage of fluid from the blood 
into the tissues the hemoglobin and erythrocyte values begin to 
rise, as the blood thus becomes more and more concentrated. The 
acme is reached within an hour, and at this time the blood count 
and the percentage of lymph in the tissues are at their maximum. 
The wave then begins to decline slowly, and as the effused fluid 
gradually is restored to the blood vessels, the hemoglobin and cor- 
puscle figures correspondingly decline, until they reach normal, 
after the lapse of two or three hours after their initial rise. The 
following table, from von Limbeck, 3 illustrates the variations in the 
red and white cells caused by taking food. 


11 15 A. M. 5>553> 7,666 98 per cent. 

12 M. Dinner of meat and farinaceous food. 

15 p. M. 5,320,000 6,166 

15 P. M. 5,480,000 8,500 

15 P. M. 4,733,000 12,000 

15 P. M. 4,872,000 14,000 89 per cent. 

15 P. M. 4,720,000 10,830 

Hayem 4 believes that meat eaters average a higher percentage of 
erythrocytes than vegetarians, on account of the more nitrogenous 
character of their food, and that a diet of fats and albuminoids is 
most favorable for the increase of the cellular elements of the 

1 Loc. cit. * Lancet, 1903, vol. ii, p. 940. 

3 Loc. cit. 4 Loc. cit. 


The habitual polycythemia of individuals liv- 
HIGH ing in high altitudes is an interesting and inade- 

ALTITUDES. quately explained fact in hematology. Viault, 1 
Wolff and Koeppe, 2 Egger, 3 and other observers 
have shown the invariable occurrence of this polycythemia both 
in inhabitants of elevated districts and in the occasional visitor. 
In the case of the latter, as the individual ascends from the sea- 
level to the mountainous district, a rapid increase in corpuscles 
and in hemoglobin develops, this increase bearing a certain rela- 
tion to the height ascended, and becoming apparent usually 
within twenty-four or forty-eight hours after his arrival in the 
highland. Campbell and Hoagland 4 computed the increase at 
the rate of 50,0x20 cells to the per 1000 feet of elevation. 
Ten hours' stay in a balloon, at a height of over 15,000 feet, pro- 
duced no morphological changes in the blood of two aeronauts, 
Schroetter and Zuntz, 5 who undertook this experiment. Viault 
counted 8,000,000 erythrocytes to the in the residents of the 
Cordilleras, at an elevation of 14,274 feet above the sea-level; 
Cazeaux 6 found 7,100,000, at a height of 5904 feet. Egger 
counted 7,000,000 at Arosa, at a height of 6100 feet; and Wolff 
and Koeppe found an average of 5,970,000 in dwellers at Reibolds- 
griin,,at a height of 2257 feet. Oliver 7 relates the interesting ex- 
perience of finding in his own blood, during a stay at Davos Platz, 
at an elevation of 5200 feet, an increase of corpuscles within 
twenty-four hours after his arrival, the maximum count, 5,550,000, 
being attained within seven days, and the number declining within 
five days after his return to London. 

The following table, by Marie, 8 illustrates the fact that the 
higher the altitude the higher is the count of erythrocytes : 



Christiania o 4,974,000 Laache. 

Gottingen 148 meters. 5,225,000 Schafer. 

Tubingen 314 

Zurich 414 

Auerbach 425 

Reiboldsgriin 700 

Arosa 1,800 

The Cordilleras. . . .4,392 

5,322,000 Reinert. 

5,752,000 Stierlin. 

5,748,000 Koeppe. 

5,900,000 Koeppe. 

7,000,000 Egger. 

8,000,000 Viault. 

1 Compt. rend. Soc. biol., Paris, 1890, vol. iii, p. 917. 

2 Munch, med. Wochenschr., 1893, vol. xl, p. 904. 

3 XII. Cong. f. inn. Med., Wiesbaden, 1893. 

4 Amer. Jour. Med. Sci., 1901, vol. cxxii, p. 654. 

5 Pfluger's Arch., 1902, vol. xcii, p. 615. 

* Cited by Tissier, Cohen's "System of Physiological Therapeutics," Philadel- 
phia, 1903, vol. x, p. 182. 

7 Loc. cii. 8 "Lecons de Clinique Medicale," Paris, 1896, p. 237. 


Foa 1 reports that animals taken to a height of 10,000 feet 
develop polycythemia within eight hours after their arrival, and 
that the increase disappears within thirty-six hours after their 
return to a normal level. On the contrary, Armand-Delille and 
Meyer 2 were unable to detect any definite changes either in the 
peripheral and heart blood or in the hematopoietic organs of ani- 
mals kept for from two to seven weeks at an altitude of 6000 feet.' 

The hemoglobin changes which accompany these cellular 
alterations are never so marked as the latter, both the rise and 
the fall being less rapid ; consequently it is common to find a low 
color index at first, whereas later, inasmuch as the rapidity of 
the cellular loss is greater than the fall in hemoglobin, a high 
color index is likely to persist for some time after return to the low- 
land. Curry 3 could determine no consistent changes in hemo- 
globin, the volume of the cells, or the specific gravity of the whole 
blood at an elevation of 6000 feet. The blood changes, as a rule, are 
more conspicuous in normal than in anemic persons. In phthisics 
living in high altitudes, Meissner and Schroder 4 found that the 
hemoglobin value rose and fell in relation to the patient's im- 
provement and decline in health. Kemp 5 calls attention to the 
fact that the morning erythrocyte count is always higher by from 
500,000 to 1,000,000 than the evening estimate. He also noted 
that a large increase in the number of blood plaques developed 
as the patient ascended to a high elevation. 

Concentration of the blood doubtless explains the polycythe- 
mia of high altitudes, this change being due largely to the great 
loss of body fluids (Grawitz), and partly to the increased arterial 
tension (Oliver) arising from a rarefied atmosphere. 'Koeppe's 
ingenious theory that the process mirrors an actual manufacture 
of new cells is scarcely tenable, for although this observer has 
found numerous microcytes and poikilocytes coincidentally with 
the appearance of the polycythemia, normoblasts were not de- 
tected, as an evidence of rapid hemogenesis, nor did such signs 
of excessive blood destruction as icterus and hemoglobirruria de- 
velop, as the increased count rapidly declined on the individual's 
descent to a lower level. 

It has been recently urged that in high elevations the effect 
upon the hemocytometer of atmospheric pressure and tempera- 
ture may be the real secret of the cellular increase, but how such 
influences act, if, indeed, they are active, is unknown. 

1 Brit. Med. Jour., 1904, vol. i, p. 1097. 

1 Sem. me'd., 1903, vol. xxiii, p. 371. s Amer. Med., 1902, vol. iv, p. 367. 

4 Cohen's "System of Physiological Therapeutics," Phila., 1903, vol. x, p. 191. 
4 Med. News, 1904, vol. Ixxxiv, p. 383; also Johns Hopkins Hosp. Bull., 1904, 
vol. xv, p. 177. 


A sea climate apparently causes a moderate 

CLIMATE. increase in the number and hemoglobin value of 

the erythrocytes. Marestang's studies 1 of the 

blood of several sailors during a sea voyage are the only available 

proof of this change. 

A tropical climate of itself probably does not affect the blood, 
although it offers fruitful factors of anemia in predisposing to such 
infections as the malarial fevers, ankylostomiasis, and filariasis. 
So-called tropical anemia is more often apparent than actual. 


True ameboid movements of the erythrocytes 
AMEBOID are sometimes observed, as the result of the effect 
MOTILITY. of globulicidal agents, or of some pathological 
state of the blood, such as a severe, high-grade 
anemia. The inherent elastic and contractile qualities shown by 
the cells, by virtue of which they undergo various changes in 
shape while floating about in the plasma, must not be confounded 
with the actual ameboid motility which they exhibit in disease. 
The oscillatory dancing movements of bits of fragmental cor- 
puscles, and the characteristic motility of the intracellular hyaline 
malarial parasite, also must be distinguished from the progres- 
sive, deliberate characteristics of the truly ameboid red blood cell. 
Within the body the hemoglobin and other 
ALTERATIONS constituents of the erythrocytes are preserved 
IN intact within the corpuscular stroma by the com- 

ISOTONICITY. position of the blood plasma, which is such that 
a perfect osmotic balance is constantly main- 
tained. Outside of the body, if this relationship is disturbed by the 
addition of distilled water to a specimen of blood, thus lowering 
the concentration of the plasma, the corpuscles swell, and a rapid 
discharge of hemoglobin into the surrounding tissue ensues, but the 
addition of saline solutions of a definite strength prevent such a 
change. Solutions of salts of just sufficient concentration to pre- 
serve the corpuscles and to prevent removal of their elements 
are known as isotonic, solutions of greater strength are termed 
hypertonic, and those of lesser strength, hypotonic. In normal 
blood it has been determined that the isotonicity of the erythro- 
cyte usually ranges from about 0.48 to 0.46 per cent. NaCl; that 
is, salt solutions of this concentration are just sufficient to prevent 

1 Cited by von Limbeck, loc. cit. 


the discharge of hemoglobin by the cell, although it may swell by 
taking up water. A 0.9 per cent, or "normal" salt solution not 
only preserves the hemoglobin within the cell, but also prevents 
alterations in its size and contour. Hamburger 1 and others have 
shown that alterations in isotonicity depend not only upon changes 
in the plasma, but also upon the constitution of the erythrocytes 
themselves. Fluctuations in the amount of the cells' diffusible 
albumins, chlorids, and phosphates are attended by corresponding 
osmotic variations. 

Owing to the conflicting results obtained by different in- 
vestigators, the isotonicity of the erythrocytes in different dis- 
eases is of little clinical value. Stengel 2 found the percentage 
0.52 and 0.6 in two cases of pernicious anemia, yet in other 
cases the figures were normal; in other diseases marked by ane- 
mia, such as carcinoma, hepatic cirrhosis, renal disease, and 
tuberculosis, he found that the variations were trivial. Von Lim- 
beck 3 found that the isotonicity was usually, but not invariably, 
increased in high-grade secondary anemias, in leukemia, and in 
many of the acute injections, while it was decreased in chlorosis 
and in calarrhal icterus. A decidedly increased isotonicity was 
found by Vicarelli 4 in pregnant and nursing women. As a gen- 
eral rule, it is believed that degenerative changes in the erythro- 
cytes, whatever their nature, predispose to dissociation of hemo- 
globin from the stroma, and that in such instances the isotonic 
percentages are higher than normal. 

In the fresh specimen of blood, exaggeration 

HYPERVIS- of the adhesive properties of the erythrocytes 
COSITY. may be observed in a number of conditions, but 
up to the present time no special clinical signifi- 
cance has been assigned to the phenomenon. It occurs to some 
extent in most inflammatory diseases, and, according. to Hayem, 5 
is often seen in the anemias associated with marked cachexia. 
Striking examples of hyperviscosity result when .the erythrocytes 
are subjected to the action of various poisons, notably snake-venom, 
and of heterogeneous pathological blood serum. (See p. 128.) 
From the effect of such influences the erythrocytes, instead of 
forming normal rouleaux, tend to adhere in large, irregular 
masses in which the distinctive characteristics of the cells are 
masked or lost. The individual cells, unattached to such a mass, 
may exhibit every possible variety of distortion, losing their typi- 
cal biconcavity and regular disc-like appearance, and "becoming 

1 Arch. f. Anat. u. Physiol., 1886, p. 476; ibid., 1887, p. 31. 

2 Loc. cit. 3 Loc. cit. 
4 Cited by von Limbeck, he. cit 8 Loc. cit. 





converted into elongated, misshapen bodies. It frequently happens 
that the cell is provided with one or more long, delicate processes 
several times the length of its diameter, this being due to the ad- 
hesion of a bit of the stroma to the cover-glass while preparing the 
specimen; in the spread film it will be noted that these processes 
all point in the same direction. 

Changes in the shape and size of the eryth- 
rocytes are common in all anemias which reach 
a severe grade, the degree of such deformities 
corresponding closely to the intensity of the 
anemic process. The diameter of the cells may 
be more or less uniformly increased or decreased, 
and such pronounced alterations in their shape may occur that 

many of them bear but 
slight resemblance to the 
typical discs of normal blood 
(Plate I; also Fig. 46). Ab- 
normal inequality in the 
size of the erythrocytes is 
expressed by the term aniso- 

When the corpuscle be- 
comes greatly enlarged in 
diameter it is known as a 
megalocyte or macrocyte, the 
presence of large numbers 
of such cells being known 
as megalocytosis or macrocy- 
tosis. The diameter of a 
megalocyte generally varies 
from 9 to 12 ,, but some- 
times much larger forms are 

seen, measuring as much as 20 p. They are present in the severer 
anemias, especially in the pernicious form, in which they constantly 
occur in large numbers. The megalocyte found in this disease is 
usually characterized by an excess of hemoglobin, while in the 
secondary anemias such cells are generally deficient in their hemo- 
globin content. 

The smaller forms, the microcyles, illustrate the extreme de- 
crease in size of the red cell under pathological conditions. The 
microcyte is an extremely small globular body, measuring from 
about 3 to 5 /J. in diameter. It is found in all the varieties of ane- 
mia, but is most commonly associated with chlorosis and with 
the moderately developed second forms. An abundance of 
microcytes in the blood is known as microcytosis. 

Illustrating various grades of cell deformity 
associated with severe anemia. The large nu- 
cleated erythrocyte is a typical megaloblast. (Ehr- 
lich's triacid stain.) 


Eichhorsfs corpuscles are deeply colored, highly refractive mi- 
crocytes, about 3 f* in diameter, and usually of regularly spherical 
shape. They were once regarded as pathognomonic of pernicious 
anemia, but are now considered diagnostic of no especial condi- 
tion, being frequently found in severe anemias of any type, and 
often being absent in pernicious anemia. 

It seems reasonable to infer that deformities in the size of the 
erythrocyte are referable chiefly to two different factors: to faulty 
hemogenesis and to degenerative changes of the corpuscle which 
lead to alterations in its histological structure. Megalocytes, for 
example, may in some instances represent an actual giantism of 
the cell, bred in the marrow from correspondingly large-sized 
nucleated antecedents; in other instances (of which those exceed- 
ingly pale, "washed-out" forms are examples) their abnormal 
size may be attributed to hydropic enlargement, resulting from 
their imbibition of fluids from the surrounding plasma. Micro- 
cytes may enter the circulating blood as such, or, as is frequently 
the case, they may be the products of corpuscular budding and 

In severe forms of anemia, characterized by excessive cellular 
loss, there appears also to be a tendency toward a compensatory 
hypertrophy of many of the erythrocytes, in order thus to in- 
crease the oxygen-carrying capacity of the blood, which, were it 
not for these numerous megalocytes, might in some instances be 
too limited to sustain life. 

Poikilocytes are erythrocytes deformed in shape as the result 
of some pathological condition of the blood. Poikilocytosis, the 
name by which this condition is designated, is .akin to crenation 
in so far as in both conditions the cells may be similarly distorted 
and misshapen. But it is unlike crenation for the reason that 
poikilocytosis is a pathological condition, and demonstrable the 
moment the blood is withdrawn from the body; while crenation 
is a physiological phenomenon depending upon external influences 
for its production, and never occurring until the blood has re- 
mained exposed to the air for some time. Poikilocytes may be 
of large or small size, the varieties of deformities being infinite, 
and the degree marked or slight in relation to the nature of the 
blood disease. Some of the cells may resemble the shape of a 
gourd or a horseshoe ; others may be drawn out at both ends until 
they form a spindle-shaped or oval body, while others appear 
sharply beaked at one or more points, or shaped like a dagger or 
the blade of a tomahawk. Occasionally very minute, rapidly 
oscillating, rod-shaped forms are .seen, morphologically not unlike 
large, unstained bacilli pseudo-bacilli of Hayem. These rod- 


shaped forms are products of corpuscular fragmentation, and 
indicate lowered vitality and feeble powers of resistance to the 
pathological influences affecting the cells. 

Poikilocytosis is not characteristic of any single disease of the 
blood, but it is generally most marked in the grave forms of 
primary anemia, such as leukemia and pernicious anemia. Oval- 
shaped red cells are considered by Cabot l as particularly abundant 
in the latter disease. 

The conditions of deformity affecting the shape and size of the 
erythrocytes are nearly always associated. As a general rule, it 
may be stated that in the milder types of anemia small-sized, 
slightly deformed poikilocytes and microcytes are most common; 
and that in the severe forms, large-sized, conspicuously distorted 

poikilocytes and megalocytes 

Loss of 

ENDOGLOBULAR color by the 
DEGENERATION, erythrocytes, 
which pro- 
gresses hand in hand with 
alterations in their size and 
shape and other structural 
changes, is regarded as a de- 
generative process of purely 
endoglobular nature. It is 
observed in the fresh specimen 

FIG. 47.-DEGENERAT.vE CHANGES m THE ERY- of blood in many severe ane- 
THROCYTES. (FRESH BLOOD FILM.) m j c conditions, especially in 

the anemias associated with 

infectious diseases, such as variola, typhus fever, and grave septi- 
cemia and pyemia (Fig. 47). 

The decoloration may commence in one or more spots, or it 
may equally involve the whole surface of the corpuscle, beginning 
at its center and spreading progressively toward its periphery. 
Clear, hyaline areas of oval, round, or elongated shape appear 
within the stroma, in some instances sharply contrasting with the 
relatively dark color of the hemoglobin, but in other instances 
imperceptibly blending with the tint of the surrounding cell body. 
The active motility of these decolorized spots must be carefully 
distinguished from the ameboid movements of the young malarial 
parasite. Complete decoloration transforms the cell into a mere 
colorless shell or "phantom," which would be practically invisible 

1<( A Guide to the Clinical Examination of the Blood," 5th ed., New 
York, 1904, p. 131. 


were it not for its faintly colored periphery. Such cells are known 
as Ponfick's shadow corpuscles or as Hayeni's achromacytes. 

Maragliano and Castellino 1 have minutely described this proc- 
ess of decoloration, along with certain other alterations in the 
structure of the erythrocyte, which they have termed endoglobu- 
lar necrosis. This process first becomes apparent by a visible 
enlargement of the central concavity of the corpuscle, together 
with a simultaneous fading away of the hemoglobin in this situ- 
ation. This central area of pallor gradually spreads toward the 
periphery of the cell, until finally the latter alone shows evidence 
of containing coloring matter. Such a corpuscle, when examined 
on cross-section, appears to be shaped like the figure 8. Frag- 
mentation of this delicate rim of coloring matter may occur, in 
the event of which numerous independent, rod-like bits of stroma 
are formed. The decolorized area is not always symmetrical, so 
that frequently various strikingly bizarre designs, widely differing 
in shape and appearance, may be observed, It has been deter- 
mined that in the dried blood film these areas of decoloration show 
a decided affinity for basic stains, such as methylene-blue and 

Total cellular necrosis, also described by the 
TOTAL authors mentioned above, represents a phase 

NECROSIS. of structural degeneration in the erythrocyte 
of more advanced development than the endo- 
globular changes. This process begins with the development of 
several small elevations or corrugations in the stroma of the cor- 
puscle, which gradually multiply, increase in size, and change in 
shape until the larger portion of the cell's surface is thus de- 
formed. Ameboid movements are seen to begin, as if the entire 
cell as a whole were involved, the final stage of the process re- 
sulting in the formation of a poikilocyte, from which body points 
and small fragments are observed to break off and to float free 
in the plasma. Decoloration, starting usually from a single point 
and in time affecting the whole stroma, also accompanies this 
necrotic alteration. On cross-section the cell appears as an 
elongated, thin rod with rounded poles. This fragmentation of the 
cells is spoken of as schistocytosis. The erythrocytes are much 
more resistant than the leucocytes, which succumb much more 
readily to necrobiotic influences. 

Endoglobular degeneration and total necrosis of the erythro- 
cytes may be observed both in normal and in pathological blood. 
In normal blood they occur as the result of prolonged contact 
with the air, the endoglobular phase becoming first apparent 

1 XI. Cong. f. inn. Med., Leipsic, 1892. 


within from thirty to seventy minutes, and the total necrosis in 
from three to four hours, after the preparation of the specimen. 
In pathological blood the changes are thought to be due chiefly 
to increased globulicidal properties of the plasma, whereby intra- 
vascular necrosis is excited, and partly to decreased resistance 
of the erythrocytes, in consequence of which their degeneration is 
abnormally hastened by contact with normal plasma and by 
exposure to extraneous influences. In disease it follows that 
they are demonstrable immediately or very shortly after the blood 
has been withdrawn, and that the development of the changes 
occurs with much greater rapidity than in normal blood. The 
endoglobular changes are regarded as a more favorable prog- 
nostic sign than the total necrosis, being usually associated with 
anemias of less severe character than those in which the latter 
process prevails. 

The normal erythrocyte, when fixed and 

ATYPICAL stained with anilin dyes, according to one of 
STAINING RE- the methods described in another section, pos- 
ACTION. sesses a strong affinity for a single, acid stain; 
it is therefore termed monochromatophilic. When 
solutions are used containing both acid and basic dyes, such 
as eosin and methylene-blue or eosin and hematoxylin, the nor- 
mal erythrocyte is always stained by the eosin; and with Ehr- 
lich's triple mixture, which is so formulated that acid, basic, or 
so-called neutral principle may be selected by the elements sub- 
jected to its action, according to their affinities, the erythrocyte 
invariably is colored by the orange G of the mixture (Plate I). 

In certain morbid conditions some, of the corpuscles lose their 
affinity for the acid stain, and with mixtures of both acid and 
basic dyes are stained atypically by either or both elements. 
Such corpuscles are said to be polychromatophilic. Thus, when 
stained with an eosin and methylene-blue mixture, they are 
tinged a dirty grayish-purple or violet, instead of the rose color 
of eosin; and with the triple mixture they may be stained pur- 
ple, reddish-brown, or pale yellowish-pink, flecked here and 
there with shadings of a darker red (Plate I). 

In polychromatophilic corpuscles the staining is likely to be 
very unevenly shaded, often being quite dark in spots, especially 
around the periphery of the cell and the margin of the nucleus, 
if the cell be nucleated. These color changes affect not only 
the protoplasm, but the nucleus as well, and are strongly em- 
phasized in megaloblasts, the nuclei of which may show every sort 
of color combination. The more deficient the corpuscle in hemo- 
globin, the more decided its polychromatophilic tendency; and 


the more strikingly the latter is developed, the more intense the 
cell's affinity toward the basic element of the stain. 

Polychromatophilia may occur in severe forms of anemia due 
to any cause, and it is especially noted in two of the primary 
varieties pernicious anemia and myelogenous leukemia in 
both of which conditions the process is a prominent character- 
istic of the blood picture. Corpuscles of Poggi, or erythrocytes 
which stain with basic dyes in the fresh, unfixed specimen, are also 
found in various anemias. They probably represent immature 
elements whose presence in the circulating blood reflects stimulated 

Nucleated erythrocytes, or erythroblasts, are 
NUCLEATION. found in the blood of the adult only during the 
existence of pathological conditions, but occur 
in large numbers in the blood of the fetus, and occasionally in 
the infant during the first few days of life. Being invisible in the 
fresh blood, they must be studied in the dried, stained specimen. 
In such preparations the finer structure of their nucleus, which 
bears a special affinity for the basic anilin dyes, may be beauti- 
fully illustrated by the use of solutions containing methylene- 
blue, methyl-green, and hematoxylin. 

According to their size and nuclear characteristics the ery- 
throblasts are designated as normoblasts, megaloblasts, and mi- 
croblasts. Certain intermediate forms are also common, some- 
times termed mesoblasts, such cells being atypical, and sharing 
characteristics of both the normoblast and the megaloblast. . 

Normoblasts (Plate I). The normoblast is a nucleated ery- 
throcyte of about the general size and shape of the normal erythro- 
cyte. In the typical mature cell the nucleus is round or ovoid in 
shape, very deeply stained, and situated rather toward the periph- 
ery of the cell than in the exact center, its diameter approximating 
more than one-half that of the corpuscle which it occupies. In the 
normoblast of an earlier developmental stage the nucleus, is rela- 
tively larger and is composed of delicate, faintly basic chromatin 
hall-marks of histological youth. In some of the typical cells 
the nucleus appears to have become partly or completely extruded 
from the protoplasm, lying either somewhat over the periphery of 
the cell or, being completely detached from it, free in the plasma 
(Fig. 49, III). The nucleus may be single, or partly divided by 
constricting bands of chromatin into a figure like a dumb-bell or 
a clover-leaf, or completely divided into several small, round 
sections. More rarely, karyokinesis may be observed, the diaster 
and early convolution stages, with an intact but plainly con- 
stricted cell body, being the phases ordinarily found. (Plate I, 


Fig. 3; also Fig. 49, II.) In carefully stained films it will be 
noted that the nuclear framework of the typical normoblast 
consists of a rather sharply defined network of chromatin hav- 
ing relatively wide intervening open spaces, so that the general 
appearance of the nucleus is not unlike that of a coarse net. 

The protoplasm of this cell is usually of regular outline along 
the periphery, stains somewhat more intensely than that of the 
normal erythrocyte, and may show distinct evidences of poly- 
chromatophilia, this characteristic being especially marked in forms 
with dividing nuclei. 

The normoblast is regarded as the immediate antecedent of the 
normal erythrocyte or normocyte, into which it becomes trans- 
formed by the loss of its nuclear structure. The exact manner 
in which the nucleus is disposed of has long been a bone of con- 
tention among histologists, and even at the present time views on 
this question should be held but tentatively, notwithstanding many 
exhaustive investigations, especially those of the German school. 
According to the views of Rindfleisch, 1 it is lost by extrusion 
from the cell body, which thus becomes a normal erythrocyte, 
while the free nucleus, to which a small fringe of protoplasm still 
remains adherent, collects from the plasma material by virtue of 
which it ultimately develops into a new erythroblast. Ehrlich 2 
believes that in blood rich in normoblasts a series of connected 
pictures may be observed, showing that the normoblast becomes 
transformed into the erythrocyte by the extrusion or emigration 
of the nucleus. The later investigations of Neumann and Kol- 
liker, 3 however, tend to prove that the nucleus is disposed of by 
its destruction and absorption within the cell, and that its ap- 
parent extrusion from the stroma is simply the result of mechan- 
ical influences. Pappenheim and Israel 4 also believe that the 
normoblast's nucleus disappears by decay and solution within the 
body of the corpuscle, and that the apparently extruded nuclei 
are to be taken as an evidence of plasmolysis, or a solution of 
the protoplasm of the cells once containing nuclei. To attempt 
a reconciliation of these diametrically opposed views is a task for 
future workers to undertake. Meanwhile, the general trend of 
opinion inclines toward the theory of nuclear solution within 
the corpuscle, and regards the so-called free nuclei of the normo- 
blasts simply as artefacts (Fig. 49, I and III). 

Normoblasts exist in the red bone marrow of the normal indi- 

1 Arch. f. mik. Anat., 1880, vol. xvii, p. i. 

2 Loc. cit. 3 Zeitschr. f. klin. Med., 1881, vol. iii, p. 411. 
4 Virchow's Arch., 1896, vol. cxlv, p. 587; also Pappenheim, Inaug. Dis- 
sert., Berlin, 1895. 


vidual, but are found in the circulating blood only when the 
marrow, in consequence of pressing demands made upon it for 
the rapid manufacture of new erythrocytes, becomes unable to 
furnish an adequate supply of perfectly developed cells, so that 
some of these immature, nucleated forms prematurely leave their 
birthplace in the marrow, and pass into the blood stream in com- 
pany with large numbers of mature, non-nucleated discs. Normo- 
blasts are associated with lesions in which active hemogenesis 
of the normal type is stimulated, being the prevailing type of 
erythroblast in the anemias resulting from hemorrhage and in 
other severe anemias of a secondary type. They sometimes ap- 
pear in the blood in successive crops of large numbers during the 
course of certain severe anemias, this phenomenon having been 
termed by von Noorden 1 a blood crisis. Blood crises, which are 
of abrupt onset and of brief duration, lasting but a few hours, 
are usually the direct precursors of an increase in the erythro- 
cyte count and in the hemoglobin percentage, being therefore 
a favorable sign, indicating regeneration of the blood. They occur 
with especial frequency after loss of blood from hemorrhage and 
in chlorosis, and are not uncommon in long-standing cases of 
myelogenous leukemia and primary pernicious anemia, in which 
diseases periods of temporary improvement are likely to take place 
from time to time. 

Megaloblasts (Plate I; also Fig. 48). The typical megaloblast 
is much larger in size than the normoblast, and contains a single, 
large, pale-staining nucleus which occupies the greater part of the 
cell body. Both cell and nucleus are round or ovoid in sljape, the 
diameter of the former being from about u to 20 fi, and that of the 
latter from 6 to 10 /A The greatest extremes of these' measurements 
apply to those forms which are seen with relative infrequency, for 
the megaloblast most commonly observed does not usually measure 
more than 12 // in diameter, with a nucleus of proportionate size. 
The nucleus, which may be situated either in or 'away from the 
center of the cell, is composed of a chromatin network having 
relatively small intervening open spaces, so that the nuclear 
structure is decidedly more delicate and less well-defined than that 
of the normoblast. With the triacid solution and with the Roman- 
owsky stain it is tinted pale green or blue, or it may show every sort 
of irregular tinctorial reaction to the anilin dyes, certain portions 
being deeply stained, while other parts are but faintly cojored; the 
undertone of green or blue is frequently stippled with fine dots of 
purple or of brilliant crimson, especially about the periphery; or 
it may be mottled and splotched here and there with areas of purple 

1 Charit-Annalen, 1891, vol. xvi, p. 217. 


or of dark blue. The nucleus, in the triple-stained film, usually 
is sharply differentiated from the body of the cell by a distinct 
white margin which encircles it and is thrown out in bold relief by 
the deep staining of the nuclear and cell bodies on either side. 
Occasionally a megaloblast shows a coarse, small nucleus of very 
basic affinity, resembling that of the normoblast characteristics of 
nuclear senility. 

The protoplasm of the megaloblast often seems swollen and 
enlarged, and appears to contain areas of depression and elevation 
at different points; it is sometimes quite round or oval in contour, 
and sometimes more or less deformed. It is usually polychro- 
matophilic, and, like the nucleus, may show the greatest variety 

of color combinations. Some 
cells stain, with the triacid mix- 
ture, a dull brownish-yellow 
color with deeper shadings of 
a burnt-sienna tint in the 
neighborhood of the nucleus 
and of the periphery; others 
have an undertone of crimson, 
as if the stain contained an 
excess of fuchsin, and are 
strea'ked and dotted with yel- 
low and tan-colored patches; 
still others stain a diffuse pur- 
ple, blending in spots into a 
light pink. With Wright's 
stain the color varies from 
greenish-blue to purple to dull 
yellow. Mitotic megaloblasts, 
with figures similar to those 
exhibited by normoblasts thus 
dividing, are met with occasionally in anemias of great severity. 
(Plate I, Fig. 3; also Fig. 49, II.) 

The megaloblast is an element of the bone marrow of the young 
fetus, and is totally foreign both to the marrow and to the blood of 
the normal adult. According to the views -of Ehrlich, it repre- 
sents the immediate antecedent of the megalocyte, into which it 
develops by the absorption of its nucleus. Apparent extrusion 
of megaloblastic nuclei is never observed. Megaloblasts are 
found in the circulating blood only under conditions in which the 
blood-making organs have reverted more or less to the fetal type, 
so that their presence in the circulation is considered to indicate that 
a sluggish hemogenesis of embryonal character exists. The 

Common types of megaloblasts, showing va- 
riations in size and shape and peculiarities of the 
nuclear structure. (Ehrlich's triacid stain.) 


significance of megaloblasts, therefore, is diametrically opposed 
to that of normoblasts, for, while the latter are regarded as an 
expression of blood regeneration and are considered to be of 
favorable prognostic significance, the former must be looked on 
as an evidence of degeneration of the hematopoietic organs, and, 
consequently, are of grave prognosis. 

In the following table the principal points of distinction between 
the typical normoblast and the megaloblast are emphasized: 




7.5 to 10 n. 

II tO 20 /I. 


Sharply defined. 
Intensely basic. ' 
Coarsely meshed. 
Occupies about one-half of 
cell body. 

Dully defined. 
Feebly basic. 
Delicately meshed. 
Occupies greater part of cell body. 


Sometimes very scanty 
and of ragged outline. 
Occasionally polychro- 

Frequently appears swollen; out- 
line fairly regular, but surface 
undulating in many cells. 
Striking tendency toward poly- 
chromatophilia . 


Typical of active, adult 

Typical of sluggish, embryonal 


Prevailing type of ery- 
throblast in anemias 
with active blood re- 

Prevailing type of erythroblast in 
anemias with megaloblastic de- 
generation of the bone marrow. 

Megaloblasts are found in the blood, almost in variably in as- 
sociation with normoblasts, in various anemias of marked severity, 
but in only three conditions, viz., primary pernicious anemia, 
certain cases of anemia due to Bothriocephalus latus infection, 
and nitrobenzol poisoning, have these cells been found to con- 
stitute the prevailing type of erythroblast. 

In pernicious anemia the prevalence of megaloblasts is. gener- 
ally admitted to be a sign that in this disease the bone marrow, 
in consequence of its reversion to a fetal type, throws into the 
blood stream large numbers of these blood cells of embryonal 
character, these degenerative changes, the presence of megalo- 
blasts, overshadowing the regenerative changes, or the presence 
of normoblasts. In bothriocephalus anemia, in which also the 
megaloblasts may outnumber the normoblasts, it is believed that 
the toxins produced by the parasite cause changes in the hema- 
topoietic organs precisely similar to those found in pernicious 
anemia. Rosenquist's 1 elaborate studies show that the excessive 

1 Berlin, klin. Wochenschr., 1901, vol. xxxviii, p. 666. 


albumin disintegration caused by the bothriocephalus toxin is 
essentially like that occurring in typical pernicious anemia. In 
a single case of nitrobenzol poisoning, reported by Ehrlich and 
Lindenthal, 1 large numbers of erythroblasts were noted; normo- 
blasts predominated at first, but in the later stages of the intoxica- 
tion they were outnumbered by megaloblasts. In other grave 
anemias, notably in leukemia, the regenerative signs appear to 
be more active than the degenerative, for, while in these condi- 
tions megaloblasts are frequently found, they are never so numer- 
ous as the normoblasts. 

Microblasts. The microblast, which is the rarest form of nu- 
cleated erythrocyte, is a cell usually not larger than 5 or 6 p in 
diameter, and often of smaller size. It consists of a deeply stained, 
round nucleus like that of the normoblast, encircled by a frag- 
ment of ragged protoplasm of a dull brownish-yellow tint, in films 
stained with the triacid solution. Wright's stain frequently 
colors this stroma blue. Microblasts are thought to be simply 
forms of the normoblast in a more or less advanced stage of 
protoplasm degeneration, this process accounting for the char- 
acteristic scantiness and frayed-out appearance of their cell body. 
Their clinical significance, naturally, is identical with that of the 

From what has been stated, it may be concluded that nor- 
moblasts and megaloblasts constitute two distinct classes of 
nucleated erythrocytes, each evidencing a separate type of blood- 
formation, and each carrying a different clinical meaning. Nor- 
moblasts, being an adult type of cell, have sharply defined, dense, 
deeply stained nuclei; megaloblasts, being an embryonal type of 
cell, have poorly defined, delicate, feebly stained nuclei. 

Atypical Erythroblasts. In some of the severer anemias, 
notably in myelogenous leukemia and in pernicious anemia, 
various atypical erythroblasts are frequently found, corresponding 
partly to one and partly to the other of the first two species of cells 
described above. These so-called " mesoblasts," which may be re- 
garded either as normoblasts with immature nuclei or as megalo- 
blasts with mature nuclei, are in some instances almost as numerous 
as the typical forms of erythroblasts. It is sometimes impossible 
accurately to determine to which type such cells belong, but 
usually they may be classified by taking size as a criterion for 
differentiation. Those approximating the normocyte in size may 
safely be classed as normoblasts; those of larger size, as megalo- 
blasts regardless of their nuclear peculiarities. The two fol- 

1 Zeitschr. f. klin. Med., 1896, vol. xxx, p. 427. 


lowing commoner forms of atypical erythroblasts may be rec- 
ognized : 

i. Corpuscles about 8 or 10 // in diameter, containing a rela- 
tively large, round or ovoid nucleus, composed of a finely meshed 
chromatin framework. The nucleus is pale, and is often filled with 



/, Megaloblasts and normoblasts showing nuclear solution; the two cells (a) show, early, and 
the three (o) late, stages of karyolysis. Five of the six cells (c) contain nuclear remains consisting 
of both coarse and delicate chromatin masses. Note the granular basophilia in the groups of cells 
at /, //, ///, and IV. II, Erythroblasts with multiple and dividing nuclei; the megaloblast (a) 
represents the wreath-shaped (monaster) stage of karyokinesis, the megaloblast (fr) the double star- 
shaped (diaster) stage, and the megaloblast (c) the convolution (daughter cell) stage. The other 
cells jllustrate the kinds of erythroblasts with convoluted and multiple nuclei ordinarily found 
in high-grade anemias. ///, Erythroblasts showing so-called nuclear extrusion. IV, Erythro- 
cytes containing ring bodies. V, Normal erythrocytes. (Wright's stain.) 

finely stippled areas of acid affinity. The cell body is usually of 
regular outline, and, as a rule, is decidedly polychromatophilic. 
Such cells may be regarded as immature forms of normoblasts, 
with which they may properly be classed in the differential count 
(Plate I; also Fig. 48). 


2. Corpuscles about 12 to 15 /* in diameter, having a small, 
coarsely meshed nucleus not exceeding 2 or 3 , in diameter, and, 
as a rule, situated eccentrically. The nucleus stains very basically, 
and may or may not be separated from the protoplasm by a color- 
less zone. The body of the cell is round or ovoid, and stains 
faintly. This form of cell appears to carry the same clinical 
significance as the megaloblast, of which it is probably a late 
developmental phase (Plate I; also Fig. 48). 

Ring Bodies. In severe anemia Wright's stain often brings out 
peculiar intra- and extra-cellular ring-shaped bodies whose outlines 
resemble those of the nuclear figures of erythroblasts (Fig. 49, 
IV). Cabot, 1 who described these bodies, suggests that they may 
represent nuclear remnants, or portions of the erythrocytes' nuclei 
which are especially resistant to whatever forces ultimately destroy 
the cells and their nuclei. Basic (blue) ring bodies within the 
erythrocytes were described and pictured, in 1901, by Strauss 
and Rohnstein, 2 who interpreted them as a variety of granular 
basophilia. The bodies have been found, in association with 
frankly nucleated erythrocytes, polychromatophiles, and basically 
stippled cells, in lead poisoning, in pernicious anemia, and in 
lymphatic leukemia. The writer has noted them in high-grade 
anemia secondary to sepsis. With Wright's solution the ring- 
shaped bodies stain red and, rarely, blue; they are variously 
shaped circles, rude rosettes, clover leaves, figures-of-eight, and 
forms with twisted threads and with netted structures. These 
designs, it will be noted, correspond accurately to the nuclear 
figures of various erythroblasts. 

In certain of the severe anemias, staining with 

GRANULAR methylene-blue shows a peculiar granular condi- 
BASOPHILIA. tion of the protoplasm in some of the erythro- 
cytes, attention first having been called to this 
fact by von Noorden, 3 who demonstrated the basophilic charac- 
ters of such granules, and described their occurrence in vari- 
ous pathological states. Many of the corpuscles thus affected 
are of the nucleated form, but non-nucleated cells may be simi- 
larly granulated; as a rule, such corpuscles are also strikingly 

The granules appear either as fine or as coarse, stippled areas, 
staining intensely with the basic stain, and distributed through 
the body of the cell either quite uniformly or in localized patches 

1 Jour. Med. Research, 1903, vol. iv, p. 15. 

"Die Blutzusammensetzung bei den verschiedenen Anamienen," Berlin, 
1901, p. 224. 

3 Charite-Annalen, 1892, vol. xvii, p. 202. 



at one or at several points. In some cells they are exceedingly 
fine and closely packed together, so that at first glance the whole 
protoplasm appears to be a homogeneous mass of purplish dis- 
coloration; in others the protoplasm is dotted here and there 
with coarse granules, not more than five or six being found 
in the whole cell; still others may contain both fine and coarse 
granules irregularly sprinkled over the surface (Plate I; also 
Figs. 49 and 50). 

The occurrence of somewhat similar granulations in the eryth- 
rocytes of the embryo has been noted by Engel, 1 Pappenheim, 2 
and others, who regard them as nuclear debris, the product of 
nuclear disintegration. Such an origin in embryonic blood is 


Erythrocytes showing various degrees of basophilia, with fine, coarse, spherical, ovoid, and spicu- 
late, granules. Note the basophilic normoblast. (Wright's stain.) 

probably physiological. In post-uterine life, however, this proc- 
ess is to be regarded as a sign of stroma degeneration, arising 
in all likelihood through the influence of various blood' poisons. 
In some instances the change precedes all other recognizable alter- 
ations in the blood, and appears as the first, and, indeed, some- 
times the only, distinct sign of anemia. 

Granular basophilia of the erythrocytes has been^ noted with 
more or less constancy in these conditions: pernicious anemia, 
leukemia, Hodgkin's disease, so-called tropical anemia, bothrio- 
cephalus anemia, malarial fever, sepsis, carcinoma, long-standing 
suppurati-ve lesions, and chronic lead poisoning. In chlorosis, if 

1 Verhandl. d. Vereins f. inn. Med. z. Berlin, 1898-99, vol. xviii, p. 216. 

2 Loc. cit. 


uncomplicated by symptoms of intestinal auto-intoxication, the 
erythrocytes do not exhibit this alteration; granule cells are also 
absent in syphilis, in acute infectious diseases, in chronic lesions 
oj the kidney and the liver, and in diabetes, according to Grawitz. 1 
Regarding the occurrence of this change in pernicious anemia, 
Ehrlich 2 believes that the number of granule cells in the blood 
bears a certain relation to the severity of the disease, stating that 
they decrease and often disappear during the periods of remission, 
reappearing as the other blood changes again become evident. 
On the other hand, Litten, 3 who asserts that he has found these 
basophilic granulations in one-tenth of all cases of anemia, has 
been unable to determine their clinical significance from either 
a diagnostic or a prognostic point of view. The studies of 
Grawitz and Hamel 4 show that granular degeneration of the 
erythrocytes occurs with great regularity in saturnism, both 
in obscure and in well-marked cases, and these authors attach 
considerable diagnostic value to this fact, concluding that the sign 
is important in the diagnosis of lead poisoning in patients in whom 
the intoxication is merely suspected, being evidenced by no other 
definite symptoms. Experimentally, basophilia has been produced 
by the administration of lead salts, tin chlorid, copper, pyrodin, 
atropin, toluylendiamin, and phenylhydrazin. A dose of any of 
the proprietary preparations of hemoglobin may also promptly 
excite the change, as may the ingestion of whole blood 
facts which lead Grawitz 5 to assume that blood in the gastro- 
intestinal canal elaborates toxic substances the absorption of which 
acts deleteriously upon the erythrocytes. This and other phases 
of basophilia have been reviewed at length by the writer elsewhere. 6 

Here may be mentioned certain areas of reddish stippling 
(Schiifiner's granules) demonstrated by polychrome methylene- 
blue in the erythrocytes of tertian malarial fever (q. v.). 

' Oligocythemia, or diminution in the number of 

OLIGOCY- erythrocytes below the normal standard, is pres- 

THEMIA. ent to a more or less marked degree in all forms 

of anemia, being associated, naturally, with an 

oligochromemia, or diminution in the percentage of hemoglobin, 

but not necessarily with an oligemia, or reduction in the volume 

of the blood mass. 

The loss of corpuscles may be slight or it may be marked, ac- 
cording to the nature of the anemia of which it is symptomatic. 

1 Amer. Jour. Med. Sci., 1900, vol. cxx, p. 277. 

2 Loc. cii. 3 Deutsch. med. Wochenschr., 1899, vol. xxv, p. 717. 

4 Deutsch. Arch. f. klin. Med., 1900, vol. Ixvii, p. 357. 

5 Deutsch. med. Wochenschr., 1901, vol. xxvii, p. 908. 
* Amer. Med., 1903, vol. v, p. 571. 


The most striking examples of oligocythemia are encountered 
after hemorrhages involving the loss of a large amount of blood 
and in pernicious anemia; while in chlorosis and in the majority 
of the secondary anemias the decrease is relatively less marked. 
The following summary of the averages of fifty consecutive counts 
each in cases of primary and secondary anemia illustrates the 
various degrees of cellular loss which ordinarily accompany these 
conditions : 


In pernicious anemia 1,152,470 

" leukemia 2,729,763 

" secondary anemia 3,642,900 

" chlorosis 4, 1 11,000 

It is impossible to designate the degree of oligocythemia which 
may exist without a fatal outcome, although a number of author- 
ities have attempted to set fixed limits beyond which reduction 
in the number of erythrocytes is supposed to cause death. The 
effects of a blood loss are so diverse in different individuals that 
all such arbitrary rules must, of necessity, prove practically 
valueless. It should be remembered that while in some persons 
a comparatively moderate decrease may prove fatal, in others a 
most astonishing loss is compatible with life. It may be stated 
in general terms that few individuals recover in whom a count of 
less than 500,0x20 erythrocytes to the is found, although 
occasional exceptions to this rule have been reported. 

Whether or not an actual, permanent polycy- 
POLYCYTHEMIA. themia, or an increase in the number of erythro- 
cytes above the normal standard, exists is still 
an unsettled question, but the majority of authorities maintain 
that such a condition is due merely to some physical change 
producing concentration of the blood, or unequal Distribution of 
the corpuscles, in favor of the peripheral blood vessels. .In health, 
it would not seem unreasonable to suppose that a moderate de- 
gree of polycythemia may be habitual in the strong, overdevel- 
oped adult, whose blood-making organs are possibly developed 
proportionately to the other parts of his system. In pathological 
conditions there is nothing tangible upon which to base the belief 
that an actual and permanent overproduction of the erythro- 
cytes ever takes place, the polycythemia associated with certain 
diseases being satisfactorily accounted for by coexisting physical 
conditions, in no way peculiar to the lesion in question. While 
it is true that in some conditions it is not always possible to ex- 
plain the increase by purely physical causes, still there is no posi- 


tive proof, in these instances, that the change is pathological. 
There seems, therefore, no evidence to warrant an arbitrary clas- 
sification of polycythemia into two divisions, actual and relative, 
as some authors have suggested. 

The cause of polycythemia, then, may be attributed to physio- 
logical factors such as concentration of the blood, increased blood 
pressure, peripheral stasis, increased viscidity of the erythrocytes, 
and their unequal distribution through the circulatory system. 

The polycythemia associated with various physiological and 
pathological conditions will be considered under their appropriate 
headings. Briefly, an increase of erythrocytes over the normal 
number is found in the following conditions: 

1. In the new-born. 

2. After taking food. 

3. In starvation. 

4. During residence in high altitudes. 

5. From the effect of cold and hot baths, muscular exercise, 
massage, and electricity. 

6. From the administration of lymphagogues, emetics, pur- 
gatives, and thyroid extract. 

7. During active blood regeneration. ' 

8. During reformation of an exudate after aspiration. 

9. After urinary crises, diaphoresis, emesis. 

10. In poisoning by illuminating gas and by phosphorus. 

11. In Asiatic cholera, dysentery, and diarrhea. 

12. In acute yellow atrophy of the liver and myxedema. 

13. In conditions of cyanosis and peripheral stasis, for example, 
uncompensated organic heart disease, emphysema, asphyxia, and 
Osier's disease. 

14. After the transfusion of blood. 


If a drop of fresh blood is examined microscopically imme- 
diately after it has been taken from the body,' a few pale, some- 
what spherical bodies, much smaller in size than the erythrocytes, 
may usually be observed. These bodies are known as the blood 
plaques or blood platelets. They are of homogeneous structure, 
either almost colorless or of a pale yellowish tint, spherical or 
irregularly ovoid in shape, and measure from i to 3 or 4 t l in 
diameter. They are non-nucleated, and react toward both basic 
and acid anilin dyes, having an amphophilic affinity. Deetjen 1 has 

1 Virchow's Arch., 1901, vol. clxiv, p. 239. 


shown that the plaques exhibit definite ameboid movement a 
property denied these bodies by the earlier investigators and 
that they are apparently nucleated. 

The plaques exist as free bodies in the general circulation, but 
directly after the withdrawal of the blood from the vessels they show 
a remarkable degree of viscosity, by virtue of which they tend to 
adhere in racemose masses, the occurrence of which at or near the 
radiating points of the fibrin network has already been described. 
Zeri and Amalgia 1 found that in malarial fever this agglutination 
of the plaques did not occur, although it was regularly observed 
in other infections, such as pneumonia, pleurisy, tuberculosis, 
enteric fever, and the exanthemata. 

The belief of Bizzozero 2 and of Hayem, 3 that the plaques repre- 
sented a so-called " third corpuscle " of the blood, is not justified, for 
it has been proved that these bodies are not distinct cellular entities, 
but rather de'bris, derived either from the blood corpuscles or from 
the plasma. It is evident, from the work of Arnold, 4 Engel, 5 
Klebs, 6 and others, that at least a large proportion of the plaques 
are simply bits of globular matter extruded from the erythrocytes, 
and in eosin-methylene-blue films the apparent eruption of plaques 
from the stroma of the erythrocytes can be readily demonstrated. 
It is possible that some of the plaques are derived by the disintegra- 
tion of the nuclei of the leucocytes (Lilienfeld ; 7 Howell ; 8 Gibson 9 ) ; 
and that still others are masses of precipitated globulin (Lowit 10 ). 
Heim 11 believes that the plaques are nuclear formations of the 
erythrocytes, and claims that regeneration of the latter and in- 
crease in the number of blood plaques progress pdri passu. 

Ducchesi's method of macroscopically demonstrating the 
plaques is of clinical interest : A few drops of blood are collected 
in a watch-glass, which is gently rocked from side to side for a few 
minutes, and then held up to the light, showing the plaques as 
groups of delicate white granules in the stratum of blood next to 
the glass. These granular masses appear within from forty seconds 
to two minutes after withdrawal of the blood, and disappear as 
coagulation commences. 

1 II Policlin., 1903, vol. ix, p. 485. 2 Virchow's Arch., 1882, vol. xc, p. 261. 

3 Compt. rend. Soc. biol., Paris, 1877, vol. ii, p. 85. 

4 Centralbl. f. allg. Path., 1897, vol. viii, p. 289. 

s " Leitfaden zur klinischen Untersuchung des Blutes," 2d ed., Berlin, 1902, 


8 Ziegler's Beitr., 1889, vol. iv, p. 528. 

7 Zeitschr. f. physiol. Chem., 1895, vol. xx, p. 155. 

8 Jour. Morph., 1891, vol. iv, p. 57. 

' Jour. Anat. and Physiol., 1886, vol. xx, p. 100. 

10 Arch. f. mik. Anat., 1891, vol. Iviii, p. 598. 

11 Deutsch. med. Wochenschr., 1903, vol. xxix, p. 588. 


Exposure to the air appears to cause an almost immediate dis- 
appearance of the plaques from the blood, and, therefore, they are 
but seldom noticed in the blood film prepared by the ordinary 
methods. With Wright's stain, however, they are readily demon- 
strable in the dry film. In studying the plaques in their fresh 
state the blood may be drawn directly through a drop of Hayem's 
solution or a weak solution of osmic acid, the mixture of the blood 
and fixative being then placed upon a slide and examined in the 
usual manner. (See p. 90.) 

The number of plaques in normal blood varies within wide 
limits, according to the statements of different authorities, but 
about 300,000 to the is generally considered the normal 
average, and from 180,000 to 500,000 the range under physio- 
logical circumstances. 

The plaques generally are increased in pernicious anemia, 
severe secondary anemias, leukemia, pneumonia, arthritis de- 
formans, myelitis, tuberculosis, bubonic plague, and as the effect of 
residence in high altitudes. They are diminished in hemophilia, 
purpura, and acute febrile diseases, such as erysipelas, typhus 
fever, and the malarial fevers. 


Muller 1 has called attention to the constant presence in normal 
and pathological blood of small, colorless, refractive bodies, of 
spheroidal or dumb-bell shape, not larger than i // in diameter. 
These bodies, to which the terms hemokonia and blood dust have 
been applied, are highly refractive, and possess active, limited 
molecular motility, but not true ameboid motion. They have 
been compared in appearance to fine fat droplets, to micrococci, 
and to granules derived from the protoplasm of the leucocytes. 
Nothing is known of their histological character and significance 
beyond the facts that they are not concerned in the process of 
fibrin formation, and that they are not fatty bodies, since they are 
neither stained by osmic acid nor dissolved by ether. Both 
Stokes and Wegefarth, 2 and Nicholls 3 regard them as free granules 
of the neutrophile and eosinophile leucocytes, and believe that 
they are probably concerned in the protective properties of the 
blood in immunity. Stengel 4 suggests that they may be simply 
the products of fragmentation of the erythrocytes, such as may 

1 Centralbl. f. Path. u. Bakteriol., 1896, vol. xxv, p. 529. 

2 Johns Hopkins Hosp. Bull., 1897, vol. viii, p. 246. 
8 Phila. Med. Jour., 1898, vol. i, p. 387. g 

4 "Text-book of Pathology," 4th ed., Philadelphia, 1903, p. 335. 



be produced by heating fresh blood to destructive temperatures, 
when bits of the corpuscles are seen to bud out, break off, and 
float free in the plasma, endowed with pseudo-ameboid motility. 
Miiller found large numbers of hemokonia in a case of Addi- 
son's disease, but these bodies were very scanty in a number of 
markedly cachectic conditions. Their occurrence in the blood 
appears to carry no definite clinical significance. 



PLATE ll. 

* - .. ?" ;'.-". 

:.: t }: * 13 

17 18 



(1-16, Triacid Stain; \~,-tf> t Eosin and Methylene-blue.) 
(E. F. FABER,/<T.) 

( Triacid Stain.) 

i, 2, 3, 4. Small Lymphocytes. 

Contrast the faintly colored protoplasm of these cells in the triple stained specimen 
with their intensely basic protoplasm in the film stained with eosin and methylene- 
blue, 17 and 18. The cell body of i is invisible. Note the kidney-shaped nucleus in 4. 

5, 6. Large Lymphocytes. 

With this stain the nucleus reacts more strongly than the protoplasm ; with eosin and 
methylene-blue (19, 20), on the contrary, the protoplasm is so deeply stained that the 
nucleus appears pale by contrast. This peculiarity is also observed in the smaller 
forms of lymphocytes. 

7, 8. Transitional Forms. 

Note the moderately basic and indented nucleus, and the almost hyaline non-granular 
protoplasm. Compare 8 with the myelocyte, 7, Plate IV, these cells differing chiefly 
in that the myelocyte contains neutrophile granules. 

9, 10, ii. Polynuclear Neutrophiles. 

These cells are characterized by a polymorphous or polynuclear nucleus, surrounded 
by a cell body filled with fine neutrophile granules. In n the nuclear structure is 
obviously separated into four parts; in 9 it is moderately, and in 10 markedly, poly- 

12, 13. Eosinophiles. 

The nuclei are not unlike those of the polynuclear neutrophile, except that they are 
somewhat less convoluted, and poorer in chromatin, staining less intensely. The pro- 
toplasm is filled with coarse eosinophile granules, the characteristics of which are 
clearly illustrated by 13, a " fractured " eosinophile. 

14. Eosinophilic Myelocyte. 

Compare with 15. 

15, 16. Myelocytes. (Neutrophilic.) 

These cells are morphologically similar to 14, except that they contain neutrophile 

instead of eosinophile granules. Note that the granules of the myelocyte are identical 

with those of the polynuclear neutrophile. A dwarf form of myelocyte is represented 

by 16. 

(Eosin and Methylene-blue.) 
17, 18. Small Lymphocytes. 

Note the narrow rim of pseudo-granular basic protoplasm surrounding the nucleus, 

and the pale appearance of the latter. 
19, 20. Large Lymphocytes. 

Budding of the basic zone of protoplasm is represented by 20. Both of these cells 

belong to the same type as 5 and 6. 
21, 22. Large Mononuclear Leucocytes. 

Compared with 19 and 20, these cells have a decidedly less basic protoplasm, but a 

somewhat more basic nucleus. In the triple stained film these differences cannot be 

detected, so that they must be classed as large lymphocytes. 

23. Transitional Form. 

The distinction between this cell and 24 is not marked ; the nucleus of the latter 
simply being somewhat more basic and convoluted. 

24, 25, 26, 27. Polynuclear Neutrophiles. 

With this stain these cells show a feebly acid protoplasm, and lack granules. Note 
that the more twisted the nucleus the deeper it is stained. Compare wuh 9, io,and n. 
28, 29. Eosinophiles. 

Compare with 12 and 13. 

30. Eosinophilic Myelocyte. 

Compare with 14. 

31. Basophile. (Finely granular.) 

This cell is characterized by the presence ol exceedingly fine 6-granules, staining the 
pure color of the basic dye. The nucleus is markedly convoluted and deficient in 
chromatin. The cell here shown was found in normal blood. 

32. 33, 34, 35, 36. Mast Cells. 

The granules take a modified basic color, as shown by their royal-purple tint in this 
illustration. Note their unusually large size and ovoid shape in 35, their peculiar 
distribution in 35 and 36, and their irregularity in size in 32 and 36. With the triacid 
mixture these granules, as well as those of the finely granular basophile, 31, remain 
unstained, showing as dull-white stippled areas in the cell body. The nuclear chro- 
matin of the mast cell is so delicate and so feebly stained that it is barely visible. 
These cells were found in the blood of a case of spleno-medullary leukemia. 



In the fresh, unstained blood film the leuco- 
APPEARANCE cytes are recognized as pale nucleated cells, the 
IN majority of which are larger in size than the 

FRESH BLOOD, erythrocytes, by which they are greatly outnum- 
bered, the proportion of the former to the latter 
ranging approximately between i : 450 and i : 1200 in normal 
blood. The size of the normal white corpuscles varies from 
about 7 // to 10 or 12 /> in diameter, and their shape, while in 
the resting stage, is irregularly round or oval. 
' By careful examination four different varieties of these cells 
may be distinguished, the distinction between these forms being 
made more striking by the addition of a small quantity of a one 
per cent, acetic acid solution to the fresh film. These varieties, 
which are essentially the same as those first described by Schultze, 
in 1865, 1 are as follows: (i) Non-ameboid cells about the size 
of the normal erythrocyte, consisting of a pale, compact, spher- 
ical nucleus encircled by a narrow zone of homogeneous proto- 
plasm. (2) Ameboid cells almost twice the size of the erythro- 
cyte, consisting of a rather coarsely meshed nucleus, spherical, 
ovoid, or indented in form, surrounded by a relatively large 
amount of clear protoplasm. The latter is highly opaque, for 
although it forms an exceedingly thin layer when spread out flat, 
it effectually obscures the outlines of objects over which it lies 
as an explanation for which characteristic Kanthack and Hardy 2 
presume that the cell matter is composed of a colorless basis 
embedding immense numbers of minute vacuoles filled with a 
substance of a different refractive index. (3) Ameboid cells of 
slightly smaller size than the second variety, consisting of a single 
twisted nucleus, or of two or more separate round or ovoid nuclei, 
embedded in a body of protoplasm crowded with exceedingly 
delicate, moderately refractive granules. The nuclear network is 

1 Arch. f. mik. Anat., 1865, vol. i, p. i. * Jour. Physiol., 1894-95, vol. xvii, p. 81. 



composed of chromatin threads closely united to form a compact, 
lobulated structure, and the protoplasm appears 'to consist of a 
transparent substance, of gelatinous character, having a refractive 
index but slightly below that of the granules which it contains. 
(4) Ameboid cells containing a convoluted nucleus, or several 
spherical nuclei, embedded in a protoplasm filled with coarse, 
highly refractive, fat-like granules. The nuclear structure con- 
sists of a coarsely meshed, knotted network, and the protoplasm 
is much less refractive than its granules, being clear and struc- 
tureless in appearance. 

Spontaneous changes in the shape of the larger 
AMEBOID varieties of leucocytes may be observed if the slide 
MOVEMENT, is placed upon a warm stage having a temper- 
ature of about 98.5 F. During these ameboid 
movements the shape of the cells constantly undergoes alteration 
by the alternate contraction and expansion of the protoplasm. 
Tentacular processes reach out from various portions of the 
cell body, while at other points its surface becomes retracted, 
so that it may appear as an irregular nucleated mass provided 
with one or more long, snake-like arms projecting from a central 
body. These ameboid cells are chiefly concerned in the process 
of phagocytosis, or the engulfing and destruction of micro-organ- 
isms and other foreign matter which may gain entrance into 
the circulating blood, and to leucocytes which exert this function 
the term phagocyte has been applied. The well-known experi- 
ments of Metschnikoff l have shown their propensity for seizing 
upon and devouring pathogenic bacteria, such as the anthrax 
bacillus and the erysipelas streptococcus, and further proof of 
such phagocytic action may frequently be found in the fragments 
of other foreign matter, such as bits of old blood clots, malarial 
pigment, and fat droplets inclosed in their protoplasm. 

It has also been suggested by Gabritschewsky 2 that it may be 
possible under some circumstances that phagocytes are capable 
not only of engulfing solid bodies, but that they may also imbibe 
liquid substances, which are thus rendered harmless to the 
organism, and to this property the term pinocytosis has been 
given by this author. Drugs injected hypodermically are taken 
up by phagocytic cells, which, according to Labbe", 3 not only 
absorb and assimilate medicaments, but perhaps carry them, by 
election, to a specific lesion of the organism mercury to a syphil- 
itic nidus, for instance. Iron, iodin, arsenic, mercury, iodoform, 

1 "L' Inflammation," Paris, 1892. 

2 Annal. de 1'Instituf Pasteur, 1894, vol. viii, p. 673. 

3 Presse me'd., 1903, vol. ii, p. 725. 


and the salicylates are among the drugs dealt with in this 

The ameboid property of the leucocytes is also responsible for 
the ease with which these cells escape from the blood vessels into 
the perivascular tissues in inflammatory lesions, and to a less ex- 
tent in health. This well-known process of diapedesis is facili- 
tated by virtue of the leucocyte's ability to elongate and flatten 
out so that it may readily emigrate through the spaces between 
the endothelial cells of the vessel wall. 

The identification of the various forms of 
CELL leucocytes depends largely upon the presence or 

GRANULES, absence of granules in their protoplasm, and upon 
the distinctive manner in which these granules 
react toward the acid, basic, and so-called neutral solutions of 
the anilin dyes. By means of this method of "color analysis" 
Ehrlich has provided a rational means by which the study of the 
leucocytes may be undertaken. 

Five varieties of granules, which are designated by the use of 
the Greek letters , /9, ^, d, and e, may be recognized in the cell 
bodies of the leucocytes, as follows : 

1. a- granules (eosinophile, oxyphile, or coarse oxyphile gran- 
ules): Coarse, spherical or ovoid, highly refractive granules of 
a peculiar fat-like appearance, showing a striking affinity for acid 
stains, especially for eosin. In normal blood they occur only in 
leucocytes with polynuclear or polymorphous nuclei, but in cer- 
tain pathological conditions they may be found in that variety of 
the leucocyte known as the eosinophilic myelocyte. 

2. p- granules (amphophile granules): Fine granules which 
are capable of reacting toward both acid and basic dyes, inva- 
riably staining with the former and sometimes with the latter, if 
the stains are used singly, while in a mixture of the two they 
always react toward the acid dye. These granules never occur 
in normal blood, but in some pathological conditions a varying 
proportion of the leucocytes may exhibit amphophilic reactions 
on the part of some of their granules. 

3. y-granules (mast cell or coarse basophile granules): Very 
coarse granules, measuring from 0.2 to 0.4 // in diameter, and pos- 
sessing an intense affinity for basic dyes. If stained with car- 
boltoluidin-blue, with thionin, or with alkaline methylene-blue, 
they are colored a distinctive deep purplish-red. These granules 
occur in a form of leucocyte known as the mast cell, which is 
abundant in myelogenous leukemia, and is met with occasionally 
in other diseases. 

4. ^-granules (fine basophile granules): Fine granules, stain- 


ing with basic dyes, and occurring under normal conditions in 
leucocytes having polymorphous nuclei. They are most clearly 
demonstrated with such basic dyes as thionin or methylene-blue, 
by which they are stained a deep blue color. 

5. e- granules (neutrophile or fine oxyphile granules) : Exceed- 
ingly fine granules, formerly thought to have a selective affinity 
for the neutral element of a solution composed of acid and basic 
dyes, but now known to have, in reality, a feeble oxyphilic ten- 
dency. They occur abundantly in the normal polynuclear 
neutrophile cells, and also in several pathological forms of leuco- 
cytes : the myelocyte, the small mononuclear neutrophile, and 
the "small neutrophilic pseudolymphocyte." 

But little is known of the real nature and function of the leuco- 
cyte granules, in spite of their elaborate study by different in- 
vestigators. Two leading views, which excite much controversy, 
to-day command attention: the hypothesis of Ehrlich 1 and the 
bioblastic theory of Altmann. 2 Ehrlich regards them as an 
evidence of a specific secretory function on the part of the cells, 
which under normal conditions contain but a single variety of 
granules. They are to be considered as products of cellular 
metabolic activity, and are destined to be given off in the vicinity 
of the cells, this elimination perhaps constituting one of the most 
important functions of the latter. Far from representing mere 
waste-products, as some authors contend, they are in reality ele- 
ments of decided, although obscurely defined, value to the or- 
ganism. Altmann, in his bioblastic theory, considers cell granules 
as definite biological entities, and believes that they "serve as a 
basis for the explanation of the many phenomena of organic metab- 
olism." In summing up their functions he remarks that "they 
effect through oxygen-transmission both reductions and oxygena- 
tion, and in this manner accomplish the disunions and the syn- 
theses of the economy without sacrificing their own individuality." 

As a rule, the cell granules are thought to be relatively simple 
bodies, although their exact composition is as yet undetermined. 
It has been proved by Weiss 3 and others that they are of albu- 
minous character. The eosinophile granules, in which iron has 
been demonstrated by Barker 4 and other observers, are more com- 
plex than the other varieties. They are of a higher histological 
structure, consisting of an external limiting portion which may 

1 Loc. cit. 

2 "Ueber die Elementarorganismen und ihre Beziehungen zu den Zellen," 
2d ed., Leipsic, 1894. 

3 "Hematologische Untersuchungen," Vienna, 1896. 

4 Johns Hopkins Hosp. Bull., vol. v, p. 93. 



be clearly differentiated from the central area. Hankin and 
Kanthack 1 have determined the fact that increased bactericidal 
power of the blood is closely correlated with the discharge of 
both eosinophile and neutrophile granules into the plasma, and 
the former observer 2 has furthermore shown that in experimental 
infections there is at the point of the infection an accumulation of 
cells containing eosinophile granules, together with a discharge 
of such granules during the conflict of the cells with the invading 

In the normal adult the number of leucocytes 

NORMAL NUM- in the peripheral circulation averages from about 

BER. 5000 to 10,000 to the of blood. In the 

majority of instances, in which the influences of 

physical factors are excluded, a count of 7500 leucocytes per 

may be regarded as the mean normal average. Variations of several 

thousand cells per above and below this number are within 

physiological limits, and frequently occur because of the extreme 

susceptibility of the leucocytes to agencies causing such transient 

fluctuations. The following table, compiled from data given by 

Hayem, 3 Grawitz, 4 and von Limbeck, 5 shows the average number 

of leucocytes determined by various authorities : 

Thoma 8687 per 

Von Limbeck 8500 

Rieder 7680 

Boeckman; Halla 7533 

Graeber; Reinecke 7242 

Tumas 6200 

Hayem 6000 

Average 7406 " 


Six distinct varieties of leucocytes may be recognized in the 
healthy adult's blood stained by the Romanowsky methocl or by 
Ehrlich's stain, according to the methods described in a previous 
section. These varieties, together with their normal relative per- 
centages and absolute number to the of blood, are as fol- 


Small lymphocytes 20-30 1000-3000 

Large lymphocytes and transitional forms 4-8 200-800 

Polynuclear neutrophiles 6075 3000-7500 

Eosinophiles -5~5 25-500 

Basophiles, as high as 0.5 25 

1 Centraibl. f. Bakt. u. Parasit., 1892, vol. xii, p. 777; ibid., 1893, vol. xiv, p. 852. 

2 Jour. Physiol., 1894-95, vol. xvii, p. 81. * Loc. cit. 
4 "Klinische Pathologic des Blutes," Berlin, 1896. ' Loc. cit. 


With the decline of life the proportion of polynuclear neutro- 
philes rises, the lymphocytic forms becoming correspondingly less 
numerous. This reversal of the youthful blood picture is, accord- 
ing to Dobrovici's researches, 1 most conspicuous after the sixtieth 
year of age. 

The variations in these numbers and percentages, which de- 
pend upon different physiological and pathological influences, are 
referred to in other sections. 

The lymphocytes, or small lymphocytes, as 
SMALL they are commonly designated in contradistinc- 
LYMPHOCYTES. tion to the large mononuclear forms, are non- 
granular cells which measure from about 5 to 
io// in diameter, their average size being that of the normal 
erythrocyte, or 7.5 p in diameter. The typical cell of this class 
consists of a single round, deeply staining nucleus surrounded by 
a narrow zone of protoplasm, and sometimes provided with one 
or two pseudo-nucleoli, situated eccentrically upon the nuclear 
surface. The nucleus is so relatively large that it almost com- 
pletely fills the cell, being its most conspicuous part, while the rim 
of protoplasm is usually so narrow and poorly defined that at first 
glance it may escape notice. These characteristics a relatively 
large nucleus and a relatively scanty amount of protoplasm are 
more conspicuously exhibited in the smaller than in the larger 
forms of these cells. (Frontispiece, II, and Plate II, Figs. 1-4, 17, 
and 1 8.) 

By Romanowsky's method the nucleus stains purple, in which 
a red tone prevails, and the protoplasm pure sky-blue. Occa- 
sionally the protoplasm is stippled with a few rather coarse purple 
or red granules. 

In films stained with simple eosin and methylene-blue solu- 
tions the nucleus shows a decided affinity for the basic dye, usually 
staining dark blue, or, more rarely, pale green. The protoplasm 
shows as a relatively narrow encircling area of deep blue color, 
which has been likened in appearance to the surface of ground 
glass; it is much more intensely basic than the nucleus, which 
looks pale by contrast. With Ehrlich's triple stain the nucleus, 
being rich in chromatin, is colored deep blue or purple, and the 
protoplasm is either entirely unstained, appearing as a narrow 
hyaline halo surrounding the nucleus, or it is tinged a delicate 
shade of pink if it happens to react toward the acid fuchsin of the 

Occasionally small lymphocytes are encountered in which the 
nucleus is atypical both in morphology and in staining properties. 

1 Sem. med., 1904, vol. xxxiv, p. 198. 


Thus, some cells contain a pale, almost hyaline nucleus, composed 
of an exceedingly scanty chromatin structure which reacts very 
feebly to the basic dyes ; others contain a deeply stained, indented 
or kidney-shaped nucleus, similar in shape to that of the so-called 
"transitional" forms; while still others are provided with a 
nucleus which has evidently become completely divided, so that 
such a cell really contains two distinct hemispherical nuclei, rich 
in chromatin, deeply stained, and situated toward the poles of 
the cell body. These irregular forms of lymphocytes occur both 
in normal and in pathological blood, but with much greater fre- 
quency in the latter, especially in both forms of leukemia. 

The small lymphocyte appears to possess greater powers of 
resistance than any other variety of leucocyte. In his studies 
of necrobiosis of the blood corpuscles Bodou 1 determined that 
the degenerative changes first involved the large mononuclear 
hyaline cells regarded as myelogenous in type; next, the transi- 
tional forms; next, the large lymphocytes of lymphatic origin; 
next, the polynuclear neutrophiles ; and last of all the small lym- 

Under this term it is convenient to include 
LARGE both the larger forms of the true lymphocyte 
LYMPHOCYTES, those measuring 1 1 p. or more in diameter and 
also that variety of hyaline cell known as the 
large mononuclear leucocyte. These two forms of cells, although 
they are generally considered as distinct histological species, one 
being a true lymphocyte and the other probably a marrow-bred 
element, may, for practical purposes, be classed together, since it 
is impracticable to differentiate one from the other in the speci- 
men prepared for an ordinary clinical examination. 2 (Frontis- 
piece, II, and Plate II, Figs. 5, 6, and 19-22.) 

Cells of this type may range in size from n to 15 p. or even 
larger in diameter, and are usually of round or ovoid shape, ex- 
cept in an occasional cell, where, in consequence of the injury 
received during the preparation of the blood film, the outline 
may be exceedingly irregular and deformed. The nucleus, which 

1 Virchow's Arch., 1903, vol. clxxiii, p. 485. 

2 Some authors, Ehrlich himself among them, maintain that a distinction be- 
tween these two forms of cells may invariably be made in the stained specimen. 
Thus, in the film stained with methylene-blue, it is held that the true lymphocyte, 
no matter what its size, always possesses a strongly basic protoplasm and nucleus, 
the latter staining less deeply than the former; while the large mononuclear leuco- 
cyte has a feebly basic protoplasm and nucleus, the latter staining more intensely 
than the former. These points of difference, although they may be distinguished 
in specimens stained by special methods, seem to be too finely drawn to justify 
their acceptance as reliable criteria for the identification of these two groups of 
cells in films prepared by the technic adapted to routine clinical work. 


is round, ovoid, or somewhat elongated, is generally situated to- 
ward the periphery of the cell body. In most of the cells the 
amount of protoplasm is relatively greater than that of the small 
lymphocyte, but occasionally this peculiarity cannot be distin- 

With Wright's stain this cell stains fainter than, but in other 
respects like, the small lymphocyte, save that it occasionally shows 
a number of coarse and fine red granules scattered through its 

With simple mixtures of a strong acid and basic dye, such as 
eosin and methylene-blue, the nuclear chromatin stains a diffuse 
sky-blue tint, and the protoplasm exhibits a more or less decided 
affinity for the basic element of the staining fluid. This tendency 
is very marked in some cells, the protoplasm of which contains 
an intensely basic pseudo-granular zone staining much deeper 
blue than the rest of the cell body, paralleling the extreme periph- 
ery of the cell, and often apparently separated from the nucleus 
by a distinct unstained area. In other cells this basic affinity is 
not so conspicuous, their protoplasm staining a diffuse purplish 
shade in which a rose-red tone prevails. 

The nucleus, being poor in chromatin, stains pale blue with the 
triple stain, and is usually so delicately tinted that it is almost 
invisible ; the protoplasm is faintly tinged with pink or with gray- 
ish-blue, or it may remain practically colorless, showing merely 
as an indefinite hyaline area surrounding the nucleus. 

Apparent extrusion of portions of the cell body is not uncom- 
monly observed, this phenomenon producing a peculiar "frayed- 
out," ragged appearance around the periphery of the lympho- 
cyte, due to the partial detachment of small bits of the peripheral 
seam of basic protoplasm, which loosely adhere to the outer 
margin of the cell. Occasionally these small basic masses be- 
come entirely detached, and may be seen lying free in the plasma, 
alongside the cell of which they were once a part. 

Typical forms of the large and small lymphocyte, such as are 
seen in the great majority of stained blood films, may be dis- 
tinguished without difficulty, but in some diseases, notably in the" 
lymphatic variety of leukemia, irregular forms of these cells are 
found, the size and nuclear characteristics of which are so confus- 
ingly atypical that it is sometimes futile to attempt the classification 
of such hybrids into two arbitrary groups, large and small. Thus 
one may meet with cells the size of the small lymphocyte, but hav- 
ing a feebly basic, eccentric nucleus and a relatively large amount 
of protoplasm; and with cells identical with the large lymphocyte 
except that they possess a small, spherical, strongly basic nucleus. 


The reddish protoplasmic granules of the large lymphocyte, shown 
by the Romanowsky stain, serve here as valuable criteria, but these 
granules, unfortunately, are not always demonstrable in irregular 
cells. In attempting to differentiate these atypical forms in the 
triple stained specimen it is safe to be guided by the suggestions 
given by Thayer, 1 who is inclined to place more emphasis upon the 
character of the nucleus than upon the size of the cell body as a 
whole. Thus, in a doubtful mononuclear, non-granular cell in 
which the nucleus is similar in size and shape to that of the small 
lymphocyte, regardless of its affinity for the basic element of the 
stain, the cell is classed as a small lymphocyte, until the size of 
such a cell exceeds that of the polynuclear neutrophile. Some 
cells no larger than the smallest lymphocyte may be classed as 
large lymphocytes if their nuclei are decidedly ovoid in shape 
and pale in color. In spite of every precaution, however, it 
must be admitted that in some instances differential counts of 
these two types of cells must be more or less inaccurate, for the 
obvious reason that so much depends upon the personal equation. 
The so-called transitional forms are cells 
TRANSITIONAL which closely resemble the large lymphocyte in 
FORMS. shape and in size, but which differ from the latter 
variety of cell chiefly in having a nucleus which, 
instead of being ovoid in shape, is indented and drawn out into the 
form of a crescent with rounded poles, the concave aspect of the 
nuclear figure lying toward the center of the cell. 'In other forms 
the nucleus may have become molded into a figure resembling 
an hour-glass, which occupies the central portion of the' cell body, 
not lying in contact with its periphery at any point. (Frontispiece, 
II, and Plate II, Figs. 7, 8, and 23.) 

With eosin and methylene-blue the nucleus shows a moder- 
ately strong affinity for the basic dye, being colored much darker 
blue than the nucleus of the large, but distinctly paler ^han that 
of the small, lymphocyte ; the protoplasm is stained a diffuse 
pale blue, in which the pink tinge of the eosin conspicuously 
prevails. With the triple stain the nucleus of this cell -is usually 
stained somewhat darker blue than that of the large lymphocyte, 
and the protoplasm is either quite colorless or, perhaps, slightly 
tinged a grayish-blue. 

Inasmuch as the clinical significance of the transitional forms 
is identical with that of the large lymphocytes, it is customary to 
class both forms together under a single heading in the percentage 
table of the different forms of leucocytes. 

1 Johns Hopkins Hosp. Reports, 1894, vol. iv, p. 103. 


Polynuclear neutrophiles are cells which, as a 
POLYNUCLEAR general rule, measure about from 10 to 12 // in 
NEUTROPHILES. diameter, although their size may vary within 
wide limits, some being not much larger than 
the small lymphocytes, while others are nearly twice this size. 
(Frontispiece, III, and Plate II, Figs. 9-11 and 24-27.) The 
distinguishing characteristics of these cells are the twisted, 
polymorphous nature of the nuclei and the so-called "neutro- 
philic" reaction of the granules embedded in the protoplasm. 
The nucleus may be of almost any shape elongated, wreathed, 
lobulated, horseshoe-shaped, or twisted into designs resembling 
various letters of the alphabet, such as S, Z, U, or E. It 
usually consists of several apparently separate masses of ir- 
regular shape, connected with each other by delicate filamentous 
strands of chromatin, which dip beneath the surface of the pro- 
toplasm, and, owing to the density of the overlying granules, are 
invisible or but dimly defined in the triple stained specimen. 
By the use of the simpler double stains, such as eosin and 
methylene-blue, the presence of these" connecting chromatin 
threads may be demonstrated with great clearness. Less com- 
monly, a cell contains several small oval or round nuclei, which 
are actually separated from each other, complete division at the 
points of constriction having resulted in the production of two or 
three, and in rarer instances even six or seven, distinct nuclei. 
The nuclear structure is rich in chromatin, which forms a dense, 
unevenly staining network possessing a marked affinity for the 
various basic dyes. It stains dark blue or greenish-blue with the 
triple stain, and still more intensely blue with eosin and methyl- 
ene-blue solutions. 

The fact that the single, twisted type of nucleus predominates 
in these cells has led to the current use of the adjective "poly- 
morphonuclear " as a substitute for " polynuclear," but it is per- 
fectly obvious that both terms may be used synonymously, the 
latter perhaps being preferable, because of its brevity, and of its 
established vogue. The irregularity of the nucleus is regarded 
as a sign of the ameboid activity of the cell, as first suggested by 
Arnold, 1 $nd not as an indication of degeneration, as 'formerly 
believed. It has been effectually demonstrated by Sherrington 2 
that if such cells are allowed to quiet down before they are killed, 
their nuclei usually return to a spheroidal form. 

The protoplasm of the polynuclear neutrophile is densely packed 
with exceedingly fine, so-called neutrophile granules, which stain 

1 Arch. f. mik. Anat., 1887, vol. xxx, p. 226. 

2 Proc. Internal. Congress of Physiologists, Liege, 1892. 


lavender or purple, or, rarely, pink, with Ehrlich's triacid mix- 
ture, but which are not stained by simple solutions of eosin and 
methylene-blue. With Wright's stain these granules are colored 
reddish-lilac or pink. Kanthack and Hardy 1 have shown that 
these granules have "a minimal attraction for acid dyes, or, 
briefly, a minimal oxyphile reaction," and, furthermore, that 
Ehrlich's neutral mixture, by which they are intensely stained, 
is not, chemically speaking, a neutral stain, but, on the contrary, 
a powerful and exceedingly differential acid dye, intensely staining 
oxyphile granules of all varieties. 2 Thus, having proved that 
the granules of the polynuclear " neutrophile " cell of Ehrlich 
display a distinct, although feeble, affinity for acid dyes, and that 
they are unstained by basic and neutral dyes, the term "finely 
granular oxyphile cell" has been adopted by these authors for 
this variety of leucocyte, the granules being known as "finely 
granular oxyphile" granules. It is doubtful, however, if the use 
of these unwieldy terms will receive general approval, except by 
certain of the British school. To designate a polynuclear leuco- 
cyte as a "finely granular oxyphile cell" is even more glaringly 
inappropriate than the use of Ehrlich's term, "neutrophile," for 
other varieties of leucocytes i.e., myelocytes and " neutrophilic 
pseudolymphocytes " may be just as fittingly described by the 
former phrase. 

The granules are of very small size and of irregular wedge- 
or spike-shape, never being spherical or ovoid in contour. They 
are usually most densely distributed about the periphery of the 
cell, whence they gradually shade off toward the nucleus, which 
is frequently found to be encircled by a perfectly hyaline, non- 
granular zone. The granules are not always confined to the cell 
protoplasm, being scattered over the nucleus, portions of which 
may be partly obscured by the overlying granular film. 

The jelly-like substance of the protoplasm in which the granules 
are embedded appears to show a slight affinity for acid -dyes, the 
intensity of this affinity varying greatly in different cells. With 
the triple stain this reaction is evidenced by the variable depth 
of fuchsin-colored undertone which may be detected beneath the 

1 Jour. Physiol., 1894, vol. xvii, p. 61. 

2 Reasoning upon the basis that eosin stains with most striking intensity in an 
aqueous solution, less decidedly in a glycerin solution, and even less strongly when 
dissolved in strong alcohol, these investigators distinguish three classes of oxyphile 
granules, according to the intensity of their affinity for acid dyes, thus: (i) Those 
which stain with eosin only in aqueous solutions or in alcoholic solutions of a per- 
centage below 60; (2) those which stain in both aqueous and glycerin solutions, 
but not in a strong alcoholic solution (90 to 95 per cent.) of the dye; and (3) those 
which stain with aqueous, glycerin, and strong alcoholic solutions. They include 
in the first class the neutrophile and the amphophile granules of Ehrlich. 


purplish color of the granules; while in the specimen stained 
with eosin and methylene-blue the protoplasm is tinted evenly 
the color of eosin. 

These cells are the most conspicuous of all the 
EOSINOPHILES. leucocytes, and may be at once identified by the 
presence of a more or less polymorphous nucleus 
embedded in a protoplasm studded with coarse, highly refractive 
granules which have a strong affinity for acid dyes, such as eosin 
and acid fuchsin. (Frontispiece, V, and Plate II, Figs. 12, 13, 
28, and 29.) Owing to the large size of their granules and 
to their striking oxyphilic reaction, these cells are also known by 
the term "coarsely granular oxyphile cells," in contradistinction 
to the "finely granular oxyphile cells" or polynuclear neutrophiles 
(Kanthack and Hardy). The size of the eosinophile varies very 
greatly, but most of them approximate the size of the polynuclear 
neutrophile, or are, perhaps, a trifle smaller. Their diameter 
commonly ranges from 8 to 1 1 ,, although occasionally forms 
not larger than the normal erythrocyte are to be observed. Their 
shape is usually that of an irregular sphere or oval. 

The nucleus may be kidney- or horseshoe-shaped, or twisted 
and drawn out into an irregular mass, but it is rarely as con- 
stricted and deformed as that of the polynuclear neutrophile. It 
is nearly always situated eccentrically, cells of this variety with 
centrally placed nuclei being very uncommonly seen. Occasion- 
ally the eosinophile contains multiple nuclei, consisting of sev- 
eral oval or round masses between which no connecting chro- 
matin threads can be distinguished, but usually such portions 
of the nucleus are joined together by extensions of chromatin 
running beneath the protoplasm. The nucleus stains faintly, in 
comparison with that of the polynuclear neutrophile, although 
more intensely than that of the large mononuclear cell; it is col- 
ored pale blue or greenish-blue by the triple stain, and dark blue 
by eosin and methylene-blue mixtures. 

The granules, which are relatively large in size and quite regu- 
larly spherical in shape (in contrast to the delicate, irregularly 
shaped granules of the polynuclear neutrophile), react strongly 
toward the acid elements of the triple stain; some are stained 
a brilliant fuchsin color, some deep red, while others are brown- 
ish-yellow or copper color, or even almost black; with mixtures 
of eosin and methylene-blue they take the brilliant color of eosin. 
There appears to be a marked tendency on the part of the gran- 
ules to overrun the nucleus, so that its morphology in some cells 
is almost indistinguishable. The granules are also prone to be- 
come readily detached from the protoplasm, which doubtless ac- 


counts for their uneven, blotchy distribution in many cells, in 
which densely packed granular areas alternate with open spaces 
merely punctuated here and there with an occasional granule. 

Eosinophiles appear to offer but feeble powers of resistance 
against external influences, so that it is common to find these 
cells so injured by the process of making the film that the nu- 
cleus has escaped from the cell body, and the granules, lying 
free in the plasma, are scattered about it in a cloud. This insta- 
bility or "explosive" character of the eosinophile is one of its 
most striking attributes, for, while observed now and then in a 
polynuclear neutrophile, it occurs with much greater frequency 
in eosinophiles than in the latter type of leucocyte. 

The protoplasm of the cell may or may not show an affinity for 
the anilin dyes; usually it does not, so that the granules appear 
to be embedded in a perfectly hyaline substance; occasionally 
the protoplasm is faintly stained by fuchsin or by eosin. With 
Wright's stain it frequently takes the color of the basic dye, 

BASOPHILE Finely granular basophile cells, containing 

CELLS. Ehrlich's 5-granules, are occasionally encountered 
in normal blood, but with such rarity that their 
real significance is not understood. (Plate II, Fig. 31.) 

In general morphology and size these cells resemble the poly- 
nuclear neutrophiles. The nucleus is invariably twisted, and usu- 
ally consists of two or three distinct lobes joined, by thin chro- 
matin bands ; in the stained specimen it is never of round or oval 
shape, but always shows evidences of polymorphism. The nuclear 
structure is composed of a delicate, scanty network of chromatin, 
and has a moderate affinity for basic dyes, staining dull blue with 
the triple stain and pale sea-green with eosin and methylene-blue 

The protoplasm of the cell is closely packed with fine, irregu- 
larly shaped granules having an intensely basic reaction-; they 
stain deep blue with solutions containing methylene-blue, but are 
not colored by the triple stain, showing in films stained with 
this mixture as groups of dull white spots scattered through the 
cell body. Wright's stain is most useful in bringing out the char- 
acteristics of these granules. 

Myelocytes, or marrow cells, are relatively 

MYELOCYTES. large round or oval cells, ranging from 10 to 20 [* 

or even more in diameter, their average size 

being somewhat larger than that of the large lymphocyte, which 

they resemble in general morphology. (Frontispiece, V and VI, 

and Plate II, Figs. 14-16 and 30.) The nucleus of the typ- 


ical myelocyte is of spherical or ovoid shape, and is situated 
eccentrically, lying distinctly toward one side of the cell, so that 
the peripheries of both cell and nucleus are often closely adjacent 
for some little distance usually for from one-third to one-half 
of their course. The nucleus reacts feebly toward the basic 
element of the triple stain, being colored a pale, delicate sky- 
blue with this solution ; it stains a moderately deep blue or purple 
with eosin and methylene-blue mixtures, and appears to be more 
coarsely netted and deeply stained than in films prepared by the 
preceding method. 

In the smaller forms of myelocytes the nucleus is frequently 
found to occupy the center of the cell body, so that it is surrounded 
on all sides by a protoplasmic zone of even width. In some of 
the larger forms the nucleus may be indented and molded along 
one margin of the cell body like that of the so-called "transitional" 
leucocyte. In rare instances actual division of the nucleus appears 
to have occurred, so that two separate nuclei, each shaped like a 
flattened hemisphere and situated at an extreme pole of the cell, 
may be found. Such cells are often mistaken at first glance for 
polynuclear neutrophiles, inasmuch as both forms of cells contain 
multiple nuclei and neutrophile granules; but the nucleus of the 
polynuclear neutrophile is always more or less twisted and of un- 
dulating surface, relatively rich in chromatin and stained with 
decided intensity, and rarely situated at the poles of the cell, while 
the nuclear halves of this type of the myelocyte are of regular out- 
line and uniformly close to the surface of the cell, relatively poor 
in chromatin and faintly stained, and invariably occupy the extreme 
poles of the cell body. 

The protoplasm of the myelocyte is filled with fine neutrophile 
granules, such as occur in the polynuclear neutrophile; they are 
most densely distributed at the periphery, and grow appreciably 
less abundant as they approach the nucleus, which they may 
overrun, spreading over its surface like a thin veil, so that its 
structure is more or less hidden. 

This one characteristic the presence of neutrophile granules 
in the protoplasm at once serves to distinguish the myelocyte 
from the large lymphocyte, which it may exactly resemble in 
size, shape, and nuclear structure; the importance of using a 
selective neutrophile stain to differentiate these granules in speci- 
mens used for differential counting is therefore patent. 

With Ehrlich's triple stain the granules stain purple or lavender, 
exactly like those of the polynuclear neutrophile. With Wright's 
solution the protoplasm has an undertone of light purple, broken 
here and there by indistinct, darker granular areas of the same 


color, indicating the presence of basophile granules, in addition to 
those of neutrophile reaction, which show as a delicate lilac or 
pink stippling. 

In certain pathological conditions, notably in myelogenous 
leukemia, an occasional myelocyte may be observed which con- 
tains both fine neutrophile and very coarse basophile granules, 
the latter being precisely identical in size, shape, and tinctorial 
qualities with Ehrlich's y or mast cell granules. They are 
situated both in the protoplasm of the cell and over the nucleus, 
and are, in the author's experience, seen most clearly in specimens 
stained in solutions containing polychrome methylene-blue. The 
basic granules show in such preparations as a coarse, brilliant, 
purple stippling, contrasting vividly with the paler, eosin-colored 
neutrophile granules which fill the body of the cell, and with the 
greenish-blue color of the nucleus. 

Eosinophilic myelocytes^ or myelocytes with a protoplasm filled 
with coarse eosinophile instead of neutrophile granules, are 
common to several pathological conditions, but occur with es- 
pecial frequency in the myelogenous variety of leukemia and 
also in. pernicious anemia, to some extent. Such cells are iden- 
tical in size and morphology of cell body and nucleus with the 
commoner neutrophilic myelocytes, from which they differ only 
in containing eosinophile granules. (Frontispiece, V, and Plate 
II, Figs. 14 and 30.) 

The normal habitat of the myelocyte is in the red bone mar- 
row, and its presence in the circulating blood must always be re- 
garded as pathological. At one time regarded as practically 
pathognomonic of leukemia, the myelocyte is now known to occur 
in many other conditions, especially those characterized by pro- 
found cachexia, by marked anemia, and by increase in the num- 
ber of leucocytes. The occurrence of myelocytes in the blood 
in various diseases and the clinical significance of these cells are 
discussed in another place. (See " Myelemia," p. 259.) 

Cells containing Ehrlich's ^-granules, known 
MAST CELLS, by the term mast cells, or mastzellen, #re oc- 
casionally present in the peripheral circulation, 
as the result of certain pathological influences, but are totally 
foreign to the normal blood of man. (Frontispiece, VII, and Plate 
II, Figs. 32-36.) They are very constantly found, generally in 
considerable numbers, in the myelogenous type of leukemia, and 
also occur, in small percentages, in many cases of pernicious 
anemia and in other grave blood disorders. 

The cells are of spherical or ovoid shape, and are characterized 
by a relatively large, structureless nucleus inclosed in an almost 


indefinable protoplasm, and by the presence of coarse basophile 
granules scattered irregularly over the surface of the cell marks 
of identification which remain unchanged whatever the size of 
the cell may be. No variety of cell found in the blood exhibits 
wider ranges in size. The forms most commonly observed meas- 
ure approximately from 9 to 12 n in diameter; some have a di- 
ameter of fully 20 or even 22 /./, but cells of this extremely large 
size are the exception rather than the rule; others are scarcely 
larger than the small lymphocyte, being but 7 or 8 p. in diameter, 
and these very small' forms are also uncommon. 

The nucleus is round, oval, or somewhat lobulated, and occu- 
pies the greater part of the cell body, in which it is usually situ- 
ated eccentrically. Owing to the similarity in the appearance of 
the nucleus and the protoplasm it is frequently impossible to de- 
termine the precise point at which the former structure begins 
and the latter ends, so that, in the stained specimen, many cells 
are met with which appear to consist simply of irregular groups 
of granules clinging to a pale nucleus, every definite trace of the 
cell body being lost. (Plate II, Figs. 35 and 36.) In films stained 
with Wright's solution (which is, by far, the most satisfactory stain 
for illustrating the finer morphology of these cells) the nucleus is 
colored a beautiful, iridescent greenish-blue, the tint of which 
is so extremely delicate that in many cell,s it is barely perceptible. 
The staining, though faint, is even and clear, indicating a structure 
almost totally devoid of chromatin. 

The granules are generally large and coarse, and vary greatly 
in size and in shape. Some are smaller than the granules of the 
eosinophile cell, while others approach or even slightly exceed 
0.5 ft in diameter. They may be spherical, egg-shaped, or roughly 
cuboid, the latter form of granule being exceedingly common. 
A single type of granules is not always found to the exclusion of 
the others, for one cell often contains granules of every possible 
variety of shape and size; this peculiarity is especially striking 
in some of the smaller forms of cells in which extremely coarse 
egg-shaped and smaller spherical granules may be distinguished 
clinging to the periphery of the nucleus, about which no evidence 
of protoplasm is demonstrable. (Plate II, Fig. 36, and Frontis- 
piece, VII.) In other forms, both large and small, the large 
spherical or ovoid granules may prevail almost exclusively. 
(Plate II, Figs. 33, 34, and 35, and Frontispiece, VII.) The 
distribution of the granules through the cell follows no constant 
rule, but it is evident that a more or less decided tendency exists 
toward their collection near the periphery. They are always 
most densely distributed at this point, sometimes extending in- 


ward over the nucleus, which is thus partly hidden, and some- 
times crowded into a limited zone, which coincides with the outer 
boundary of the cell for the greater part of its extent. 

The granules of the mast cell show an intense affinity for basic 
anilin dyes, toward which they react metachromatically in a 
highly characteristic manner. With Wright's solution they are 
stained a deep royal purple color in which the red tone is dis- 
tinctly evident, thus differing from the granules of other basophile 
cells, which are stained a pure blue with this mixture. Dr. H. F. 
Harris has called the writer's attention to another distinctive 
method of identifying these granules, by first staining with carbol- 
toluidin-blue or with thionin, and then by differentiating with 
Unna's glycerin-ether mixture. In specimens thus treated the 
mast cell granules are of a dark red color, while other basophile 
granules stain blue, so that the former must be regarded as hav- 
ing a modified basic reaction. They are stained reddish-violet 
with Ehrlich's acid dahlia solution, and deep blue with aqueous 
solution of methylene-blue. They are not stained by the tri- 
acid mixture, and appear as coarse, dull white spots through 
the cell body in films stained with this solution. The distinc- 
tive manner in which they react toward selective stains for 
mucin has been discovered by Harris, 1 who, in view of this 
fact, suggests that the term mucinoblast be applied to the mast 

The author questions the identity of these coarsely granular 
basophilic blood cells with the well-known mast cell of the tissues, 
although most hematologists consider them identical. Both, 
it is true, contain granules which tinctorially and morpholog- 
ically are identical, but it is obviously impossible to determine 
cell identity by criteria such as these. The mast cell of the 
tissues differs from that of the blood in having a nucleus which is 
smaller in relation to the size of the cell body, .more centrally 
situated, and richer in chromatin, hence being more deepjy and 
more unevenly stained. The "explosive" nature of the tissue 
mast cell is also usually more striking, for while cells with this 
tendency are met with not infrequently in the blood, they seem 
to be the rule rather than the exception in the tissues, large 
numbers of them consisting of a nuclear structure surrounded by 
dense clusters of granules, which are frequently drawn out in 
long tentacular extensions. In view of these differences it may be 
well to be more specific, by designating the mast cell found in the 
blood as the hemic mast cell. 

1 Phila. Med. Jour., 1900, vol. v, p. 757. 


This term has been applied by Capps 1 to a 
MONONUCLEAR form of leucocyte which he found in certain cases 
NEUTRO- of general paralysis of the insane, its appearance 
PHILES. in the blood having been noted after apoplecti- 
form attacks and preceding death. This cell is 
as large as, or larger than, the polynuclear neutrophile, contains a 
round or ovoid nucleus which is deeply stained by basic dyes, 
and has a protoplasm thickly sprinkled with fine neutrophile 
granules. Capps suggests that the cell may be a form of leuco- 
cyte of slightly more mature development than the large lympho- 
cyte, one in which the development of the granules has preceded 
the nuclear changes. The close resemblance of these cells to 
the smaller forms of myelocytes, however, makes it reasonable 
to class them as such. 

Ehrlich has described 2 as a "small neutro- 
NEUTROPHILIC philic pseudolymphocyte " a cell of the same 
PSEUDOLYM- size as that of the small lymphocyte, and char- 
PHOCYTES. acterized by a relatively large, round, intensely 
basic nucleus, surrounded by a narrow zone of 
protoplasm filled with neutrophile granules. This cell, it is 
maintained, is of very rare occurrence, having been found in the 
blood only in a case of hemorrhagic small-pox and in the exudate 
of a recent pleural effusion. Ehrlich differentiates it from a myelo- 
cyte by its small size, deeply staining nucleus, and scanty amount 
of protoplasm, but these points of distinction do not appear con- 
clusive, for many of the smaller, "dwarf" forms of myelocytes 
have similar characteristics. It does not appear unreasonable, 
therefore, to regard this cell as an exceedingly small form of 
myelocyte, in which the nucleus is relatively larger and richer in 
chromatin than is the rule in the larger, more typical varieties. 

These cells, first described by Tiirk 3 as "Rei- 
REIZUNGS- zungsformen " (or, literally, " stimulation forms "), 
FORMEN. are said to occur in the same pathological condi- 
tions in which myelocytes are found, but as yet 
their exact significance is undetermined. They may be found in 
any condition provoking decided anemia or leucocytosis and thus 
causing active stimulation of the bone marrow. The writer has 
seen such cells in the blood of the post-typhoid anemias of infancy, 
always in association with lymphocytosis. The size of the cell is 
usually midway between that of the small and large lymphocyte, 
more often approximating the size of the former. The cell con- 
tains a round nucleus, deficient in chromatin, often eccentrically 

1 Amer. Jour. Med. Sci., 1896, vol. cxi, p. 650. 

3 Loc. cit. 3 " Klinische Hamatologie," Vienna and Leipzig, 1904, p. 368. 



placed in the cell body, and reacting with moderate intensity to- 
ward the basic dyes. The protoplasm is non-granular, and stains 
purple with Wright's stain and intense brown with the triacid 
mixture. Ehrlich suggests that this cell may possibly represent 
an early stage of the erythroblast, but reasons for such an inference 
do not seem clear. 

The chief points of distinction between the different forms of 
leucocytes, as recognized in specimens stained with Wright's 
Romanowsky mixture, are tabulated below: 





Small lymphocyte. 

6 to 9 ft. 


Relatively small amount. 


Occasionally granular. 

Relatively large. 

Pale sky-blue. 

Dark blue or purple . 

Large mononuclear 

10 to 15 ft. 


Relatively large amount. 

leucocyte or large 

Round or ovoid. 

Often granular. 


Relatively small. 

Pale blue. 

Very pale blue. 

Transitional leuco- 

10 to 15 p. 


Relatively large amount. 


Indented, kidney- 

Often granular. 

shaped, or cres- 

Pale blue. 


Relatively small. 

Pale blue. 

Polynuclear neutro- 

7.5 to 12 p.. 

Polymorphous or 

Relatively large amount. 



Contains fine lilac or pink 

Relatively small. 

neutrophile granules. 

Moderately dark 

Relatively large amount. 




7.5 to 12 //. 

Polymorphous or 

Contains coarse rose-col- 


ored eosinophile gran- 

Relatively small. 


Pale blue. 


7-5 tO 12 p. 


Relatively large amount. 

Relatively small. 

Contains fine blue baso- 

Dull blue. 

phile granules. 


10 to 20 /'. 


Relatively large. or small 

Round or ovoid. 


Relatively large or 

Contains fine lilac or pink 


neutrophile granules. 

Very pale blue. 

Mast cell. 

7 tO 22 ft. 


Relatively small amount. 

Round, ovoid, or 

Contains coarse royal 

slightly lobulated. 

purple basophile gran- 

Relatively large. 


Very pale blue. 


6 to 15 ft. 


Relatively large amount. 



Relatively small. 

Intense lilac or purple. 

Deep blue. 


Two different views are current at the present 
ORIGIN AND time regarding the origin and development of the 

DEVELOP- leucocytes, the first being that of Ehrlich 1 and 

MENT. his followers, and the second that maintained by 

the Russian school, led by Uskow 2 and his pupils. 

According to Ehrlich's teachings, the small lymphocyte and 
its mother-cell, the large lymphocyte, are developed in the lym- 
phatic tissues in the various parts of the body, wherever such 
structures exist. The large mononuclear leucocytes and transi- 
tional forms are considered probably of myelogenous origin. 
The polynuclear neutrophiles are thought to develop exclusively 
in the bone marrow, the great majority being evolved from the 
neutrophilic myelocytes of this tissue, while a very limited num- 
ber perhaps arise from the non-granular large mononuclear cells. 
The eosinophiles develop from the eosinophilic myelocytes in the 
bone marrow, while the basophilic leucocytes similarly have their 
origin in basophilic marrow antecedents. Thus, it is maintained 
that all varieties of leucocytes may .be classed in two distinct 
groups which have separate origins, functions, and relations. 
The first group consists of the lymphocytes, large and small, 
which are produced solely by the lymphatic tissues; and the 
second group includes the mononuclear leucocytes and transi- 
tional forms, the polynuclear neutrophiles, the eosinophiles, and 
the basophiles, all of which cells are produced exclusively by the 
marrow. 3 Cellular reproduction, except in rare instances, does 
not take place in the circulating blood stream. Labbe 4 considers 
that no such distinction between lymphoid and myeloid cells is 
possible in early life, at which period he believes that all blood- 
forming tissues are capable of producing both hyaline and granular 
cells. But as adult life approaches the origin of the lymphocytes 
may be traced to the lymphatic tissue, and the birthplace of the 
granular leucocytes to the bone marrow. 

The scheme devised by the Russian school contends for the 
continuous evolution of the leucocyte from its earliest to its most 
mature stages. Accordingly, all varieties of the leucocyte, except 
the basophilic cells, of which no account apparently is taken, are 
but different developmental stages of one and the same cell. 
The youngest form of leucocyte, the small lymphocyte, originates 

1 Loc. cit. 

2 "The Blood as a Tissue," 1890 (Russian); also series of articles by Uskow's 
pupils in Arch. d. Soc. biol., St. Petersburg, 1893-97. 

8 Muir's article on the relations of the bone marrow to leucocyte formation 
(Jour. Path, and Bacteriol., 1901, vol. vii, p. 161) admirably discusses the natal 
differences of the lymphoid and myeloid cells of the blood. 

4 "Le Sang," Paris, 1902. 


in the lymph glands, the lymphocytic bone marrow, and the 
spleen, from which sources of origin it reaches the circulation. 
The small lymphocyte enlarges until it becomes identical in 
appearance with the cell recognized as the large lymphocyte, its 
nucleus at this period of its growth having become somewhat 
less intensely basic, although the basic affinity shown by the cell 
protoplasm is unaltered. The large lymphocyte in turn under- 
goes a simple increase in size, its nucleus meanwhile becoming 
progressively paler and its protoplasm more feebly basic, until 
it develops into the large mononuclear form. The nucleus of 
the latter now becomes indented and molded into a roughly 
crescentic figure, its nuclear and protoplasmic characteristics re- 
maining unchanged, and the so-called transitional form thus origi- 
nates a type of cell which is regarded as the immediate ante- 
cedent of the polynuclear neutrophile. During the next stage of 
development the size of the transitional cell decreases and the 
whole cell becomes ameboid; the nucleus becomes denser, more 
basic, and polymorphous or polynuclear; while the protoplasm 
loses the last trace of its basic tendency and becomes sprinkled 
with fine neutrophile granules, until finally the mature form of 
leucocyte, the polynuclear neutrophile, is fully developed. The 
final, or "over-ripe," stage of the leucocyte is represented by the 
eosinophile, which is thought to be derived from the polynuclear 
form by a transformation of the latter's neutrophile into 
eosinophile granules. It is maintained that these, transitions 
from one form of cell to the other occur partly in the circulat- 
ing blood and partly in the blood-forming tissues most largely 
in the latter. 

It is beyond the province of this book to discuss the merits 
and demerits of these two opposing views, but it may be remarked 
that Uskow's theory, which up to the advent of Ehrlich's observa- 
tions commanded general attention among hematologists, has 
been generally supplanted by the latter. The investigations of 
Ehrlich in this direction constitute the only dependable' means 
by which many of the pathological changes in the leucocytes 
may be explained, and his views may be accepted, on the whole, 
as accurate. 

Disintegration of the leucocytes occurs chiefly in the spleen 
and to a less extent in the liver. Bain's perfusion experiments, 1 
of repeatedly passing blood through these organs, indicate that 
the hemolytic action of the spleen is directed mainly toward the 
leucocytes and especially toward the polynuclear neutrophile 
cells. From Lewis' studies 2 it is to be inferred that the hemo- 

1 Jour. Physiol., 1903, vol. xxix, p. 352. * Ibid., 1902, vol. xxviii, p. 8. 



lymph glands have a similar function, by virtue of their phago- 
cytic endothelium. 

In a number of diseases associated with 
IODIN anemia and with bacterial or chemical toxemia 

REACTION, the protoplasm of the polynuclear neutrophile 
leucocytes, and rarely of the myelocytes and baso- 
philes, shows a more or less marked affinity for iodin, as shown by 
staining with a weak solution of this metal. Extracellular iodin- 
stained masses also are found in both normal and pathological 
blood, but they are without significance unless in association with 
leucocytes showing a similar reaction. Some of these extracellular 
iodophile masses are doubtless bits of protoplasm torn from the 
leucocytes; others may be blood plaques, while some are but 

Goldberger and Weiss 1 recommend the following reagent for' 
demonstrating the iodin reaction : 

Iodin i 

Potassium iodid 3 

Distilled water 100 

Mix and add sufficient gum arabic (about 50 parts) to make a 
syrupy mixture. 

With a camel's-hair brush a layer of this solution is painted 
over the surface of the dried, unfixed blood film, upon which it 
is allowed to act for from one to five minutes. The excess is 
then removed by blotting with a bit of filter-paper, and the speci- 
men is mounted in cedar oil. Or, as Wolff 2 advises, Zollikofer's 
method may be used: placing the fresh film for a few minutes 
in a stoppered bottle containing crystals of pure iodin. 

In films thus treated the iodin reaction is recognized by a 
slight or intense, diffuse brown coloring of the entire protoplasm, 
or by the presence throughout the protoplasm of numerous in- 
tensely stained, reddish-brown granules, the latter change being 
the more common. In normal blood the protoplasm of the leu- 
cocytes is stained a pale yellow and the nuclei remain almost 

The above reaction, known as iodophilia, is constant in all 
purulent conditions, persisting as long as the suppurative focus 
exists, its intensity appearing to bear no parallelism to the extent 
of the pus collection. It signifies simply a general toxemia severe 
enough to cause a form of leucocyte degeneration, and, like 
its corollary, leucocytosis, is merely a symptomatic sign. Kaminer 3 
believes that the reaction depends upon three factors for its pro- 

1 Wien. klin. Wochenschr., 1897, vol. x, p. 601. 

2 Zeitschr. f. klin. Med., 1904, vol. li, p. 407. 

3 Deutsch. med. Wochenschr., 1899, vol. xxv, p. 235; also Berlin, klin. 
Wochenschr., 1899, vol. xxxvi, p. 119. 


duction, pyrexia, leucocytosis, and toxemia, and that it is 
caused by the action of some unknown chemotactic substance. 
He groups the iodophile cells, according to their grade of reac- 
tion, into three stages: diffuse brownish staining, circumscribed 
granulation, and complete metamorphosis. 

The reaction is absent in pure tuberculous abscesses. It is 
present with great constancy in puerperal sepsis and in other 
forms of septicemia, frequently in pneumonia, pulmonary tuber- 
culosis, and malignant disease, and occasionally in marked cachexias. 
Hofbauer l found iodophile granules in the leucocytes in all cases of 
pernicious anemia, their number being greatest in the gravest cases ; 
they were also present in severe forms of secondary anemia and 
in leukemia, but were absent in chlorosis and in pseudoleukemia. 
This author also observed numerous iodin-stained extracellular 
masses in a case of purpura hamorrhagica. Dunn, 2 study- 
ing the iodin reaction in children, found it invariably present in 
croupous and catarrhal pneumonia, influenza, cerebrospinal men- 
ingitis, empyema, and miscellaneous purulent conditions; the reac- 
tion was generally present in enteric fever and in acute miliary 
tuberculosis. Cabot and Locke 3 obtained uniformly positive 
reactions in septicemia, pneumonia, empyema, and suppurative 
appendicitis; in serous pleural effusions and in catarrhal appendi- 
citis the test was negative. In about one-half of the cases of 
enteric jever examined by these writers the test was positive, 
usually only in those complicated by hemorrhage, perforation, 
furunculosis, or lung lesions. These studies have been recently 
substantiated by Gulland. 4 

Experimentally, iodophilia has been caused by Spezia 5 by the 
subcutaneous injection of peptone, glucose, and olive oil, and the 
sign has also been noted by him during the digestive leucocytosis 
following a hearty meal. Kaminer 6 produced iodophilia in ani- 
mals by injecting cultures and toxins of the pneumococcus, the 
Klebs-Loffler bacillus, the typhoid bacillus, and various pyogefiic 
bacteria; he failed to cause it by injecting tetanus toxin. 

The practical value of this test is considerable if it is properly 
interpreted merely as a symptom. Its constancy in purulent condi- 
tions, however slight in extent the focus of pus, is useful in diagnos- 
ing a deep-seated abscess, if other causes which also may give rise 
to the reaction can be ruled out. The sign is also a valuable aid in 

1 Centralbl. f. inn. Med., 1900, vol. xxi, p. 153. 

2 Boston Med. and Surg. Jour., 1903, vol. cxlix, p. 511. 
8 Jour. Med. Research, 1902, vol. vii, p. 43. 

4 Brit. Med. Jour., '1904, vol. i, p. 880. 

5 Lancet, 1903, vol. i, pp. 655 and 1444. 
* Brit. Med. Jour., 1902, vol. i, p. 1049. 


differentiating serous from purulent effusions and inflammations 
serous pleurisy and empyema; catarrhal appendicitis and appendic- 
ular abscess. It is also useful, according to Sorochowitsch, 1 in dif- 
ferentiating gonorrheal and rheumatic arthritis, since positive reac- 
tions occur in the former and negative in the latter. The absence of 
the iodin reaction in pure tuberculous abscess and its presence in all 
other forms of abscess may aid in distinguishing between the two, and 
of ascertaining whether a mixed infection exists. The persistence 
of iodophilia for forty- eight hours or so after a pneumonia crisis 
and after the incision of a pus cavity suggests, in the first instance, 
delayed resolution or some other post-pneumonic complication, 
and, in the second, imperfect drainage. Although the writer has 
followed out rather at length the suggestion made by Hofbauer, 
that the intensity of the reaction serves as an index to the severity 
of an anemia, the results from this study have not shown the relia- 
bility of such a presumption. 

Neusser, 2 in 1894, described certain basic 
PERINUCLEAR granules about the nuclei of the leucocytes which 
BASOPHILIA. he regarded as pathognomonic of the uric acid 
diathesis, asserting that this so-called "perinu- 
clear basophilia" could be demonstrated constantly in gout, 
lithiasis, rheumatism, leukemia, and a number of other diseases. 
These statements were soon corroborated by Kolisch, 3 and for a 
time enjoyed more or less general credence. The later researches 
of Futcher 4 and of Simon, 5 however, have entirely disproved the 
claims of Neusser and his school, for these investigators, working 
independently, have proved that perinuclear basophilia is not 
only quite uncharacteristic of the uric acid diathesis, but that it 
can be constantly demonstrated in every sort of blood, whether 
from healthy or from diseased persons. It is now clear that 
Neusser's granules are simply artefacts, due to some slip in the 
technic of staining. Ehrlich 6 believes that their presence is but 
rarely noted if perfectly pure crystalline dyes are used in pre- 
paring the stain. 


Leucocytosis may be described as an increase above the normal 
standard in the number of leucocytes in the peripheral blood, this 

1 Zeitschr. f. klin. Med., 1904, vol. li, p. 245. 

2 Wien. klin. Wochenschr., 1894, vol. vii, p. 71. 

3 Ibid., 1895, vol. viii, p. 797. 

4 Johns Hopkins Hosp. Bull., 1897, vol. viii, p. 85. 

5 Amer. Jour. Med. Sci., 1899, vol. cxvii, p. 139. 8 Loc. cil 


( Triacid Stain.) 

The blood field from a case of croupous pneumonia. The leucocytes are all of the polynuclear 
neutrophile type. The erythrocytes show no deformity, and stain a normal orange 

Contrast this illustration with leukemia, Plates IV and V. 

(E. F. FABER,/fC.) 


change either (a) involving both an absolute and a relative increase 
in the polynuclear neutrophile cells with a consequent relative dimi- 
nution in the proportion of mononuclear non-granular forms, or (6) 
affecting all varieties of leucocytes alike. 

A leucocytosis of the first kind, also termed a polynuclear neu- 
trophile leucocytosis, is by far the more common of the two types : 
it may be symptomatic either of pathological or of physiological 
conditions, being found almost invariably in the former and fre- 
quently in the latter. A leucocytosis of the second kind, or a 
general increase unattended by any disturbance in the normal 
relative proportions of the different forms of cells, is compara- 
tively rare: it is more frequently dependent for its production 
upon physiological than upon pathological factors, but it may 
occur under either of these circumstances. 

From these facts it is obvious that simply an increase in the 
total number of leucocytes, without regard to the differential 
changes involved, does not of necessity constitute a leucocytosis. 
Nor is it possible to recognize the condition with certainty by 
any such criterion as a deviation from the ratio of red to white 
cells maintained in health. To state that a patient's blood con- 
tains, say, 50,000 leucocytes to the suggests both leuco- 
cytosis and leukemia, but to add to such a statement the fact 
that of these 50,000 leucocytes 90 per cent, are of the polynuclear 
neutrophile variety, at once stamps the condition as a genuine leu- 
cocytosis. In order, therefore, to distinguish leucocytosis with 
absolute certainty the character of the leucocytes involved in the 
increase must be determined by a differential count of the s'tained 
specimen of blood. (Plate III.) 

Leucocytosis may be of a more or less transient character, or 
may persist for a long period, its duration being dependent upon 
the nature of the underlying cause. In acute diseases it is usually 
a temporary condition, but in long-continued affections it is .pro- 
longed in relation to the chronicity of the lesion by which the 
increase is excited. 

For clinical purposes all forms of leucocytosis may be classed 
under two main groups, physiological and pathological, these be- 
ing further divided as follows: 

Physiological Leucocytosis. 

1. Leucocytosis of the new-born. 

2. Digestion leucocytosis. 

3. Leucocytosis of pregnancy and parturition. 

4. Leucocytosis due to thermal and mechanical influences. 

5. Terminal leucocytosis. 


Pathological Leucocytosis. 

1. Inflammatory and infectious leucocytosis. 

2. Leucocytosis of malignant disease. 

3. Post-hemorrhagic leucocytosis. 

4. Toxic leucocytosis. 

5. Experimental leucocytosis. 


The leucocytoses associated with a number of 
CHARACTER, purely physiological conditions are generally of 
brief duration, and, as a rule, involve a moderate 
increase in the white corpuscles, the gain in many instances being 
trifling and never excessive. As noted in a preceding paragraph, 
the increase sometimes affects equally all forms of leucocytes, so 
that, although the total number of cells is higher than the normal 
standard, the relative percentages of the different varieties remain 
in the ratio observed in normal blood. In other instances the 
gain is due to an absolute and relative increase in the polynuclear 
neutrophiles, with a consequent decrease in the percentage of 
non-granular, mononuclear forms. 

The increase of leucocytes under such condi- 
CAUSAL tions is to be regarded usually as a physical 
FACTORS. phenomenon depending upon temporary concen- 
tration of the blood or upon an unequal distri- 
bution of the cells in favor of the peripheral vessels. Evidence 
is wholly lacking to show that it is caused by an actual overpro- 
duction of leucocytes by the blood-forming organs, thus produc- 
ing a general increase through all parts of the body. It is, there- 
fore, reasonable to believe that the high leucocyte counts may 
be accounted for by such factors as decrease in the total volume 
of blood plasma, and the transference of cells from the vessels 
of the deeper tissues to those of the superficial parts of the body, 
i. Leucocytosis of the New-born. The blood of the infant at 
birth contains two or three times the number of leucocytes found 
in the normal adult, the count usually ranging from 15,000 to 
20,000 or higher during the first forty-eight hours of life. After 
this time the number of cells gradually decreases until, by the 
end of the first or second week, it has fallen to an average of from 
10,000 to 15,000, which figures may be considered normal for 
children under one year of age. In ten babies examined by War- 
field 1 these averages were found: first day, 26,090; third day, 

1 Amer. Med., 1902, vol. iv, p. 457. 


13,270; and eleventh day, 15,740 leucocytes' per In all 
of these cases it was found that the younger the infant, the 
higher the count, which during the first five hours after birth 
commonly ranged between 28,000 and 34,000. Gundobin 1 and 
Carstan jen 2 have determined, by a series of differential counts, that 
the increase is due chiefly to an excessive gain in the polynuclear 
neutrophiles, the proportion of these cells during the first ten days 
after birth averaging from 60 to 70 per cent, of all forms of leu- 
cocytes. The extent of this increase becomes apparent when one 
recalls the fact that in the infant the relative proportion of these 
cells to the other varieties is usually not more than 40 per cent. 
Japha 3 considers 42 per cent, the average of the polynuclear neutro- 
philes in infants from several weeks to twelve months of age. 
By the tenth day this polynuclear increase usually subsides, and 
the percentage of mononuclear forms rises to the figure normal 
at this period of life. (See Section VI.) In prematurely-born 
infants a similar increase in the number of leucocytes is present, 
but the mononuclear forms rise to their normal percentage more 
rapidly than in the full-term baby; thus, in the case of an eight- 
months' child, examined by Whitney and Wentworth, 4 the large 
and small lymphocytes, which averaged together but 26 per cent, 
at birth, rose to 80 per cent, by the fourth day, remaining at practi- 
cally this figure through subsequent counts. 

The leucocytosis of the new-born is probably attributable 
partly to concentration of the blood by venous stasis 'and by the 
drain on the body-fluids incident to the early days of life, and partly 
to the influence of digestion leucocytosis, which is especially active 
at this period. Schiff 5 first noted the influence of the latter factor, 
and particularly drew attention to the marked gain in cells after 
the baby's initial feeding. 

2. Digestion Leucocytosis. Within an hour after taking food 
an appreciable increase in the number of leucocytes may be ob- 
served in the great majority of healthy individuals, the/count 
reaching its maximum within from two to four hours after the 
meal, and then gradually declining. Rieder 6 estimates the aver- 
age increase at about 33 per cent, in excess of the normal figure. 
Meals rich in albuminoids are followed by a more marked increase 
than those consisting chiefly of vegetable articles of diet. In in- 
dividuals whose process of digestion is slow from any cause the 
appearance of the leucocytosis is also delayed. The following two 

1 Jahrb. f. Kinderheilk., 1893, vol. xxxv, p. 187. 

1 Ibid., 1900, vol. Hi, p. 333. 3 Ibid., 1901, vol. liii, p. 179. 

4 Cited by Rotch, "Pediatrics," Philadelphia, 1896, p. 348. 

s Zeitschr. f. Heilk., 1890, vol. xi, p. 30. 

* " Beitrage zur Kenntniss der Leukocytose," Leipsic, 1892. 



instances, taken from von Limbeck, 1 illustrate the development of 
the leucocytosis in the normal adult: 





11.15 A - M - 2 


11.30 A. M. 2 


12.15 p - M - 


12.30 P. M. 


1.15 P. M. 


1.30 P. M. 


3.15 P. M. 


2.30 P. M. 


5 . 1 5 P.M. 


3.30 P. M. 


7.15 P.M. 


6.00 P. M. 


The gain is due usually to a predominance of polynuclear neu- 
trophile forms, with a consequent relative diminution in large and 
small lymphocytes; but in some instances the differential count 
remains normal, all forms of cells sharing equally in the increase. 

Digestion leucocytosis is not invariably present even in those 
who apparently enjoy perfect health, its absence in such instances 
remaining entirely unexplained. It is also absent occasionally in 
chronic constipation, frequently in chronic gastric catarrh and 
anemia, and is found in only a small proportion of cases of gastric 
carcinoma. Other lesions of the gastro-intestinal tract and dis- 
eases characterized by high-grade anemia and by marked de- 
bility may greatly delay or even entirely prevent the increase. 
Rieder 3 is authority for the statement that digestion leucocytosis 
does not occur during pregnancy, and Bohland 4 finds that it fails 
to develop during the administration of tannic acid. Digestion 
leucocytosis ordinarily does not occur when the leucocytes are 
already increased by some pathological factor. 

In children, especially in young breast-fed infants, the increase 
is very decided; in the new-born counts of from 30,000 to 35,000 
may follow the first few feedings. (See Section VI.) The leucocy- 
tosis is also marked after fasting and in diabetics. 

The factors of digestive leucocytosis are, theoretically, two- 
fold : the chemotactic action of the absorbed albumins, which calls 
from the bone marrow an excess of polynuclear neutrophiles ; 
and the increased lymph flow from the thoracic duct, which 
accounts for the lymphocyte gain. The older writers, notably Poll 
and Hofmeister, 5 thought the latter due to hyperactivity of the 
gastro-intestinal lymphatic tissue, but this view has been con- 
troverted by the careful work of Goodall, Gulland, and Paton, 6 who 
failed to find the slightest sign of adenoid proliferation in the walls 

1 Loc. cit. * Meal of nitrogenous and farinaceous food. 

3 Loc. cit. 4 Centralbl. f. inn. Med., 1899, vol. xx, p. 361. 

8 Cited by von Limbeck, loc. cit. 
8 Jour. Physiol., 1903, vol. xxx, p. i. 


of the gut, and who, furthermore, found the quantitative and 
qualitative count of leucocytes identical in the mesenteric veins and 
arteries and in the general circulation. These investigators also 
found that in animals digestion leucocytosis is unaffected by 
removal of the spleen. 

3. Leucocytosis of Pregnancy and Parturition. In the majority 
of primiparae a moderate leucocytosis, not usually involving an in- 
crease in excess of double the normal count, is observed during 
the later months of pregnancy. The increase is less constant 
and much less marked in multipart, occurring in a smaller per- 
centage of the latter, and amounting to a cellular gain of about 
one-sixth the original count on the average. The maximum 
number of cells is usually found immediately before and after de- 
livery, at which time the number of leucocytes commonly rises 
to about 15,000 per During convalescence the leucocytosis 
gradually declines, and disappears before the end of the first week 
after delivery, in uncomplicated cases. As a general rule, in both 
primiparae and in multiparae the degree of increase is more decided 
in young women than in those of middle age. It is also marked 
in the late rather than in the early stages of gestation and of labor. 
The leucocytosis of pregnancy is best explained by the unusually 
active metabolism of this physiological period and by the hyper- 
activity of the pelvic lymph glands. 

The careful blood studies by Hibbard and White 1 in 55 preg- 
nant women (33 primiparae and 22 multiparae) furnish the most 
reliable data concerning the leucocytosis of this condition. These 
authors found that leucocytosis occurred before delivery in 84 
per cent, of primiparae and in 75 per cent, of multiparae, the aver- 
age counts in 32 of the former being 15,021 (50 per cent, above 
normal) and in 20 of the latter 11,700 (17 per cent, above nor- 
mal). In normal labor the number of leucocytes fell rapidly after 
delivery, gradually reached the normal standard by the fourth or 
fifth day, and then again slowly rose until the seventh day, when 
a decline to normal was again observed. In 37 cases at full term 
(including both primiparae and multiparae) examined by J. Hen- 
derson 2 the leucocyte count averaged 21,365, falling by the tenth 
day after labor to an average of 12,327. 

After version, forceps application, and other operations the 
leucocytosis persists for a longer period. Post-partum hemorrhage 
and lacerations of the genital tract may also prolong the leucocy- 
tosis. Zangmeister and Wagner 3 found that post-partum pyrexia 

1 Jour. Exper. Med., 1898, vol. iii, p. 639. 

2 Amer. Jour. Obstet., 1902, vol. xlv, p. 745. 

8 Deutsch. med. Wochenschr., 1903, vol. xxviii, p. 549. 


with fetid lochia is not a factor of any decided leucocytosis, such 
as that supervening in genuine sepsis. Pray 1 found in the blood 
of a woman delivered by Cesarean section the same leucocyte 
changes which accompany a normal labor. 

Differential counts in 19 cases of Hubbard and White's series 
showed that the leucocytosis was of the polynuclear neutrophile 
type, a marked relative and absolute increase in these cells being 
constantly present; as a rule, their percentage was from 85 to 95 of 
all forms of leucocytes, usually being higher the higher the leu- 
cocytosis. Henderson's counts showed similar differential changes. 
These results are unlike those obtained by Rieder 2 and by Bjork- 
man, 3 the former having stated that the various forms of cells 
remain practically normal, while the latter attributes the increase 
to a predominance of mononuclear elements. 

Lactation of itself has no appreciable effect upon the leuco- 
cytes, so that a leucocytosis occurring in a nursing woman should 
be attributed to inflammatory conditions of the breast or nipple 
even a mild mastitis or a slight irritation of the nipple may be 
capable of causing a prompt leucocytosis. 

The number of leucocytes is somewhat in excess of normal 
for a few days preceding and during menstruation in the majority 
of healthy women, according to the investigations of Sfameni, 4 
but the increase scarcely ever reaches a degree which may be 
regarded as a genuine leucocytosis. 

4. Leucocytosis Due to Thermal and Mechanical Influences. A 
transient increase in the number of leucocytes of the peripheral 
blood, not involving a disturbance of the normal ratio between the 
different forms of cells, is produced by active local or general 
muscular exercise;* by brief exposure to atmospheric cold; 6 by 
cold baths, either local or general; 7 and by the application of 
electricity* and of massage? The number of leucocytes is also 
increased by the effect of prolonged dry or moist heat. 9 Hot 
tubbing likewise causes considerable leucocytosis, 10 as does free 
sweating, whether natural or induced. It occurred in 24 of 29 
instances studied by Hannes, 11 the increase amounting to between 

1 Amer. Gyn., 1902, vol. i, p. 337. 2 Loc. cit. 

3 Amer. Medico-Surg. Bull., 1894, vol. vii, pp. 17 and 79. * Loc. cit. 

s Oliver, loc. cit.; Larrabee, Jour. Med. Research, 1902, vol. ii, p. 76; Schultz, 
Deutsch. Arch. f. klin. Med., 1893, vol. li, p. 234; Willebrand, Skandin. Arch. 
f. klin. Med., 1903, vol. xiv, p. 176. 'Oliver, loc. cit. 

7 Winternitz, Centralbl. f. klin. Med., 1893, vol. xiv, p. 1017; Thayer, 
Johns Hopkins Hosp. Bull., 1893, vol. iv, p. 37. 

8 J. K. Mitchell, Amer. Jour. Med. Sci., 1894, vol. cvii, p. 502; Ekgren, 
Deutsch. med. Wochenschr., 1902, vol. xxviii, p. 519. 

9 Friedlander, Cong. f. inn. Med., Berlin, 1897. 

10 Knapfelmacher, Wien. klin. Wochenschr., 1893, vol. vi, p. 810. 

11 Centralbl. f. inn. Med., 1901, vol. xxii, p. 823. 


3000 and 5000 cells per Checking the perspiration resulted 
in a fall of the leucocytes to normal within about half an hour. 

The increase under these circumstances is generally attributed 
to blood concentration due to the influence of increased vaso- 
motor tension, whereby the liquid elements of the blood arc tem- 
porarily decreased, and, in addition, many of the cells lodged in 
the deeper tissues of the body are swept into the peripheral cir- 
culation. As a rule, all varieties of leucocytes share equally in 
the process, no single form being unduly increased at the ex- 
pense of the others. In the case of long-distance runners, how- 
ever, a very decided polynuclear neutrophile increase has been 
found. 1 

5. Terminal Leucocytosis. Terminal or preagonal leucocyto- 
sis is the term applied to an increase in the number of leucocytes 
of the peripheral circulation frequently observed just before death. 
It occurs during the terminal stages of a number of different 
diseases, and is especially marked in those conditions in which 
death comes slowly, being ushered in by a more or less mori- 
bund state of the patient, lasting for a considerable length of time. 
The increase is usually moderate, and the counts do not often 
exceed 20,000 or 30,000 per, except in those cases in 
which decided circulatory embarrassment has existed for some 
time. Most commonly the blood picture is one of ordinary 
polynuclear neutrophile leucocytosis, although occasionally the 
large and small lymphocytes show disproportionately high per- 
centages, and still more rarely all forms of cells may be in- 
creased equally. The presence of myelocytes, in small* numbers 
is also common, especially when the leucocyte count is high. 

In pernicious anemia the increase may be so great as to simu- 
late lymphatic leukemia, according to Cabot, 2 who found the fol- 
lowing blood changes on the day of death in this disease : ratio of 
white to red corpuscles, i to 15; and a differential count of 91.7 
per cent, of lymphocytes, 7.7 per cent, of polynuclear "neutro- 
philes, and 0.5 per cent, of eosinophiles. Four megalobiasts to 
TOCO leucocytes were also found. 

The following data were obtained by the writer in a case of 
pernicious anemia eighteen hours before death: hemoglobin, 12 
percent.; erythrocytes, 622, 500 per; leucocytes, 18,600 per The differential count of 1000 leucocytes showed: lymph- 
ocytes, 46 per cent.; polynuclear neutrophiles, 49.7 per cent.; 
eosinophiles, 2.3 per cent.; myelocytes, 1.6 per cent.; and mast 
cells, 0.4 per cent. Megalobiasts outnumbered normoblasts 3 to i, 

1 Cabot, Blake, and Hubbard, Annals of Surg., 1901, vol. xxxiv, p. 372. 
' Loc. tit. 


24 of the former being found in the count of 1000 leucocytes. 
The number of leucocytes in four previous counts having ranged 
from looo to 2400 per, and the proportion of lymphocytes 
from 42 to 48 per cent., this case illustrates the occurrence of a 
terminal leucocytosis without a notable change in the relative 
percentage of different forms peculiar to the case in question. 

The principal cause of this form of leucocytosis is thought to 
be peripheral stasis dependent upon failure of circulatory com- 
pensation, but in many instances there is good reason to believe 
that terminal infections also act as the causal factors. 


Increase in the number of leucocytes, involving 
OCCURRENCE, chiefly the polynuclear neutrophile cells in the 
great majority of instances, is associated with a 
wide variety of pathological conditions, mainly inflammatory, in- 
fectious, and toxic in character, and in such conditions the under- 
lying cause of the phenomenon is radically different from that 
which determines the increase in physiological leucocytosis. 
Prominent examples of pathological lesions in which leucocytoses 
of this character are observed are pneumonia, diphtheria, scar- 
let fever, erysipelas, rheumatic fever, variola, and various septic 
processes. Enteric fever, paratyphoid fever, tuberculosis, typhus 
fever, Malta fever, the malarial fevers, influenza, measles, and 
rotheln are notable exceptions, for in these infections leucocy- 
tosis occurs only as the result of some complication. 

The extent of the leucocytosis, inasmuch as it 
DEGREE OF depends both upon the nature of the exciting cause 
INCREASE, and upon the individual's reactive powers, varies 
within wide limits in different cases. It is safe to 
state, however, that in the great majority of instances the number of 
leucocytes is rather below than above 20,000 to the, counts 
in excess of this figure being noted in only about one-fourth of 
the cases in which the leucocytes exceed the normal limits of 
health. A count of 25,000 cells per may be regarded as 
a decided leucocytosis, while an increase of from 40,000 to 50,000 
is of extremely rare occurrence. In an analysis of 100 consecu- 
tive counts made by the writer in pathological conditions, in which 
the number of leucocytes reached or exceeded 10,000 per, 
it was determined that the counts were below 20,000 in 65 per 
cent, of cases, and between 20,000 and 30,000 in 28 per cent.; in 
4 per cent, the increase was between 30,000 and 40,000; in 2 per 
cent., between 40,000 and 50,000; and in only one per cent, did it 


exceed 50,000. Judging from these figures, which, it should be 
remembered, are applicable only to the average case, it appears to 
be the rule that in most leucocytoses the increase amounts to 
about double the maximum normal number. 

With rare exceptions the increase affects 
DIFFERENTIAL chiefly the polynuclear neutrophile cells, which 

CHANGES. commonly constitute at least 85 per cent, of the 
different forms of leucocytes. In many instances 
the percentage is much higher, as, for example, in a case of sup- 
purative meningitis, reported by Stengel, 1 in which a differential 
count showed 99.5 per cent, of this variety of cells. The excep- 
tional cases in which these disproportionately high percentages of 
polynuclear neutrophiles are sometimes wanting are encountered 
in the leucocytoses of malignant disease, after hemorrhage, in the 
moribund, and in children. The relative lymphocytosis which is 
occasionally observed under these circumstances is considered in 
connection with these conditions. Coincidentally with the increase 
in polynuclear forms there is a marked decrease in the relative 
percentages of large and small lymphocytes and of eosinophiles, 
the latter variety of cells sometimes entirely disappearing from 
the blood. In cases in which the increase is marked, small num- 
bers of myelocytes usually may be observed, together with an 
occasional cell whose characteristics at once suggest a stage of 
development intermediate between that of the myelocyte and the 
typical polynuclear neutrophile. 

Several times the writer has found in typical infectious leuco- 
cytoses practically normal differential counts, notwithstanding 
the high total estimates. Still rarer are those instances in which 
an inflammatory or infectious lesion excites a high polynuclear 
neutrophile percentage with no total increase in the leucocytes. 
The latter blood change has been, on insufficient grounds, in- 
terpreted in the same light as a frank general increase. 

The exact manner in which pathological leu- 
CAUSAL cocytosis arises is a question about which many 

FACTORS. conflicting views are held by different authorities, 
but the general trend of opinion at the present 
time attributes the increase chiefly to the influence of chemotaxis. 
According to the chemotactic theory of leucocytosis, the pres- 
ence in the blood of certain chemical substances, produced by in- 
fective principles, is capable of exerting both an attractive and a 
repellent influence upon the ameboid leucocytes. If the collec- 
tions of cells are attracted by such substances, the phenomenon is 
known as positive chemotaxis, but if, on the other hand, they are 

1 Loc. cit. 


repelled, the condition is termed negative chemotaxis. This mass- 
ing and repulsion of the leucocytes may be caused by various 
agents by thermal and mechanical irritants, by bits of necrotic 
tissue which have gained entrance to the circulation, and espe- 
cially by the presence in the blood of bacteria or of their meta- 
bolic products. Thermotaxis, or attraction by heat, may also 
attract the leucocytes to an inflamed area. Mendelson 1 has shown 
that a local temperature of 39 C. is most active in provoking such 
a massing of the cells. In the light of our present knowledge it 
appears that the different varieties of ameboid leucocytes respond 
to different kinds of chemotactic influences, as an instance of which 
the behavior of the neutrophiles and eosinophiles to this sort of 
stimulus may be cited. Certain substances, which for one of 
these groups of cells are either positively or negatively chemo- 
tactic, are, as a rule, indifferent to the other group, and some- 
times even antagonistic, for substances which serve to attract one 
group either fail to influence or in fact repel the other. Clin- 
ically, this theory seems to find corroboration, for in the great 
majority of instances an increase in either variety of these cells is 
associated with a constant decrease in the other. In infections with 
certain animal parasites this general rule does not apply notably in 
trichiniasis, in ankylostomiasis, and in filariasis, in which it is appar- 
ent that substances chemotactically active for both neutrophile and 
eosinophile cells are at work. Ehrlich 2 has shown that the mast cells 
are wholly uninfluenced by those substances which exert a strong 
chemotactic influence upon the neutrophiles and eosinophiles. 

The intense cellular activity excited by the en- 
FUNCTIONS. trance of bacteria into the organism indicates an 
attempt on the part of the leucocytes to destroy 
the invading principle and to counteract its noxious influences. 
In this endeavor it is probable that in a restricted sense Metsch- 
nikoff's hypothesis holds true, and that the immense numbers of 
phagocytic leucocytes which crowd the blood stream mechan- 
ically engulf and destroy many of the invading micro-organisms. 
But of still greater significance is the faculty which the leuco- 
cytes possess of producing, both by secretion and by cellular 
disintegration, certain chemical substances (alexins) acting either 
as directly bactericidal or as antitoxic agents. The researches of 
Buchner, 3 Lowy and Richter, 4 Goldscheider and Jacob, 5 and others 

1 Russkiy Vrach, 1903, vol. ii, p. 4; abst. in Phila. Med. Jour., 1903, vol. xi, 

P- 78S- 

2 Loc. cit. 3 Arch. f. Hyg., 1890, vol. xvii, p. 112. 

4 Deutsch. med. Wochenschr., 1895, vol. xxix, p. 240; also Virchow's Arch., 
1898, vol. cli, p. 220. 

5 Zeitschr. f. klin. Med., 1894, vol. xxv, p. 373. 


tend to show that such substances either actually destroy the in- 
fecting micro-organisms, or at least antidote and render innocuous 
their poisonous products. This joint process of phagocytosis and 
bactericidal action is most intensely developed at the period of 
maximum leucocytosis, according to the statements of Gabri- 
tschewsky. 1 

Alexin is an unstable nucleo-proteid, acting as an enzyme and 
corresponding to the complement of Ehrlich. It is, according to 
Metschnikoff, a product of the leucocytes, and does not exist free 
in the blood plasma. In the process of bacteriolysis the alexin's 
activity depends upon the intermediation of the amboceptor or 
immune body, according to the hypothesis of immunity elaborated 
by Ehrlich (p. 151). The French school holds that phagocy- 
tosis is excited by the action of the amboceptor, the chief source 
of which is also the leucocytes. Metschnikoff believes that the 
functions of the different phagocytic cells in immunity are dis- 
tinctly dissimilar, the action of the polynuclear cells (or micro- 
phages) being simply bacteriolytic, while that of the large lympho- 
cytes (or macro phages) is solely hemolytic. 

It has been suggested by Wright and Douglas 2 that phago- 
cytosis is materially aided by the body fluids, which may so influ- 
ence invading bacteria as to make them easy prey for the phago- 
cytic leucocytes. They attribute this effect to the presence in the 
blood serum of an unknown body, "opsonin," which is thought 
to develop as immunity is established. In a number* of patients 
suffering from furunculosis Wright succeeded, by treating them 
with a sterile antistaphylococcus vaccine, in increasing .the phago- 
cytic power of the blood which before this treatment was dis- 
tinctly below the normal. 

In experimental leucocytosis, caused by the 
HYPOLEUCOCY- injection of such irritants as bacteria and bacterial 
TOSIS AND HY- products, organic extracts, various albumins, and 
PERLEUCOCY- even by simple trauma, it has been found that the 
TOSIS. first effect of the irritant is to cause a rapid, tran- 
sitory diminution in the number of leucocytes in 
the peripheral blood, known as hypoleucocytosis, this decrease 
being succeeded in turn by an increase of these cells in excess of 
the normal standard, termed hyperleucocytosis. Frequently in 
simple traumatic leucocytoses after the disappearance of the stage 
of hyperleucocytosis, the duration of which is variable, Sher- 
rington 3 was able to distinguish a secondary stage of hypoleuco- 

1 Centralbl. f. Bakt. u. Parasit., 1898, vol. xxiii, p. 365. 
1 Prpc. Roy. Soc., London, 1903, vol. Ixxii, p. 357. 
3 Ibid., London, 1893, vol. Iv, p. 161. 


cytosis during which the leucocyte count again fell below the 

Within certain limits the extent of this preliminary decrease 
and of the subsequent increase varies directly in accordance with 
the intensity of the irritant. If the irritant is slight, the repellent 
influence is feeble, and the consequent cellular increase is in- 
conspicuous in fact, it is the opinion of many that in such in- 
stances there may be merely a local accumulation of leucocytes 
at the site of the injection, without any real increase in the whole 
mass of cells. If the effects of the irritant are severe, both the 
repellent and the attractive stages are promptly excited and 
markedly developed, and a general increase in the number of 
leucocytes through the whole circulatory system promptly re- 
sults. If, on the contrary, the effects of the irritant prove to be 
too intense, the organism suffers a depression so profound that 
reaction is stifled, and leucocytosis docs not develop. It some- 
times happens that the attractive influences of the chemotactic 
principle predominate over its repellent action, in which case the 
stage of hyperlcucocytosis may develop without the initial stage 
of hypolcucocytosis. Clinically, the preliminary decrease is prac- 
tically never observed, perhaps partly for the reason last given, 
but also in a large measure because the repellent action of the 
irritant has passed off by the time the disease has developed into 
a clinical picture. In artificially excited leucocytoses, however, 
its appearance is quite constant, for under such circumstances the 
irritant is introduced into the organism suddenly and in a rela- 
tively massive dose, thus producing a decidedly repellent in- 

The initial stage of decrease was termed the leucopenic phase 
by Lowit, 1 who attributed the change to an actual destruction of 
the leucocytes, or a leucocytolysis. The subsequent increase he 
spoke of as the leucocytic phase, maintaining that for the produc- 
tion of the latter the preliminary development of the former was 
in some unexplained manner essential. The work of Goldschei- 
der and Jacob 2 definitely proved the error of Lowit's leucocyto- 
lytic hypothesis, and demonstrated the fact that the leucopenia 
was dependent purely upon an altered distribution of the cells 
in favor of the vessels of the deeper circulation. Extensive 
investigations carried on by these authors showed that at the 
time a decided diminution occurred in the number of leucocytes 
of the peripheral blood there was a simultaneous increase of 
these cells in the capillaries of the lungs and other internal organs. 
Furthermore, it was also shown that in some instances a marked 

1 "Studien z. Physiol. u. Pathol. d. Blutes," Jena, 1892. 2 Loc. cit. 


leucocytosis may occur without the initial decrease, this being 
the case after the injection of such substances as the glycerin 
extract of spleen. From these experiments it seems reasonable 
to attribute the initial stage of decrease to a repellent action of 
the irritant, and to infer that the stage of hyperleucocytosis is 
due to an active stimulation of the hemogenic organs which results 
certainly in an increased cellular output from, and probably in 
an increased cellular proliferation in, this situation. Muir's 
investigations 1 tend to strengthen this belief, and to throw 
additional light on the phenomenon of pathological leucocytosis. 
This author found that in experimental leucocytosis in animals, 
produced by the injection of pathogenic bacteria, changes occurred 
in the bone marrow, consisting of absorption of the marrow fat, 
together with a corresponding hyperplasia of the cells from which 
he believes the leucocytes originate, many of these cells under- 
going rapid multiplication by mitosis. In inflammatory leuco- 
cytosis Muir found the following suggestive changes : first, a local 
increase in the polynuclear neutrophile cells; second, an increase 
of the same variety of cells in the circulating blood; and third, 
a marked increase in the marrow of their direct antecedents. Opie 2 
found in experimental bacterial infections an accumulation of 
eosinophiles at the point of inoculation, where these cells undergo 
nuclear fragmentation and other degenerative changes and thereby 
probably repel infection, although, unlike the polynuclear neutro- 
philes and large mononuclear leucocytes, they rarely if ever act 
as phagocytes. Coincident with this massing of eosinophiles at 
the site of the infection these cells are exceedingly scanty in the 
peripheral blood, but collect in large numbers in the spleen. Ac- 
cording to Ehrlich's latest views, 3 leucocytosis involving mainly an 
increase of the polynuclear neutrophiles (" polynuclear neutrophile 
leucocytosis") is the expression of an independent chemotactic 
reaction on the part of these cells, caused by the remote influence 
of dissolved substances upon the bone marrow, whereby this, tissue 
throws into the blood current excessive numbers of these elements. 
Schultz 4 and others, on the contrary, attribute leucocytosis en- 
tirely to changes in the distribution of the cells, maintaining that 
increase in the number of leucocytes in the peripheral vessels 
goes hand in hand with a decrease in their number in the vessels 
of the internal organs, and vice versa. This view, however, has 
been shown to be untenable. 

1 Brit. Med. Jour., 1898, vol. ii, p. 604; also Trans. Path. Soc., London, 1902, 
vol. liii, p. 379. 

2 Amer. Jour. Med. Sci., 1904, vol. cxxvii, p. 988. * Loc. cit. 
4 Tagebl. der Naturforsch. Vers., Heidelberg, 1889, p. 405. 



In summing up the various experimental and clinical data bear- 
ing upon the nature of the leucocytoses associated with patho- 
logical conditions, the evidence tends to confirm the view that the 
process is, in all instances, save perhaps those of trivial local in- 
fections, a general one throughout the entire circulatory system, 
and that it is symptomatic of an excessive output and rapid de- 
velopment of leucocytes by the bone marrow, due to the influence 
of chemotactic principles. It must be remembered that this view 
is in part based upon hypotheses, but it nevertheless represents 
the belief current at the present time. 

i. Inflammatory and Infectious Leucocytosis. In this class are 
included the leucocytoses occurring during the course of a num- 
ber of diseases of inflammatory and infectious character, in which 
the increase may be attributed either to simple inflammation or to 
bacterial infection, or to both. The presence of such a leucocy- 
tosis is to be regarded as symptomatic of an attempt on the part 
of the organism to overcome the noxious invading principle, what- 
ever its nature may be, through the protective action of the white 
corpuscles. Bearing in mind this construction of the phenom- 
enon, it is possible in many instances to derive valuable clinical 
information from the presence or absence of a cellular increase. 

A genuine inflammatory or infectious leucocytosis is a much 
more potent defensive agent than a leucocytosis excited artificially 
by the injection of nuclein, for example. (See p. 251.) Accord- 
ing to Labbe, 1 an induced leucocytosis, while hypothetically an 
antidote to disease, practically is of trivial value as a means of 
defense against a definitely established infection. As a preventive 
measure, however, an artificially provoked leucocyte increase 
may confer a certain degree of protection. 2 Bezancon and Labbe 3 
maintain that a polynuclear neutrophile leucocytosis follows 
active, rapid infections, and confers simply a transitory immunity, 
while an increase affecting chiefly the mononuclear cells attends 
slow infections and generally gives permanent protection. 

The view expressed by von Limbeck, 4 that the height of the 
leucocytosis is dependent upon the extent of the inflammatory ex- 
udate, is not tenable, for processes characterized by insignificant 
exudates are capable of causing as great an increase as those in 
which this outpouring is extensive. As a rule, leucocytoses as- 
sociated with purulent exudates are much more marked than 
those due to serous effusions. The essential factor in determining 

1 Presse med., 1903, vol. ii, p. 725. 

2 For an account of the practical value of induced leucocytosis in wound in- 
fections (reviewing the studies of Lowy, Richter, Jacob, Hahn, Goldscheider, Hof- 
bauer, Salieri, and Miyake) see Mikulicz-Radecki, Lancet, 1904, vol. ii, p. i. 

3 Presse m^d., 1902, vol. ii, p. 1071; also "Traite d'Hematologie", Paris, 1904. 

4 Loc. cit. 


the degree of the increase is not the extent of the exudate nor, 
in all cases, its character, but rather the systemic reaction to which 
it gives rise. 

The degree of leucocytosis may be considered a general index 
to the intensity of the infection and to the strength of the indi- 
vidual's resisting powers in reacting against it. It follows, 
therefore, that intense infections occurring in individuals whose 
resisting powers are strong produce a decided increase; but 
the presence of an infection of like intensity in one whose re- 
sisting powers are greatly crippled fails to cause leucocytosis, for 
in such an instance the organism is s<5 overpowered by the effects of 
the morbid process that it is incapable of reacting. The increase is 
either absent or slight when a trifling infection is associated with 
vigorous resisting powers, and moderate when a moderately intense 
infection is linked to fairly well-developed resisting powers. 

The clinical inferences to be drawn from these facts are of 
value chiefly as corroborative of other w r ell-known physical signs, 
but are obviously untrustworthy when considered apart from the 
latter. A marked leucocytosis indicates simply an intense in- 
fection in a person whose resisting powers are normally developed 
and actively exerted against the disease, but it is of no prognos- 
tic value in itself, for it conveys no idea of the final outcome of 
the conflict 'between the disease and the organism. Absence of 
leucocytosis or a slight increase may be either of very v favorable 
or of very grave significance, inasmuch as these signs occur both 
in trivial and in overwhelming infections. If the absence is asso- 
ciated with clinical manifestations which point to a severe"infec- 
tion, the sign may be depended upon as being of grave prognosis. 

The clinical significance of the leucocytoses associated with 
various inflammatory and infectious processes will be discussed 
in Section VII. A more or less decided increase in the num- 
ber of leucocytes occurs with great constancy in the following 
groups of diseases of this nature : 

I. General Infectious Diseases. 

Actinomycosis. Pneumonia. 

Asiatic cholera. Pyemia. 

Bubonic plague. Relapsing fever. 

Cerebrospinal meningitis. Rheumatic fever. 

Diphtheria. Scarlet fever. 

Dysentery. Septicemia. 

Erysipelas. Spotted fever (Montana). . 

Filariasis. Syphilis (secondary). 

Glanders. Trichiniasis. 

Malignant endocarditis. Vaccinia. 

Multiple neuritis. Varicella. 

Osteomyelitis. Variola. 

Pertussis. Yellow fever. 


II. Simple and Infective Local Inflammations. 

Acute yellow atrophy of the liver. Pellagra. 

Appendicitis, catarrhal. Pemphigus. 

Arthritis, serous. Pericarditis. 

Bronchitis, acute. Peritonitis. 

Burns. Prurigo. 

Cholangitis. Purulent lesions: 

Cholecystitis. Appendicular abscess. 

Cystitis. Cerebral abscess. 

Conjunctivitis, acute. Hepatic abscess. 

Dermatitis. Ischio-rectal abscess 

Eczema. Ovarian abscess. 

Endocarditis. Pancreatic abscess. 

Endometritis. Pelvic abscess. 

Enteritis. Perinephritic abscess. 

Epididymitis. Prostatic abscess. 

Gangrene: Pulmonary abscess. 

Appendicular. Retropharyngeal abscess. 

Cancrum oris Splenic abscess. 

Hepatic. Superficial abscess. 

Pancreatic. Arthritis, suppurative. 

Pulmonary. Carbuncle. 

Gastritis, acute. Empyema. 

Gastro-enteritis, acute. Felon. 

Herpes zoster. Furuncle. 

Hydatid disease. Gonorrhea. 

Infected wounds. Otitis media, suppurative. 

Mastitis. Phlebitis. 

Meningitis. Pyelonephritis. 

Nephritis, acute. Pyonephrosis. 

Orchitis. Pyosalpinx. 

Ovaritis. Quinsy. 

Pancreatitis. Splenitis. 

Under this heading it is convenient to include post-operative 
leucocytosis, or the increase commonly occurring after a surgical 
operation, as a symptom of the normal process of wound repair. 
The reparative process, however, is not always the sole factor of 
the leucocytosis, since the effects of the anesthetic and of hemor- 
rhage also must be taken into account. In non-septic uncom- 
plicated cases the increase amounts to between 5000 and 10,000 
cells per in excess of the pre-operative count. The maxi- 
mum increase, which is transitory, is generally attained within 
from twelve to twenty-four hours, and the normal standard is 
again reached, by a progressive decline in the leucocyte count, 
within two or three, or at the latest four, days. In uncomplicated 
cases the count falls to normal within twenty-four hours in Blood- 
good's experience 1 ; within thirty-six hours in Cabot's 2 ; within 
eighty-four hours in Frazier's 3 ; and within five days in C. Y. 
White's. 4 Persistence of a post-operative leucocytosis is signifi- 

1 Med. News, 1901, vol. Ixxix, p. 321. 

2 Annals of Surg., 1901, vol. xxxiv, p. 361. 

3 Univ. of Pa. Med. Bull., 1901, vol. xiv, p. 363. 

4 Univ. Med. Mag., 1900, vol. xiii, p. 260. 


cant of some such complication as defective drainage, infection, 
hemorrhage, or extensive inflammation. 

King 1 could determine no relation, in non-septic cases, be- 
tween the height of the leucocytosis and the ranges of the pulse 
and temperature. Frazier and Holloway 2 believe that the in- 
crease corresponds in general with the extent of the operation, 
but that it is uninfluenced by the degree of the temperature 
and by the anesthesia. 

2. Leucocytosis of Malignant Disease. A moderate leucocy- 
tosis is commonly, but by no means constantly, associated with 
the various forms of carcinomata and sarcomata, but the cases in 
which no increase is observed are even more numerous than those 
in which it occurs. It is more common and the increase is usually 
regarded as more marked in sarcoma than in carcinoma, but in 
neither condition are excessively high leucocytoses met with fre- 
quently. In the writer's experience the increase, when it does 
occur, is generally moderate in most forms of malignant disease, 
counts of less than 20,000 leucocytes per being the general 
rule. Cases in which the number of cells exceeds this figure 
are comparatively rare, but are distinctly more common in sar- 
coma than in carcinoma; it is especially in rapidly growing 
neoplasms of the lung, liver, and kidney that the cells rise to 30,000, 
40,000, or even 50,000 or more. In a series of 68 consecutive 
cases of malignant disease in the German and Jefferson hospitals 
less than one-half were accompanied by a leucocyte ' count of 
10,000 or more, this figure being reached or exceeded in approx- 
imately 45 per cent, of cases of carcinoma, while in sarcpma such 
an increase was noted in almost 65 per cent. Five per cent, of 
all cases showed a high leucocytosis that is, counts ranging 
between 30,000 and 50,000. (For further data concerning the 
leucocytosis of these conditions see "Malignant Disease," Section 

The behavior of the leucocytes in malignant disease app'ears 
less contradictory when we inquire into the actual influence which 
these growths exert in provoking leucocytosis. It is the cur- 
rent belief that malignant disease per se has little if any influ- 
ence of this sort, and that the increase, if any occurs, is attribu- 
table to local inflammatory complications and to secondary septic 
infections, rather than to the specific toxic effects of the neoplasm 
itself. In some instances it is reasonable to suppose that the 
profoundly cachectic state of the patient also is an important de- 
termining factor. Clinically, it is observed that tumors of rapid 

1 Amer. Jour. Med. Sci., 1902, vol. cxxiv, p. 450. 

2 Univ. of Pa. Med. Bull., 1901, vol. xiv, p. 363. 


development, involving a large area of tissue and complicated by 
extensive metastases, cause decided, often high, leucocytoses; while 
localized tumors, of small size and of slow growth, give rise to 
trifling, if any, increase. Variations from this general rule are 
the result of differences in the resisting powers of different indi- 
viduals, for the effects of this factor in causing leucocytosis are 
potent in this as in other pathological conditions. 

Qualitatively, the leucocytes usually show a marked absolute 
and relative increase in the polynuclear neutrophiles, with a con- 
sequent diminution of the mononuclear forms. But in some in- 
stances, both of carcinoma and of sarcoma, the polynuclear forms 
are relatively below the normal percentage, and the lymphocytes 
increased, so that the blood picture is not one of leucocytosis, but 
rather one of relative lymphocy tosis ; such a change seems es- 
pecially prone to occur in sarcoma of the lymphatic system, in 
which it may be so marked that it suggests lymphatic leukemia. 
There are certain cases of malignant disease in which the propor- 
tion of polynuclear neutrophiles rises, although the total number 
of leucocytes is not increased; the polynuclear gain is less than 
is usually found with high leucocyte counts, and is not 
to be regarded as of the same significance as a frank leu- 
cocytosis involving an increase in the total number of cells. (See 

P- 2 37-) 

The eosinophiles are variously affected in different cases : some- 
times they are greatly diminished, if not, indeed, entirely absent, 
as in most leucocytoses; sometimes they are normal; and some- 
times they are largely increased in number. The increase may 
be pronounced in sarcoma, this being due probably to involve- 
ment of the bone marrow by the growth, either directly or by 

Small numbers of myelocytes 0.5 to i or 2 per cent. are ex- 
ceedingly common, being found with great constancy in cases 
with marked cachexia, especially in carcinoma. In malignant 
disease involving the bone marrow the percentage of this variety 
of cells is much higher. 

3. Post-hemorrhagic Leucocytosis. A leucocytosis of moderate 
grade commonly occurs as the result of hemorrhage due to trau- 
matism or to other causes. It has been found that in animals 
the stage of increase is preceded by a well-defined leucopenia, 
which develops immediately after the loss of blood. This initial 
leucopenia, however, has not yet been demonstrated in man, 
although it probably occurs. In an extensive traumatic hemor- 
rhage the increase sometimes may be recognized in the periph- 
eral blood within an hour after the accident, but usually it is 


not distinguishable until after the lapse of a longer period from 
five to ten hours, as nearly as can be ascertained. In hemor- 
rhage accompanying various pathological conditions, such, for 
example, as gastric ulcer, lung tuberculosis, or uterine disease, 
the appearance of the leucocytosis is less prompt than in hemor- 
rhage from trauma. The maximum increase is usually within 
moderate limits approximately two or three times the normal 
standard. The injection of a salt solution decidedly aggravates 
the leucocytosis. As an illustration of the degree of leucocytosis 
which is commonly encountered, Rieder 1 noted in four cases 
(hematemesis from gastric ulcer, fatal hemophilia, and uterine 
hemorrhage) an average count of 22,625, tne maximum being 
32,600, and the minimum, 15,100. Inasmuch as the height of 
the increase is thought to correspond to the strength of the organ- 
ism's reaction in compensating the blood loss, it varies in different 
cases. In two individuals of equally strong regenerative powers 
a severe hemorrhage will produce a greater leucocytosis than a 
slight one. The duration of the increase also varies with the 
individual case, for it depends upon a similar factor; but in the 
majority of instances it does not last for more than three or four 
days, according to the investigations of Lyon. 2 Leucocytoses 
excited by traumatic hemorrhage are prone to persist longer than 
those due to other causes, and the long-persisting increases which 
are sometimes associated with other pathological lesions should 
be attributed to factors other than the actual loss of .blood. In 
an instance of leucocytosis following venesection Rieder 3 found 
that the increase persisted for twelve days. Head 4 found that 
in dogs the leucocytosis following extensive hemorrh'age lasted 
for at least seven days. 

Usually the qualitative changes involve chiefly the polynu- 
clear neutrophiles, which are greatly increased at the expense of the 
other forms of cells, but in an occasional instance it will be found 
that the mononuclear varieties are greatly in excess of their normal 
percentages, so that a lymphocytosis is observed. Myelocytes 
may also be found in considerable numbers in many cases. 

4. Toxic Leucocytosis. Typical examples of toxic leucocytosis 
are found in poisoning by ptomains and by coal-gas, in both of 
which conditions the predominant influence of a toxic agency in 
producing the increase is self-evident. For the same reasonjthe 
leucocytoses occurring as the result of ether and chloroform nar- 
cosis, and in convulsions and acute delirium, are included in this 
classification. In certain diseases, notably in the uric acid diath- 

1 Loc. cit. * VirchoVs Arch., 1881, vol. Ixxxiv, p. 207. 

3 Loc. cit. 4 Jour. Amer. Med. Assoc., 1901, vol. xxxvii, p. 501. 


esis, in cholemia, and in uremia, the presence in the blood of 
toxic principles is thought to be the underlying factor of the in- 
crease; the same probably is true of a number of other diseases, 
which, for obvious reasons, have been classed with the infectious 
and inflammatory leucocytoses. 

The effect of gas-poisoning is well illustrated by the blood 
examination of a patient admitted to the German Hospital, 
fatally poisoned by illuminating-gas. The leucocytes were in- 
creased to 32,000 per, the gain being due to an excessive 
predominance of polynuclear neutrophiles, as determined by the 
following differential count: small lymphocytes, 3.5 per cent.; 
large lymphocytes, 2.5 per cent.; polynuclear neutrophiles, 92 
per cent.; eosinophiles, 0.5 per cent.; and myelocytes, 1.5 per 
cent. To what extent this increase depended upon the actual 
toxic effects of the gas and to what extent it was attributable 
to peripheral stasis (which was marked in this^patient) are con- 

The leucocytosis caused by ether narcosis has been exhaus- 
tively studied by von Lerber 1 and by Chadbourne. 2 The inves- 
tigations of von Lerber included 101 cases, of which number leu- 
cocytosis was found in more than 95 per cent., the increase fre- 
quently amounting to two or three times the original count; in 
the majority of instances the maximum count was observed sev- 
eral hours after the anesthesia was produced. Chadbourne has 
carefully studied 21 cases, all of which showed a more or less 
decided leucocytosis, the minimum gain being 6 per cent., the 
maximum 73 per cent., and the average 37.3 per cent. He found 
that the leucocytosis developed most rapidly during the early 
part of the etherization, and that only exceptionally did it persist 
for more than twenty-four hours. Differential counts in five 
cases showed that all forms of cells were proportionately in- 
creased. This author attributes the increase to the irritating 
effects of the ether vapor upon the mucous membrane of the 
respiratory tract. J. Chalmers Da Costa and Kalteyer, 3 in 50 
cases, found an average leucocytosis of about 5000 per 
Results similar to the above also have been obtained by Ewing 4 
and by Ames, 5 in the experimental etherization of animals. 

Ether also causes moderate polycythemia, but this is due in 
partjto inspissation of the blood by the preparatory treatment of 

1 "Ueber die Einwirkung der Aethernarkose auf Blut u. Urin," Inaug. Dis- 
sert., Berlin, 1896. 

a Phila. Med. Jour., 1899, vol. iii, p. 390. 
8 Annals of Surg., 1901, vol. xxxiv, p. 329. 

4 N. Y. Med. Jour., 1895, vol. Ixi, p. 257. 

5 Jour. Amer. Med. Assoc., 1897, vol. xxix, p. 472. 


the patient. A slight diminution in hemoglobin is commonly, 
but not invariably, found. 

The effects of chloroform are similar to those of ether, but 
are 'more marked and persist longer. Solimei 1 found that the 
leucocytosis induced by chloroform narcosis, while sometimes dis- 
appearing within a few hours, often continues for several days, 
the duration of the increase being related to the amount of the 
anesthetic used and to the rate of its elimination from the system. 
This observer also detected a loss of hemoglobin and erythrocytes 
with moderate poikilocytosis, and, after protracted narcosis, de- 
layed coagulability, the presence of a methemoglobin spectrum, 
and an increase in the carbon dioxid of the blood, with a corre- 
sponding diminution of oxygen. Holman 2 warns the operator 
against the danger from hemolysis by chloroform in patients 
whose hemoglobin percentages arc law. 

The leucocytosis associated with acute delirium and with con- 
vulsive seizures, due to a variety of causes, has been studied in 
detail by Capps 3 and by Burrows. 4 Under such circumstances the 
increase is usually marked, and the height of the count in a gen- 
eral way is dependent upon the severity of the attack. The 
polynuclear neutfophiles are chiefly concerned in the increase, 
with a consequent decline in the proportion of mononuclear 
forms. The leucocytosis of this class of diseases is discussed 
more fully in Section VII. 

5. Experimental Leucocytosis. Leucocytoses not, differing es- 
sentially from those associated with various local and general 
infections may be caused by the administration of many drugs, 
chemicals, organic principles, bacteria, bacterial proteins, and by 
the application of intense irritants and revulsives to the surface 
of the body. No doubt many of these leucocytoses should be 
classed either as inflammatory or as toxic, owing to the character 
of their exciting causes, but for the sake of convenience they may 
be grouped under this heading. v 

Leucocytoses resulting from the administration, subcutaneously 
and by the mouth, of various drugs and other substances have 
been studied chiefly by the Continental investigators, to whom we 
are indebted for most of our present knowledge of this subject. 
The manner in which such agencies act in causing the increase is 
not at all clear in many instances, but, as a rule, the change is 
thought to be dependent upon chemotactic influences, as in in- 

1 Gaz. degli Ospedali e. d. Clin., 1902, vol. xxiii, p. 108. 

2 St. Paul Med. Jour., 1902, vol. iv, p. 618. 

3 Amer. Jour. Med. Sci., 1896, vol. cxl, p. 650. 

4 Ibid., 1899, vol. cxvii, p. 503. 


flammatory and infectious leucocytoses,. as well as upon concen- 
tration of the blood from vasomotor changes. 

Lowit 1 determined that a preliminary leucopenia succeeded by 
a more or less decided leucocytosis followed the subcutaneous in- 
jection of the following substances : hemialbumose, pepsin, nuclein, 
nucleic acid, curare, leech extract, tuberculin, filtered yeast cultures, 
pyocyanin, sodium urate, and uric acid. This change was not ob- 
served, however, after the injection of urea. 

Goldscheider and Jacob, 2 conducting a large number of experi- 
ments with various organic animal extracts, obtained results simi- 
lar to Lowit's from the injection of the extracts of spleen, thymus, 
and bone marrow, but found negative results from the use of the 
extracts of thyroid, pancreas, and liver. 

Winternitz 3 studied the effects resulting from the subcutaneous 
injection of substances causing both transient inflammatory edema 
and aseptic abscess formation at the site of the injection. In the 
former class, which includes neutral salts and dilute acids and 
alkalis, an increase in the number of leucocytes, amounting to 
from 40 to 75 per cent, of the original count, was noted; and it 
was furthermore found that even although a local necrosis was 
produced by the injection of an irritating salt, the increase still 
did not become excessive. In the second class of more active 
irritants which produced local abscesses turpentine, oil of 
mustard, carbolic acid, croton oil, sapotoxin, digitoxin, silver 
nitrate, cupric sulphate, and salts of mercury and of antimony the 
leucocytosis was much more decided and of less transient dura- 
tion. As a rule, in these experiments the height of the leucocy- 
tosis ran parallel to the intensity of the local irritation provoked. 

Pohl 4 noted a moderate leucocytosis following both the inges- 
tion and the injection of absinthe, acetic ether, extract of gentian, 
peppermint, piperin, the oils of anise and fennel, egg albumin, 
and sodium albuminate. With the last two substances he deter- 
mined that the increase was greater when they were given by the 
mouth than when administered subcutaneously ; the gain usually 
ranged from about 5 to 50 per cent, of the normal count. This 
investigator also found that quinin, caffein, calomel, sodium bicar- 
bonate, ethyl alcohol, and hydrochloric acid did not cause a leucocy- 
tosis, while bismuth subnitrate and oxid of iron produced irregular 
results. Many of the above experiments have been substantiated 
by the later work of von Limbeck. 5 Memmi 6 produced a moderate 

1 Loc. cit. 2 Loc. cit. 

3 Arch. f. exp. Pathol. u. Pharmak., 1895, vol. xxxv, p. 77. 

4 Ibid., 1889, vol. xxv, p. 51. 

5 Loc. cit. * Gaz. degli Ospedali e. d. Clin., 1903, vol. xxiv, p. 995. 


persistent leucocytosis by the daily intravenous injection of mer- 
curic bichlorid, in ordinary doses, the increase first becoming ap- 
parent a few days after beginning the treatment and continuing as 
long as the drug was given. A single injection generally caused 
no increase unless the dose was excessive. In such an instance the 
maximum increase occurred within six hours, and the normal count 
of cells was reached within twenty-four hours after the injection. 
The intravenous injection of lecithin was found by Stassano 1 to 
increase the number of leucocytes, especially those of the mononu- 
clear variety, the leucocytosis thus produced lasting four or five 

Wilkinson 2 observed leucocytosis, preceded by leucopenia, 
after the injection of potassium iodid, camphor, quinin, antipyrin, 
salicin, salicylic acid, nuclein, and pilocarpin; by the repeated 
administration of the latter drug it was found that the -granules 
of the polynuclear neutrophiles disappeared, although no effect 
was produced upon the granules of the eosinophile cells. Von 
Jaksch 3 also studied the effects of the injection of pilocarpin, and 
of the administration by the mouth of nuclein, and found that 
by either procedure a temporary and sometimes very marked 
leucocytosis resulted. Borini 4 found a similar effect was pro- 
duced by aleuron, but that with digitalis a more prolonged leuco- 
cytosis occurred. The leucocytosis caused by the ingestion of 
salicylic acid, according to Schreiber and Zandy, 5 gradually dis- 
appears after the drug has been given for a few days. 

The effects of the ingestion of the essential oils of peppermint, 
turpentine, and cinnamon have been studied by Meyer, 6 while 
Hirt 7 has investigated the influences of the simple oitters and 
drugs, such as the tincture 0} myrrh. Such drugs were found to 
cause a moderate but easily recognized leucocytosis. Krus- 
man, 8 by the injection of spermin and of protalbumose, and Bes- 
redka, 9 by a similar use of carmin and of arsenic trisulphate, have 
obtained varying degrees of increase in the number of leucocytes. 
A marked increase is produced, according to Bohland; 10 by the 
injection of morphin, Dover's powder, sodium salicylate, pilocar- 
pin, antipyrin, phenacetin, and antifebrin. 

Shaw 11 finds that a marked polynuclear neutrophile leucocytosis 

I Me"d. mod:, 1902, vol. xiii, p. 63. 2 Brit. Med. Jour., 1896, vol. ii, p. 836. 

3 Centralbl. f. inn. Med., 1892, vol. xiii, p. 81. 

4 Centralbl. f. Bakt. u. Parasit., 1902, vol. xxxii, p. 207. 

5 Deutsch. Arch. f. klin. Med., 1899, vol. Ixii, p. 242. 
* Cited by von Limbeck, loc. cit. 

''Ibid. 'These de St. Petersbourg, 1898. 

8 Annal. de 1'Institut Pasteur, 1899, vol. xiii, p. 49. 
10 Centralbl. f. inn. Med., 1899, vol. xx, p. 361. 

II Jour. Path, and Bacteriol., 1902, vol. viii, p. 70. 


develops after the administration of sodium cinnamate. The 
favorable action of this drug in tuberculosis is attributed partly to 
its ability to excite and to maintain leucocytosis. 

In addition to the substances already mentioned, the leucocy- 
tosis-producing effect of various purgative drugs, 1 of the transfusion 
of blood and of normal salt solution? of the subcutaneous use of 
fibrin ferment, 3 of hemoglobin* and of bacterial cultures, 5 bacterial 
extracts, 6 and protein 1 has also been demonstrated. 

Thymectomy in animals is followed by a well-marked leuco- 
cytosis, associated with an increase in the bactericidal properties 
of the blood. In a number of experiments upon dogs Cazin 8 
determined that after contusions of the abdomen the number of 
leucocytes was tripled or even quadrupled within three or four hours 
after the injury; the height of the count and the rapidity of the 
onset of the leucocytosis corresponded in these experiments to the 
severity of the visceral injuries inflicted. The same author found 
but a trifling leucocytosis after bullet wounds of the abdomen with 
intestinal perforation. 


An increase, whether relative or absolute, in the lymphocytes 
above the number normal in health is known as lymphocytosis. 
The word mononucleosis also is used to denote this change. Rela- 
tive lymphocytosis involves simply a gain in the percentage of 
lymphocytes without a coincident increase in the total leucocyte 
count. Absolute lymphocytosis, on the other hand, is character- 
ized by an increase above normal both in the percentage of lympho- 
cytes and in the total number of leucocytes. Barring lymphatic 
leukemia, in which the lymphocytes are both relatively and abso- 
lutely in excess, lymphocytosis is almost always a relative condi- 
tion, or at least it is not accompanied by a decided rise in the 
total leucocyte count. 

The increase in the proportion of lymphocytes is moderate in 
most instances, the greater number of differential counts showing 
percentages of these cells ranging from 50 to 70, in comparison 
with the maximum normal percentage, about 30. These figures, 
of course, refer to the blood of adults, for in children the increase 

1 De Rienzi and Boeni, Gaz. degli Ospedali e. d. Clin., 1898, vol. xix, p. 1570. 

2 Hand, N. Y. Med. Jour., 1900, vol. Ixxi, p. 556. 

3 Birk, "Das Fibrin-Ferment im lebenden Organismus," Dorpat, 1880. 

4 Bojanus, "Exp. Beitrage z. Physiol. u. Pathol. d. Blutes," Dorpat, 1881. 

5 Hankin and Kanthack, Centralbl. f. Bakt. u. Parasit, 1892, vol. xvii, p. 782. 
* Buchner, Arch. f. Hyg., 1890, vol. x, p. 84. 7 Ibid. 
8 Sem. med., 1903, vol. xxiii, p. 351. 


is generally greater, owing to the higher proportion of lympho- 
cytes normally found at this period of life. A differential count, 
which shows, for instance, 60 per cent, of lymphocytes, means a 
decided lymphocytosis in the adult, but is entirely within the 
normal limits in the young infant. Either type of cells, large or 
small, may predominate, or the change may not involve any con- 
spicuous deviation from the normal ratio of one form to the other. 
Frequently it happens that the two varieties possess such similar 
characteristics that it is impossible to determine which prevails. 
Occasionally the lymphocytosis depends largely upon unusually 
large percentages of the so-called "transitional" forms, while in 
other instances, notably in the lymphocytosis of von Jaksch's 
anemia, rickets, syphilis, variola, and scarlatina, the increase 
chiefly involves Ehrlich's large mononuclear forms, reputed to 
originate in the marrow. 

Lymphocytosis may be due either to changes in the distribu- 
tion of the cells through the circulatory system, or to their in- 
creased production and output by the lymphatic tissues. Ehr- 
lich 1 attributes lymphocytosis to the local irritation of certain 
areas of lymphatic glands which produces an increased circulatory 
activity in these situations, in consequence of which large numbers 
of lymph elements are swept mechanically from the lymphatics 
and enter the general circulation. He does not regard the change 
as an expression of an active chemotactic reaction, to which the 
lymphocytes are insensible. It also appears reasonable to pre- 
sume that the lymphocytosis which often accompanies leucopenia 
may be traced to still another factor, that of negative chemotaxis, 
which diminishes the number of polynuclear neutrophiles and 
thus brings about a relative increase in the lymphatic elements, 
upon which the repellent action is not exerted. 

Lymphocytosis has been observed in a number of pathological 
conditions, but its presence may be considered physiological in 
but a single instance in the blood of infants and young, ctiildren, 
in whom such a change is entirely normal. During the decline 
of life, on the other hand, the lymphocytes are relatively dimin- 
ished. This tendency toward a lymphocytic increase in infantile 
life, which becomes less notable as the child matures, is prone to 
become markedly exaggerated in many of the forms of secondary 
anemia from which children suffer, especially the anemias second- 
ary to syphilis, tuberculosis, rachitis, gastro-enteritis, and scurvy; 
less commonly, it has been observed in the acute infections. 

Lactation, conditions of cachexia, and great debility in the 
adult are in many instances accompanied by abnormally high per- 

1 Loc. cit. 


centages of lymphocytes in the blood. It is a well-known fact 
that differential counts show a higher percentage of mononuclear 
non-granular elements in the blood of the enfeebled and poorly 
nourished than in that of the active and vigorous individual. 

Similar changes are frequently associated with the terminal 
stages 0} a number 0} diseases, and may be found ajter hemorrhage 
from various causes trauma, hemophilia, purpura, and following 

Lymphocytosis, sometimes decidedly marked, is common in 
certain of the severe anemias, especially in chlorosis, pernicious 
anemia, Addison's disease, and in syphilitic and tuberculous 
secondary anemias; it may be observed in certain of the injections, 
such as enteric fever, malarial fever, Malta fever, scarlet fever, 
measles, diphtheria, pertussis, variola, phthisis, pneumonia, 
plague, and trypanosomiasis. A high degree of lymphocytosis 
has been reported by Labbe and Bernard 1 in tropical hemato- 
chyluria resulting from filariasis. Roger's contention, 2 that in- 
fections characterized by lymphocytosis are of protozoan type, 
is scarcely tenable, in view of the occurrence of this change in 
enteric fever, in Malta fever, and in tuberculosis, to name but 
three bacterial infections in which the lymphocytes are increased. 

Diseases involving the spleen and lymphatic glands are often the 
cause of a varying degree of increase in the lymphocytes, common 
examples of such conditions being chronic malarial splenic tumors ; 
kala-azar; simple, syphilitic, and tuberculous adenitis; and ma- 
lignant neoplasms, especially sarcoma, of the lymph glands. On 
the contrary, extreme decrease in the lymphocytes develops in 
consequence of obliteration of the lymph channels by malignant 
tumors. Enlargement of the thyroid gland relatively increases the 
lymphocytes, as does that systemic taint known as the constitutio 
lymphatica. Chloroma may provoke high absolute lymphocytosis. 

Distinct lymphocytosis has been observed by Wilkinson 3 after 
the injection of quinin hydrochlorate; and by Perry 4 and Lepine 5 
as the result of the administration of thyroid extract. It also follows 
the injection of tuberculin, pilocarpin, and emulsion of cancerous 

From a clinical viewpoint lymphocytosis is of value chiefly in 
the diagnosis of lymphatic leukemia. Marked absolute increase 
in the number of lymphocytes, associated with enlargement of the 
lymphatic glands, forms a pathognomonic picture of this disease. 

1 Sem. me"d., 1902, vol. xxii, p. 433. 

J "Infectious Diseases," Eng. trans, by Gabriel, New York and Philadelphia, 

1 Loc. cit. 4 Med. Record, 1896, vol. 1, p. 289. 

8 Sem. med., 1902, vol. xxii, p. 409. 


The recognition of a doubtful case of syphilis may be facilitated 
by the occurrence of lymphocytosis plus eosinophilia. 


The term eosinophilia is used to denote an increase above the 
normal standard in the number of eosinophiles in the circulating 
blood, this change usually, but not necessarily, being associated 
with a coincident increase in the relative percentage of these cells 
to the other forms of leucocytes. Thus interpreted, eosinophilia 
is a condition of absolute increase, in contradistinction to a purely 
relative gain in percentage, to which the term is not strictly appli- 

For the sake of uniformity it is customary to speak of the per- 
centage of eosinophiles rather than of their actual number, but 
in order to determine accurately the presence or absence of eosino- 
philia it is also essential in every instance to calculate the number 
of eosinophiles to the of blood from data obtained by a 
numerical estimate and a differential count of the leucocytes, 

Total number oj leu- .. Percentage of eosinophiles to Total number of eosin- 

cocytes per other forms of leucocytes. ophiles per 

The necessity for such a calculation is forcibly illustrated in 
myelogenous leukemia, since in this condition the ' relative per- 
centage of eosinophiles is often well within the normal limits, 
and yet a striking degree of eosinophilia may exist. Eor example, 
in a given case of this form of leukemia, the blood count shows 
300,000 leucocytes per with 5 per cent, of eosinophiles. 
This percentage, interpreted into the actual number of cells, 
means an eosinophilia of 15,000 per, or an increase of 
thirty-fold in excess of the highest normal figure. 

On the basis of a variation in the normal number of leucocytes 
of from 5000 .to 10,000 per, the absolute number of eosino- 
philes may range from 25 to 500 per in the blood of the 
healthy adult. An increase in excess of this maximum standard, 
regardless 0} the percentage indicated by the differential count, con- 
stitutes eosinophilia. 

Granting the accuracy of the current view that the hemic 
eosinophiles are purely myelogenous elements, their increase in 
the blood may be attributed to the influence of chemotaxis, prob- 
ably of a specific and selective character. Under the influence of 
such a stimulus the eosinophiles are attracted from, and perhaps 
overproduced by, the bone marrow, and are thrown into the gen- 


eral circulation in large numbers. It is also possible that to a 
slight extent their proliferation from like cells may occur in the 
blood stream as well. 

Increase in the number of eosinophiles occurs as a physiolog- 
ical change in young infants, in women during the menstrual period, 
and after coitus. With these three exceptions the presence of 
eosinophilia is always to be regarded as an evidence of some 
pathological condition. 

Once believed to be a pathognomonic sign of leukemia, in the 
light of more recent investigations eosinophilia is now known to 
be associated with diseases of almost every conceivable nature; 
in fact, it has been reported in so large a number of conditions 
of such widely dissimilar pathogenesis that its value as a clinical 
sign must be largely restricted. Inasmuch as many of these re- 
ported instances of eosinophile increase lack verification, it can 
only prove confusing to give here a list of the many pathological 
states in which the change is reputed to have been observed. 
The following list, based upon the work of Cannon, 1 Zappert, 2 
Gollasch, 3 T. R. Brown, 4 von Noorden, 5 and others, includes only 
those diseases in which eosinophilia is observed with a great de- 
gree of constancy. Such conditions are: 

I. Diseases of the Skin. 

Dermatitis herpetiformis. Pellagra. 

Eczema. Pemphigus. 

Erythema multiforme. Prurigo. 

Herpes zoster. Psoriasis. 

Leprosy. Scleroderma. 

Lupus. Urticaria. 

II. Helminthiasis. 

Ankylostomiasis. Oxyuridiasis. 

Ascariasis. Strongyloides intestinalis infection. 

Bilharziasis. Teniasis. 

Filariasis. Trichiniasis. 

Hydatid disease. 

III. Diseases of the Bones. 

Hypertrophy. Multiple periostitis. 

Malignant neoplasms. Osteomalacia. 

1 Deutsch. med. Wochenschr., 1892, vol. xviii, p. 206. 

2 Zeitschr. f. klin. Med., 1893, vol. xxiii, p. 227; also Wien. k!in. Wochenschr., 
1892, vol. v, p. 347. 

3 Fortschr. d. Med., 1889, vol. vii, p. 361. 

4 Johns Hopkins Hosp. Bull., 1897, vol. viii, p. 79. 

5 Zeitschr. f. klin. Med., 1892, vol. xx, p. 98. 


IV. Post-febrile. 

Malarial fever. Scarlet fever. 

Pneumonia. Septicemia. 

Rheumatic fever. Varicella. 

V. Bronchial Asthma. 

VI. Myelogenous Leukemia. 

In addition to the conditions listed above, eosinophilia also 
occurs, but with less constancy, in some forms of the high-grade 
secondary anemia of childhood, purpura, hemorrhagic effusions, 
gonorrhea, syphilis, malignant disease, and in fibrinous bron- 
chitis. It is also seen in many cases of splenomegaly and after 
splenectomy, its development under the latter circumstance 
being regarded as a compensatory condition. An increase in 
the percentage of eosinophiles has also been noted in conditions of 
starvation. In scarlet fever the eosinophiles usually persist, in 
spite of the coexisting polynuclear leucocytosis, and the same 
peculiarity may often be found in trichiniasis, filariasis, hydatid 
disease, and infection with intestinal parasites. 

Eosinophilia may be produced experimentally by the injection 
of a number of medicaments, such as antipyrin, camphor, nu- 
clein, phosphorus, pilocarpin, tuberculin, and many of the iron 
salts. T. R. Brown 1 found high eosinophilia (12 per cent, in a 
leucocyte count of 30,000) in acetanilid poisoning. The writer 
detected eosinophilia in 3 cases of poisoning by nitrites. > 

Neusser 2 and his school have contended that eosinophilia is 
symptomatic of an extensive group of diseases, chiefly . those in- 
volving the sympathetic nervous system, the sexual organs, and a 
long list of disorders which they attributed to the "xanthin dia- 
thesis." Most of these views have been unsubstantiated, many 
are misleading, and a few can be shown to be fanciful. .TRose 
who are inclined to investigate some of the remarkable claims 
made by Neusser as to the diagnostic and prognostic value of 
eosinophilia are referred to his original communication on the 

Diminution in the number of eosinophiles occurs as a physiolog- 
ical process during digestion and after active muscular exercise. 
It is observed usually in lymphatic leukemia, tuberculosis, during 
the febrile stages of diphtheria, influenza, pneumonia, enteric fever, 
and septicemia, frequently after hemorrhage, and in the terminal 

1 Md. Med. Jour., 1902, vol. xlv, p. 307. 

2 Wien. klin. Wochenschr., 1894, vol. vii, p. 737. 



stages of many diseases. The number of eosinophiles is said to be 
diminished ajter castration. The writer has found a decrease or 
even absence of eosinophiles in the majority of cases of chlorosis 
and pernicious anemia. 

The chief clinical value attached to eosinophilia relates to its 
presence in trichiniasis, in filariasis, in intestinal helminthiasis, 
and in hydatid disease, in which infections it has been shown to 
be a sign of great reliability. It may, however, be absent in the 
later stages of these infections. 

In the diagnosis of an exanthema which is suggestive either of 
scarlet fever or -of measles, eosinophilia points to the former dis- 
ease, for it does not occur in the latter. 

The association of eosinophilia and lymphocytosis constitutes a 
blood change which may be helpful in the recognition of an ob- 
scure case of syphilis. 

High percentages of eosinophiles in chlorosis, in pernicious 
anemia, and after hemorrhage are generally regarded as an evi- 
dence of good regenerative powers of the hemogenic organs, and 
are therefore of favorable import (Rieder). 1 


Increase in the number of basophiles in the circulating blood 
is of rare occurrence, having been observed in but few diseases 
except the myelogenous variety of leukemia, in which this change 
is quite constant, and sometimes most striking; the basophiles in 
this disease may constitute 10, 20, or even 30 per cent, of all 
forms of leucocytes. The increase may involve the finely 
granular or the coarsely granular (mast cell) forms, or both. 

Until recently but little attention has been paid by hematolo- 
gists to the general circulatory form of basophilia, although the 
local increase of the basophiles under various conditions has been 
well investigated. Canon 2 has reported an increase of the mast 
cells in a case of chlorosis and in various skin diseases. Sherring- 
ton 3 has observed a similar blood change in patients dying in the 
reaction stage of Asiatic cholera. A. E. Taylor 4 states that he has 
seen a notable circulatory basophilia in a case of carcinoma, with 
marked cachexia, but without bone metastases; in a case of 
gonorrhea; in a case of mycosis fungoides; and in two cases of 
septic bone disease. Basophilia has also been observed in variola, 

1 Loc. cit. * Deutsch. med. Wochenschr., 1892, vol. xviii, p. 206. 

3 Proc. Roy. Soc., London, 1894, vol. Iv, p. 189. 

4 "Contributions from the William Pepper Laboratory of Clinical Medicine," 
Philadelphia, 1900, p. 148. 


splenic anemia, Hanoi's cirrhosis, and in hydatid disease and other 
forms of helminthiasis. The writer found a mast cell basophilia 
of 2 per cent., with a leucocytosis of 13,0x20, in a case of plumbism, 
and has noted large numbers of mast cells in suppurative appen- 
dicitis, pernicious anemia, filariasis, and trichiniasis. 

Owing to our imperfect understanding of this condition no 
theory regarding the production of basophilia is as yet generally 
acceptable. It is possibly due to the influence of a specific chemo- 
tactic substance, in response to which the basophiles are attracted 
from the bone marrow and enter the general circulation. 


The presence in the circulating blood of myelocytes, in small 
or in large numbers, is known as myelemia. As previously re- 
marked, this condition is invariably to be regarded as patholog- 
ical, since myelocytes are never found in the blood of the normal 

The most striking example of myelemia is to be found in 
the myelogenous form of leukemia, in which condition this 
change constitutes one of the most conspicuous features of the 
blood picture. Myelocytes occur in the blood in this disease in 
greater absolute and relative numbers and with greater constancy 
than in any other condition a fact which is of the greatest diag- 
nostic value. The degree of increase may be enormous, as illus- 
trated by a case of the author's, in which the actual .number of 
myelocytes was found to be 192,738 per, or 27.3 per cent, 
of all forms of cells in a total leucocyte count of 706,000. In 
instances of splenomegaly with anemia and moderate leucocytosis 
small numbers of myelocytes have been reported as a constant 
finding by Emile Weil and Clerc. 1 Kurpjuweit, 2 in malignant dis- 
ease with bone metastases, found as high as 17 per cent, of mye- 
locytes, and he regards this myelemia as a sign of gravely altered 
hemogenesis. (See " Malignant Disease.") Von Jaksch 3 noted 
marked myelemia in his symptom-complex, termed multiple peri- 
ostitis with myelocythemia. 

Less frequently myelocytes are observed in lymphatic leu- 
kemia and in Hodgkin's disease, but in these conditions their 
occurrence is inconstant and their increase trivial. Small num- 
bers of myelocytes (from 0.5 to 2 or 3 per cent.) are found in 

1 Arch. gen. de Med., 1902, vol. cxc, p. 560. 

2 Deutsch. Arch. f. klin. Med., 1903, vol. Ixxvii, p. 553. 

3 Zeitschr. f. Heilk., 1901, vol. xxii, p. 259. 


almost every case of primary pernicious anemia, and are not un- 
common in marked cases of chlorosis and in many of the severe 
forms of secondary anemia due to various causes. They are fre- 
quently met with in such conditions as pneumonia, septicemia, 
diphtheria, syphilis, malignant disease, rachitis, tuberculosis, osteo- 
myelitis, osteomalacia, Addison's disease, and the malarial fevers. 
The writer has found them also in the following conditions: car- 
bon monoxid poisoning, hepatic cirrhosis, acute gout, malignant 
endocarditis, exophthalmic goiter, as well as in the above-named 
affections. One is forcibly impressed with the almost constant 
presence of myelocytes in the estivo-autumnal type of malarial 
fever, in severe septic infections, and in enteric fever in childhood, 
both in the early stages of the disease and during the later, post- 
febrile anemic period. Small numbers of myelocytes have been 
reported also in many other conditions, chiefly those associated 
with leucocytosis, with anemia, or with both. 

Increased activity of the bone marrow, whereby the myelocytes 
are forced into the blood stream, is in all probability responsible 
for the production of myelemia. In response to an increased 
demand for leucocytes the marrow becomes so overstimulated 
that many immature forms of leucocytes, or myelocytes, acciden- 
tally find their way into the general circulation, their passage from 
the marrow no doubt being accomplished largely by emigration. 
It is furthermore now believed that substances which are posi- 
tively chemotactic for the polynuclear neutrophiles also exert a 
similar attractive influence upon their immediate precursors, the 
myelocytes, stimulating their increased proliferation in the bone 
marrow and exciting their emigration from this tissue into the 
blood stream. 


Decrease below the normal standard in the number of leuco- 
cytes in the peripheral blood is known as leucopenia or hypoleuco- 
cytosis. Such a condition, like its antithesis, leucocytosis, may be 
the result of either physiological or pathological causes. Owing 
to the variation in the normal number of leucocytes in different 
individuals, it is difficult to determine arbitrarily just what degree 
of decrease may be considered as a leucopenia, but it is safe to 
apply the term to any leucocyte count decidedly below 5000 cells 
to the The number of leucocytes is rarely reduced to 
less than 3000, except in certain of the essential anemias, in 
which their decline to one-tenth the maximum normal figure 
or even less is occasionally to be observed. The most extreme 


instance of leucopenia on record has been reported by Koblanck, 1 
who found but a single leucocyte in a careful search through 
twenty stained cover-glass preparations of blood from a man of 
twenty-five years, suffering from epilepsy; the exact numerical 
estimate of the leucocytes in this case is not given in detail. 

The decrease may be accompanied by no deviation from the 
normal percentages of the different varieties of leucocytes, or it 
may involve a more or less decided gain in the lymphocytes, the 
latter being the more common change of the two. 

According to the nature of its underlying causes, leucopenia may 
be considered clinically as either physiological or pathological. 


The decrease in the number of leucocytes observed in several 
physiological states is generally attributed to vasomotor influ- 
ences which produce changes in the distribution of the leucocytes 
throughout the system. Such changes occur from the effect of 
prolonged cold, and brief hot, baths. 2 Decastelle 3 has found that 
a temporary leucopenia may be produced experimentally, by stim- 
ulation of sensory nerves, this procedure causing a reflex con- 
traction of the abdominal vessels and a consequent retention of 
large numbers of circulating leucocytes in this part of the vascu- 
lar system. The variations in the number of cells range from 
20 to 30 per cent, of the original count; the maximum decrease 
occurs usually within three or four minutes, and in most instances 
does not persist longer than ten or fifteen minutes. Reduction 
of blood pressure is promptly followed by a very transient* diminu- 
tion in the leucocytes of the peripheral blood. 

Malnutrition and starvation are also potent factors in the pro- 
duction of leucopenia, the decrease dependent upon such causes 
frequently being most pronounced. The much-cited case of the 
faster, Succi, is a good example of the effects produced upon the 
leucocytes by abstinence from food. Luciani 4 found in the blood 
of this individual a decrease in the number of leucocytes from 
14,530 to 861 per after a seven days' fast; on the eighth 
day an increase to 1530 occurred, this being the average count 
noted during the remaining twenty-one days of the fast. The 
subnormal leucocyte counts which are often met with in many 
of the infirm and the greatly enfeebled can be traced to the effects 
of faulty nutrition and to the malassimilation of food. 

1 Inaug. Dissert., Berlin, 1889. l Winternitz, loc. cit. 

3 Wien. klin. Wochenschr., 1899, vol. xii, p. 395. 

4 " Das Hungern" (German translation by O. Frankel), Hamburg and Leipsic, 



Leucopenia, or at least an absence of leucocytosis, occurs dur- 
ing the course of a number of general infectious diseases, promi- 
nent among which are the following: enteric fever, paratyphoid 
fever, measles, rotheln, influenza, leprosy, Malta fever, the malarial 
fevers, trypanosomiasis, and non-septic tuberculosis. In acute 
infections which are ordinarily accompanied by leucocytosis the 
combined influences of an intense infection and feeble resisting 
powers on the part of the individual may produce a distinct 
leucopenia, or may prevent the development of the characteristic 
increase. This is well illustrated by the low counts which some- 
times are found in severe cases of pneumonia and of appendicitis. 

Leucopenia, often pronounced, is not uncommon in chlorosis 
and in pernicious anemia, being much more frequent and more 
decided in the latter disease. A well-marked leucopenia may be 
expected in about one-fourth of all cases of chlorosis, and in 
quite three-fourths of cases of pernicious anemia. It is also often 
met with in some high-grade secondary anemias, notably in those 
due to syphilis and to rachitis, and in splenic anemia. 

D'Orlandi 1 has called attention to the frequency with which 
leucopenia is observed in certain of the severer forms of chronic 
gastro-enteritis in young infants. In a fatal case of primary 
infectious pharyngitis reported by P. K. Brown, 2 the leucocytes 
never numbered more than 400 to the 

In the anemias accompanied by a decrease in the leucocytes,, 
especially in primary pernicious anemia, the rule holds good that 
the more intense the oligocythemia and oligochromemia, the 
greater the degree of leucopenia. Ehrlich 3 attributes the de- 
crease in such cases to a lessened proliferative function of the 
bone marrow, in consequence of which there is a diminution in 
the output of leucocytes by this organ. 

In leukemia an acute intercurrent infection may produce an 
abrupt and marked fall in the number of leucocytes, as in Cabot's 
remarkable case of lymphatic leukemia, 4 in which, as the conse- 
quence of a fatal septicemia, the leucocytes fell in three weeks 
from 40,000 to 419 per 

Decrease in the number of leucocytes may be caused experi- 
mentally by the administration of various drugs and other sub- 
stances. Bohland 5 found that it followed the injection of ergot y 
sulphonal, tannic acid, camphoric acid, atropin, agaricin, and 

1 Rev. mensuelle des malad. de PEnfance, 1899, vol. xvii, p. 300. 

1 Amer. Med., 1902, vol. iii, p. 649. 3 Loc. cit. 

4 Loc. cit. 5 Loc. cit. 


picrotoxin. DelezeneV investigations showed that a marked 
decrease results from the injection of various anticoagulant sub- 
stances, such as peptone, diastase, and eel serum; he attributes 
the leucopenia thus produced to two factors actual destruction 
in the circulation of some of the leucocytes and dilatation of the 
blood vessels, in which the undestroyed cells tend to accumulate. 
The transient leucopenia which precedes an increase in the 
leucocytes has been discussed elsewhere. (See p. 239.) 

O. K. Williamson 2 has shown that an increased destruction 
of leucocytes is attended by an increased excretion of uric acid in 
the urine. His studies apparently prove that in cases in which 
a rise in the phosphoric acid curve follows a fall in the leucocyte 
curve and in the number of granular cells especially, this rise 
corresponds with a rise in the uric acid curve. 

1 Nouveaux Montpel. med., 1898, vol. vii, pp. 694, 733, 765, and 789. 
3 Lancet, 1903, vol. i, p. 657. 





Lorrain Smith, 1 by his carbonic oxid method, 
GENERAL estimates that the total volume of blood is greatly 
FEATURES, increased in chlorosis, this excess being due to an 
increase in the bulk of normal plasma, and being 
more marked the severer the case. Taking the normal blood 
volume as 3240 c.c., Smith found in 21 chlorotics an average of 
4883 c.c. Although the oxygen capacity of a blood unit is dimin- 
ished about one-half, the total oxygen capacity of the blood remains 
approximately normal 95 per cent. Lloyd Jones 2 also believes 
in a plasma increase in chlorosis, but further confirmation of the 
above experimental work is needed before this presumption can 
be registered as a fact. 

The proportion of dry residue of the whole blood is-subnormal, 
approximating in the severe case between one-half and one-third 
the normal amount, with a corresponding increase in water. 

There is a decided loss of blood albumin, mainly referable to 
the oligochromemia, but, as Biernacki 3 has pointed out, also due, 
at least in severe cases, to a deficiency in the amount of serum 
albumin. In high-grade chlorosis this author found that the 
dry residue of the whole blood might contain a normal iron con- 
tent, the inference being that in such instances the albumin .loss 
was not always confined to the hemoglobin. 

The blood drop is exceedingly pale and 
APPEARANCE watery-looking, and flows so abundantly from the 
OF THE puncture that it actually seems as if the whole 
FRESH BLOOD, mass of blood in the body must be increased; a 
large -sized drop usually follows the slightest prick 
of the needle, in spite of the obviously anemic appearance of the 
patient a marked contrast to the difficulty commonly experi- 
enced in pernicious anemia of obtaining enough blood for the 

1 Jour. Physiol., 1900, vol. xxv, p. 6. 2 "Chlorosis," London, 1897, p. 24. 

3 Zeitschr. f. klin. Med., 1894, vol. xxiv, p. 500. 



examination. The blood spread out in a film over the finger is 
transparent rather than opaque, and its fluidity is most striking. 
Microscopical examination of the fresh film shows excessive 
pallor of most of the erythrocytes, together with the presence of 
a variable number of cells of smaller diameter than normal, in 
the average case, and of cells decidedly deformed in shape, in 
severe cases. The resistance of the erythrocytes, as shown by 
their hypotonicity, 1 is increased. The practised observer can 
determine at first glance that the number of erythrocytes is not 
greatly decreased, except in an occasional case in which the 
oligocythemia may be so marked as to lead him to infer that he is 
dealing with a well-defined secondary anemia. 

Coagulation of the blood drop, in spite of the 

COAGULATION, fact that hyperinosis is absent, is generally very 

rapid in chlorosis often so rapid as to interfere 

with the technic of the examination, if one delays during this 


The specific gravity of the whole blood is more 
SPECIFIC or less diminished, the degree of decrease being 
GRAVITY, closely parallel with the loss of hemoglobin. 
Lloyd Jones, 2 who has made elaborate researches 
concerning this subject, believes that chlorotic blood exhibits an 
exaggeration of the fall in specific gravity which occurs in healthy 
girls at about the age of puberty. In the 36 cases studied by 
this author the specific gravity ranged from 1.030 to 1.049, these 
figures corresponding to 17 and 58 per cent, of hemoglobin, re- 
spectively, as estimated by the von Fleischl hemometer. In 30 
cases Hammerschlag 3 found that the density of the whole blood 
averaged 1.045, an d of the serum, 1.030. 

Most observers maintain that in this dis- 
ALKALINITY. ease the alkalinity of the whole blood generally 
remains normal, or suffers but a trifling diminu- 
tion, this being in direct contrast to the condition found in other 
forms of anemia, in which the fall in the alkalinity figure is usually 
pronounced. Burmin, 4 in 18 examinations of 9 cases, found that 
it ranged between 128 and 200 mgm. NaOH, the normal figures 
of this investigator being 182 to 218 mgm. In 6 of these cases 
the administration of iron was followed by a marked increase in 
the alkalinity of the blood, closely paralleling the gain in hemo- 
globin and erythrocytes. On the other hand, Graeber 5 states that 

1 Von Limbeck, loc. cit. 

2 Loc. cit. 3 Wien. med. Presse, 1894, vol. xliv, p. 1068. 
4 Zeitschr. f. klin. Med., 1900, vol. xxxix, p. 365. 

6 "Zur klin. Diag. d. Blutkrankheit.," Leipsic, 1890, p. 289. 



in many cases he discovered abnormally high alkalinity figures, 
so constantly, indeed, that he regarded them as "specific for this 
condition." Funke's studies, reported by Dare, 1 show that the 
alkalinity is diminished, and that it closely corresponds to the 
color index. 

The decrease in the percentage of hemoglobin 
HEMOGLOBIN is usually excessive in comparison with the re- 
AND duction in the number of erythrocytes, this dis- 

ERYTHROCYTES. proportionate oligochromemia being the most 
conspicuous and most constant feature of the 
changes affecting chlorotic blood. Naturally, such a change 
gives rise to very low color indices. This statement applies only 
to the majority of cases, for a low 
color index, while it is the rule 
in chlorosis, and is, diagnostic- 
ally, a most important feature of 
the blood picture, is by no means 
invariably found no more in- 
variably than a high color index 
in pernicious anemia. To illus- 
trate this point, of 106 consecu- 
tive cases of chlorosis studied by 
the author, 49, or more than 46 
per cent., showed an index below 
0.50, the average for the series 
being 0.51, the maximum i.oi, 
and the minimum 0.22. 

The average loss of hemo- 
globin, as evidenced by the 155 
cases tabulated below, amounts 
to about 53 per cent., in contrast 

to which stands the mean average erythrocyte decrease, which is 
equivalent to about 23 per cent., the hemoglobin loss thus averag- 
ing about two-and-one-half times that of the corpuscles. Indi' 
vidually, the hemoglobin percentage ranged in these cases from 
12 to 87, averaging 46.8, and the count of erythrocytes from 
1,720,000 to 5,600,000, averaging 3,816,486. On account of 
their rather close correspondence, it is interesting to compare 
with these figures the results obtained by Cabot 2 in 109 cases 
and those of Thayer 3 for 63 cases. Cabot's cases gave the 
following mean averages: hemoglobin, 41.2 per cent.; erythro- 

1 Johns Hopkins Hosp. Bull., 1903, vol. xiv, p. 179. * Loc. cit. 

3 Cited by Osier, "American Text-book of Theory and Practice of Medicine," 
Philadelphia, 1894, vol. ii, p. 196. 


Showing a general decrease in the diam- 
eter of the corpuscles, striking decolorization. 
and moderate poikilocytosis. The nucleated 
cell near the center of the. field is a normo- 


cytes, 4,112,000, with individual counts ranging from 1,932,000 
to 7,100,000. In Thayer's series the hemoglobin averaged 42.3 
per cent, and the erythrocyte count 4,096,544. Somewhat lower 
figures are given by Bramwell, 1 who found the following averages 
in a series of 80 cases : hemoglobin, 34 per cent., or from 10 to 60 
per cent.; erythrocytes, 3,437,300, or from 1,425,000 to 5,200,000 
per; and color index, 0.49, or from 0.20 to 0.96. 

While numerous examples may be found of typical cases of 
chlorosis in which the hemoglobin estimate and erythrocyte count 
resemble those commonly occurring in pernicious anemia or in 
the secondary anemias, nothing is more characteristic of chlorosis 
than the averages above mentioned. The great difference is 
between chlorosis and pernicious anemia, the index usually being 
high and the corpuscular loss extreme in the latter disease. As 
compared with the secondary anemias, the difference is too slight 
and its occurrence too inconstant to enable one to regard it with 
any degree of certainty from a clinical standpoint. Theoretically, 
in secondary anemia the hemoglobin loss is fairly proportionate 
to the erythrocyte decrease, thus producing color indices at or 
somewhat below the normal, but cases of secondary anemia having 
a so-called "chlorotic" type of blood are far too common to render 
any information reliable gained by a simple inquiry into the 
changes affecting the erythrocytes and their hemoglobin content. 





From 80-90 5 Above 5,000,000 13 

7~ IJ From 4,000,000-5,000,000 63 

00-70 17 

50 _6o ' 3,000,000-4,000,000 49 

40-50 '.'.'.'.'.'.'.'.'.'.'.'.'.'.'. 24 " 2,000,000-3,000,000 28 

30-40 25 " I,OOO,000-2,OOO,OOO 2 

20-30 28 


Average, 46.8 per cent. Average, 3,816,486 per 

Maximum, 87.0 " " Maximum, 5,600,000 " " 

Minimum, 12.0 " " Minimum, 1,720,000 " " 

The most conspicuous change to be observed in the stained 
film of chlorotic blood is the presence of large numbers of under- 
sized, pale erythrocytes, such cells usually being so numerous 
that one is forcibly impressed with the fact that there must be a 
general decrease in the average diameter of all the erythrocytes 
in the field. As a rule, this decrease in size involves a large 
number of corpuscles moderately, rather than a few to an ex. 

1 "Anemia," London, 1899, p. 35. 


treme degree, and therefore, except in severe cases associated 
with marked oligocythemia, striking examples of microcytosis are 
wanting. This alteration is just the opposite of what is generally 
found in pernicious anemia, for in this disease a tendency to- 
ward an increase in the average diameter of the erythrocytes, 
frequently in association with the presence of many extremely 
small microcytes, is the rule. If well denned, this feature of the 
blood changes carries a certain amount of diagnostic significance, 
although it cannot be distinguished in every case of chlorosis, 
since in some the diameter of the erythrocytes appears to be un- 
altered, while in others the deformities of size may so affect the cells 
that the blood picture resembles that of a severe secondary anemia. 

The pallor of the erythrocytes, shown by their feeble reaction 
toward the plasma stain, is at once apparent. The great majority 
of the cells are affected alike, being pale, often quite colorless in 
the center, and gradually becoming of darker color toward the 
periphery, in which a certain amount of hemoglobin still remains. 
This portion of the cell is usually well stained, so that the cor- 
puscles frequently appear as hoops or rings; some, however, do 
not show even this narrow hemoglobin-filled zone, being practi- 
cally decolorized throughout. Stroma degeneration, as shown by 
the changes described by Maragliano, is not demonstrable in the 
average case of moderate severity, but this process has been ob- 
served in occasional cases of high grade. Polychromatophilia, 
except in cases of the latter class, does not occur. 

Basophilie granulations in the erythrocytes are not present in 
this condition, according to Grawitz, 1 a finding which 'the author 
has substantiated. 2 Stengel, 3 on the contrary, found them in u of 
1 8 chlorotics. 

Deformities of shape are not noticeable, as a rule, except in 
the severer types of the disorder. In such cases, in which both 
the hemoglobin and the cellular losses are excessive, ppikilo- 
cytosis may be very striking as great, in fact, as in any blood 
disease, not excepting pernicious anemia. Poikilocytes, should 
they occur, are almost invariably of small size. 

Nucleated erythrocytes are very rare. In the average case 
they are usually sought for in vain, and even in the severer forms 
of chlorosis these cells are not numerous. Erythroblasts con- 
forming to the normoblastic type are found almost exclusively; 
megaloblasts, although they are seen now and then, are extremely 
uncommon, and have never been found in a large relative or 
absolute proportion to the other form of nucleated erythrocytes. 

1 Loc. cit. * Amer. Med., 1903, vol. v, p. 571. 

3 Amer. Jour. Med. Sci., 1902, vol. cxxii, p. 873. 





Above 20,000 i 

From 15,000-20,000 3 

" 10,000-15,000 14 

" 5,000-10,000 104 

Below 5,000 33 

Average, 6,457 per c. mm. 

Maximum, 21,000 " " 

Minimum, 800 " " 

The number of leucocytes per is, 
LEUCOCYTES, as a rule, normal in the typical case of chloro- 
sis. If leucocytosis occurs, as it does occasion- 
ally, it should be attributed to some hidden or frank complication ; 
if leucopenia exists, as it sometimes does, it may nearly always 
be regarded as a sign of the severity of the disease, since it is 
rarely met with except in cases in which the hemoglobin and ery- 
throcyte losses are decidedly marked. The mean average number 
of leucocytes in the 155 cases of chlorosis to which reference has 
already been made (Table II) is 6457 per, or approximately 
the same as the average count of these cells in normal blood. 
Counts as low as 800 and as high as 21,000 were made in this series ; 
and in 18 of the cases (or n.6 per cent.) the increase was suf- 
ficiently in excess of the normal standard to justify the application 
of the term leucocytosis that is, it was in excess of 10,000. These 
figures do not differ materially from those of Cabot and of Thayer, 
alluded to above, Cabot's counts in 104 cases averaging 7400, and 
Thayer's estimates in 63 cases being but slightly higher 7485. 

Relative lymphocytosis, usually marked in relation to the se- 
verity of the case, is a common, but not a constant, qualitative 
change. It occurs in both mild and severe cases, but is much 
more common in the latter. In the author's experience this in- 
crease involves chiefly the larger forms of these cells, both the 
non-granular mononuclear cells with spherical nuclei and the so- 
called "transitional" forms with indented nuclei; striking in- 
crease in the last-named variety of cells was a notable differential 
change in a large proportion of the 37 cases listed in Table III. 
In many of the cases in this series the large and small lympho- 
cytes together made up from 45 to as high as 67.5 per cent, of 
all varieties of leucocytes, the percentage of large forms being 
repeatedly estimated at 20 or 30, and even 40, in one instance. 

Deviations from normal in the relative percentage of polynu- 
clear neutrophiles are governed by the behavior of the lympho- 
cytes, low differential counts of the former type of cells accom- 



panying high percentages of the latter, and vice versa. Should 
leucocytosis exist, it is of the pure polynuclear neutrophile type. 
The eosinophiles are notably decreased, both absolutely and 
relatively. The author has never found an increase of these cells 
in chlorosis, although considerable pains were taken to verify the 
statements made by some writers that this variety of leucocytes 
is occasionally observed to be greatly above normal in this con- 
dition. Eosinophiles were absent entirely in more than seven- 
tenths of all the cases collected in Table III, and were never found 
to exceed two per cent, of all the forms of leucocytes. 















n. o 3.0 





20.0 3.5 





25.0 3.0 




19.5 20.5 





22.5 15.0 





20.5 19.0 




32.2 17.8 




25.5 16.0 




18.5 12.0 





18.0 10.0 




17.0 21.5 



I 3 

19.5 17.0 





22.0 14.5 

6 3-5 



7-5 12.4 

80. i 



19.3 21. I 


o.o * 



18.5 16.0 





18.0 16.0 





18.3 19.0 




15-5 24.5 





26.O 21. O 




12.0 3O.O 





13-5 12.5 


o-5 o . 


14.0 15.9 


0.0 . O 











27 20.5 


68 . o o . o o 




67.0 o.o o 










O.I O 





o.o o 

3 2 










I.O 2 










o.o o 












Average: 20.1 









Specific gravity. 


Exceptionally, small percentages of myelocytes may be en- 
countered, as a rule only in cases of a severe character. These 
cells are ordinarily present in not more than six per cent, of all 
cases, and their relative proportion to the other forms of leuco- 
cytes is always trifling, being rarely over one or two per cent. 

In the great majority of cases it has been gen- 

BLOOD erally observed that the number of plaques is 

PLAQUES. considerably in excess of normal. It appears 

that these elements are especially numerous in 

blood which clots rapidly. 

The changes in the blood associated with the 
DIAGNOSIS, well-defined case of chlorosis may be summarized 
as follows : 

Pale and watery. 
Usually rapid. 

Marked absolute decrease, in most instances 
relatively greater than the loss of erythrocytes, 
thus producing a low color index. 
Moderately decreased, ordinarily to about 
4,000,000 per Counts of 3,000,000 
or lower are common in severe cases. Ery- 
throblasts very rare; if present, cells of the 
normoblastic type invariably predominate. 
General decrease in the average diameter of 
the erythrocytes. In severe cases microcy- 
tosis may be marked. 

Poikilocytes not numerous, except in severe 

Polychromatophilia rare. 
Usually normal in number. 
Relative lymphocytosis common. 
Small percentages of myelocytes, only in severe 

Eosinophiles notably decreased. 
Increased in number. 

He who attempts the diagnosis of chlorosis solely by the blood 
examination is indeed a rash clinician. This point cannot be 
emphasized too strongly, that there is no blood picture peculiar 
to this condition, since changes precisely similar to those seen in 
many a case of typical chlorosis are often observed in the secondary 
anemias, especially in those dependent upon such factors as syph- 
ilis, septicemia, malignant disease, and chronic renal lesions. 
Futhermore, rare cases of chlorosis with typical symptoms 




have been reported in which no alterations in the blood were 
discoverable by ordinary clinical methods. In these cases, mis- 
termed "pseudo-chlorosis," a diminished volume of plasma may 
mask the blood impoverishment (Lloyd Jones 1 ), or this oligo- 
plasmia may be combined with both plasma and cellular hy- 
dremia (Biernacki 2 ), the dropsical erythrocytes being actually 
deficient in hemoglobin, although the hemoglobin and erythrocyte 
estimates of the whole blood remain normal. 

The blood changes enumerated above (especially such features 
as a low color index, the general decrease in the diameter of the 
erythrocytes, the absence or scantiness of erythroblasts, and the 
normal number of leucocytes associated with relative lymphocytosis 
and a decrease of the eosinophiles) are not then, pathognomonic, but 
simply highly suggestive of the disease under discussion, in view of 
which fact it becomes essential to seek for other clinical signs 
and to consider them carefully in connection with the blood find- 
ings. One of the most important points which should be borne 
in mind is the fact that chlorosis is practically confined to females, 
usually those in early womanhood, at or near the period of puberty. 
Chlorosis is about as compatible with the male sex as is pregnancy 
the so-called "male chlorosis" is nothing more than a diagnostic 
myth. Osier 3 remarks that in girls in whom the disease occurs 
early in their teens precocity and almost premature appearance of 
the menses are likely to exist. A large proportion of those in 
whom the disease develops later in life complain of scantiness or 
total suppression of the menstrual flow and of dysmenorrhea, these 
symptoms being especially common in chlorotics in. the early 
twenties or thereabouts. 

The question of heredity also is of some diagnostic value, for 
it has been frequently noted that the disease exists, for instance, 
in two or more sisters, the statement being elicited upon further 
inquiry that their mother suffered from chlorosis at an earlier 
period. Thus, Allbutt 4 speaks of meeting in his consulting-room 
the chlorotic daughters of women whom years before he had 
treated for the same disorder. 

Among the other manifestations of the disease to which atten- 
tion should be paid the following are the more important : a pecu- 
liar greenish-yellow color of the complexion and blanching of the 
mucous membranes (except in those rare instances of chlorosis 
florida, in which the color is high); the occurrence of various 
gastro-intestinal disturbances, of edema of the face and lower 

1 Loc. dt.. 

2 Zeitschr. f. klin. Med., 1894, vol. xxiv, p. 500. * Loc. cit. 
4 "System of Medicine," London and New York, 1898, vol. vi, p. 483. 


limbs, of vertiginous attacks, and of dyspnea upon physical exer- 
tion; and the presence of systolic basic heart murmurs and a 
venous hum most distinctly audible over the great vessels of the 
neck. Slight enlargement of the thyroid gland, frequently asso- 
ciated with Joffroy's sign (absence of horizontal wrinkling of the 
skin of the forehead and of upward curving of the eyebrows when 
the patient glances suddenly at the ceiling without elevating her 
head), is a physical sign which should always make one suspicious 
of chlorosis. 

The distinctions between chlorosis and pernicious anemia, as 
shown by the blood examination, will be described under the 
latter disease. (See p. 289.) 


Lorrain Smith's experiments 1 argue a de- 
GENERAL crease, averaging 48 per cent., in the total oxygen 
FEATURES, capacity of the blood, and also show that the 
blood -volume fluctuates greatly in different cases. 
Any decided increase, however, is at variance with both the clini- 
cal and the postmortem findings. 

Diminution in the albumin of the whole blood, due chiefly to 
the cellular poverty, is a conspicuous change, the dry residue in 
extreme instances amounting to but one-half of the normal figure. 2 
The albumin of the blood serum is also diminished, 3 but not greatly 
a point of difference between Addisonian and secondary anemia, 
since in severe types of the latter the serum albumin is strikingly 

In marked cases of pernicious anemia it is 
APPEARANCE sometimes almost impossible to obtain from a 
OF THE puncture of the finger-tip a sufficient quantity of 
FRESH BLOOD, blood for an ordinary clinical examination, owing 
to the bloodlessness of the superficial vessels. 
This fact naturally prompts the query, Does an actual reduction 
in blood volume or oligemia exist in such an instance, or can the 
dryness of the superficial tissues be attributed to vasomotor dis- 
turbances causing an unequal distribution of the blood mass, in 
favor of the internal viscera and deeper circulation ? In a patient 
in whom puncture of the finger fails to give the requisite amount 
of blood, the lobe of the ear will generally be found to yield a 
drop of sufficient size. But even this very vascular part of the 

1 Loc. cit. 

2 Gumprecht and Stintzing, Deutsch. Arch. f. klin. Med., 1894, vol. liii, p. 265. 

3 Diabella, ibid., 1806, vol. Ivii, p. 302. 


body may in extreme cases seem practically bloodless. The 
writer recalls a case of fatal anemia in which, in order to secure 
a drop of blood large enough to fill the lumen of an erythrocy- 
tometer only to the 0.5 graduation, it was necessary to open a 
small superficial vessel of the scalp, repeated deep punctures of 
the ringers, toes, and ear-lobes having given negative results. 

The drop as it emerges from the puncture wound is exceed- 
ingly pale, thin, and hydremic, lacking the characteristic opac- 
ity of healthy blood, and being of a fluidity and general color 
which have been likened to those of meat-washings. In an oc- 
casional instance the color of the blood may be practically normal, 
or, rarely, of a brownish-red or chocolate tint; but, as a rule, it 
resembles a watery, pinkish fluid, deficient both in depth of color 
and in density. It has been frequently observed that, after having 
stood for a short length of time, the drop shows a tendency to 
separate into two more or less distinct parts, consisting of a dark 
stratum of corpuscles and a clear, watery-looking layer of plasma; 
or it may be irregularly mottled at different points, as if the cor- 
puscles had become concentrated in isolated, compact groups in 
various parts of the plasma, thus producing the effect of alternating 
dark and light areas distributed through the drop. 

Microscopical examination of the fresh film shows a great re- 
duction in the number of the erythrocytes, together with the pres- 
ence of many forms of these cells which exhibit every possible 
variation in size and in shape. The color of the individual eryth- 
rocyte varies, some being normally dark and well colored, while 
others appear as mere washed-out rings or " phantoms. " In 
some of the cells the hemoglobin appears to be quite evenly dis- 
tributed throughout the stroma, so that their typical biconcav- 
ity is obliterated. The endoglobular degenerative changes and 
those structural alterations denoting total necrosis of the eryth- 
rocytes, previously described, may be demonstrated with great 
clearness in this condition. Rouleaux formation is either entirely 
absent, or incomplete and atypical. 

The erythrocytes are abnormally vulnerable, as shown by their 
increased isotonicity 1 and by the readiness with which their con- 
tained hemoglobin crystallizes. 2 Von Jaksch's analyses 3 point to 
a decided relative excess of cellular albumin. 

Owing to the extreme oligocythemia common in pernicious 
anemia, it is advisable in making the films to use a somewhat 
larger drop of blood than is chosen for making ordinary spreads, 

1 Von Limbeck, loc. cit. 

2 Copeman, Brit. Med. Jour., 1901, vol. i, p. 161. 

3 Zeitschr. f. klin. Med., 1893, vol. xxiii, p. 187. 


so that the field will not contain such a pronounced scarcity of 
cellular elements. 

The obvious fluidity of the blood, the defici- 
COAGULATION. ency of the fibrin network, and the slowness with 
which coagulation occurs are marked features of 
this disease. In fact, in some cases coagulation may be said not to 
occur at all, according to the experiments of Hayem 1 and others 
of the French school, as in the case quoted by Lenoble, 2 in which 
no clotting of a sample of arterial blood was observed even after 
a lapse of seventy-two hours after its withdrawal from the vessels. 
Many authors attribute considerable diagnostic value to this 
absence of clotting, and others go so far as to state that it ren- 
ders a patient suffering with pernicious anemia especially prone 
to troublesome hemorrhages, even from a slight finger-prick an 
accident which must be extremely rare, however, for it practically 
never complicates an ordinary clinical examination. 

The density of the whole blood is much 

SPECIFIC below the normal standard, specific gravities as 

GRAVITY. low as 1.027 having been reported. It is to 

be recalled that in cases with a high color index 

erroneous results may occur from attempting to estimate the 

hemoglobin percentage by Hammerschlag's table of equivalents, 

since the hemoglobin, in reality, is somewhat higher than the 

percentages corresponding to the specific gravity figures. (See 

P- I33-) 

Up to the present time the reaction of the 

ALKALINITY, blood in pernicious anemia has not been very 
thoroughly studied, but the work already accom- 
plished is sufficient to show that the alkalinity is much diminished, 
as in other severe anemias. That it may be strikingly below 
normal is shown by a case reported by Waldvogel, 3 who in one 
case estimated the alkalinity figure at 40 mgm., using Salkowski's 
method. This author has determined that the normal alkalinity 
for men ranges from 350 to 400 mgm., and for women from 300 
to 350 mgm. 

Both the percentage of hemoglobin and the 
HEMOGLOBIN number of erythrocytes are greatly diminished, 
AND the former, as a general rule, relatively less so 

ERYTHROCYTES. than the latter. Thus, inasmuch as the individual 
corpuscles contain often a normal or even an ex- 
cessive amount of hemoglobin, it follows that high color indices 
are common common but by no means constant, as seems to be 

1 Loc. cit. 2 "Charact. s6miol. du caillot et du serum," Paris, 1898. 

3 Deutsch. med. Wochenschr., 1900, vol. xxvi, p. 685. 





































r i 










Red, Hemoglobin. 

Black, Erythrocytes. 

Blue, Leucocytes. 


the current impression among many students, for although it is 
true that while the average color index is about i .00 in pernicious 
anemia cases, the same statement cannot always be applied to 
the individual case. The author's series of 81 cases (Table 
IV) showed, at the first examination, hemoglobin percentages 
varying from a minimum of 10 to a maximum of 70, with a mean 
average of 27.1; the color index of these cases averaged 0.99. 
During remissions, as the erythrocytes increase, it is common to 
find low indices, this peculiarity being especially conspicuous 
should the improvement in the patient's condition be rapid, 
since in such instances the corpuscular increase is relatively much 
more rapid than the gain in hemoglobin. In cases in which im- 
provement takes place more slowly, the color index is likely to 
remain higher, for here the corpuscles and the hemoglobin are 
more prone to increase proportionately along parallel lines. 




From 60-70 i Above 3,000,000 2 

50-60 2 From 2,000,0003,000,000 13 

1 40-50 4 1,000,000-2,000,000 41 

' 30-40 16 " 500,000-1,000,000 23 

' 20-30 30 Below 500,000 2 

' 10-20 28 

Average, 27.1 per cent. Average, 1,361,777 per 

Maximum, 70.0 " " Maximum, 3,240,000- " " 

Minimum, 10.0 " Minimum, 450,000 " " 

The oligocythemia is most striking, counts of from r,ooo,ooo 
to 2,000,000 erythrocytes per being not uncommon when 
the patient first comes under observation, the number of cells 
frequently diminishing to about 750,000 or even 500,000 later 
during the course of the disease. In Quincke's often-quoted 
case the remarkable count of 143,000 per was observed 
just before the death of the patient, an instance which is -almost 
paralleled by a case recorded by Hills, 1 in which the erythrocyte 
count fell to 155,760 one day before death. In the series just 
mentioned (Table IV) the count of erythrocytes per 
averaged 1,361,777, ranging between 450,000 and 3,240,000; 
this represents an average loss in corpuscular matter of some- 
what less than 75 per cent., the greatest decrease amounting to 
91 per cent, of normal a much more striking oligocythemia than 
is found in any other form of anemia. Thirty per cent, of the 
cases showed a count of 1,000,000 or low r er. 

1 Boston Med. and Surg. Jour., 1898, vol. cxxxix, p. 542. 



Periods of temporary increase in the hemoglobin and erythro- 
cytes, followed sooner or later by relapses, are commonly observed, 
the gain during such periods sometimes being very pronounced. 
Thus, in one of the cases tabulated above a gain of more than 
2,500,000 erythrocytes to the was noted during six weeks' 
time with a subsequent loss of over 1,000,000 cells in the fol- 
lowing eight days, the color index during this time ranging from 
1.25 to 0.74. Such stages of remission may or may not follow 
vigorous treatment by arsenic or other medicaments. Elder, 1 
by the use of antistreptococcic serum, caused a gain of 4,000,000 
cells per, while DeWitt, 1 by the same means, caused an in- 
crease in hemoglobin 
from 30 to 90 per cent, 
and in erythrocytes 
fr m 1,000,000 to 
4,960,000 in three 
weeks' time. 

These periods of 
improvement in the 
condition of the blood 
are generally associ- 
ated with an amelio- 
ration of the other 
clinical manifestations 
of the disease, the pa- 
tient's general condi- 
tion improving so sub- 
stantially that he be- 
gins to consider him- 
self on the high road 
to recovery, but in the 
course of time the old 
symptoms return, and 

the characteristic blood picture again becomes evident. In most 
cases death is preceded by extreme oligochromemia and oligo- 
cythemia, the hemoglobin often falling to 15 or 20 per cent, of 
normal and the erythrocyte count declining to 750,000 or less; in 
some cases, however, these losses are not so marked, and the count 
does not fall below 1,500,000 during the whole course of the disease. 
A prominent characteristic of the blood in pernicious anemia 
is the wide dissimilarity in the size of the erythrocytes, due to the 
presence of large numbers of megalocytes and microcytes; so 

1 Cited by Packard and Willson, Amer. Jour. Med. Sci., 1902, vol. cxxiv, 
p. 1015. 


Showing a general increase in the diameter of the cor- 
puscles, and marked poikilocytosis. The nucleated cell in 
the right of the field is a megaloblast. 


striking may this feature of the blood picture be that it is some- 
times difficult to find any two cells in the same field of the micro- 
scope which are of the same diameter. In the great majority of 
cases it will be found that the megalocytes distinctly outnumber 
the microcytes, to such an extent and in so large a proportion 
of cases that some writers consider this change an almost con- 
stant blood finding in this disease. Large erythrocytes, measur- 
ing slightly below or above 10 , in diameter, are very common, 
while those measuring in the neighborhood of 15 [J- or even 20 
are met with more rarely. Undersized erythrocytes, about 3 or 
4 // in diameter, are also numerous, but, as remarked above, 
much less so than those of larger size. The presence of small, 
dark-colored, spherical microcytes of this size (the so-called 
"Eichhorst's corpuscles"), once regarded as pathognomonic of 
pernicious anemia, is neither constant nor diagnostic of this dis- 
ease, since they are found in many other anemic conditions, and 
are absent in a large proportion of cases of true pernicious anemia. 
The fact that, of these alterations in the size of the erythrocytes, 
megalocytosis predominates, constitutes a sign of valuable diag- 
nostic significance. 

Poikilocytosis, to a more or less marked degree, is constantly 
observed, the conspicuousness of the deformities being in some 
cases extreme, while in others the change is a less notable feature. 
While marked poikilocytosis usually goes hand in hand with 
excessive diminution in the number of erythrocytes and in the 
amount of hemoglobin, the association of these three changes 
cannot be invariably counted upon, for in some cases, in spite of 
the fact that both oligocythemia and oligochromemia are marked, 
deformities in the shape of the corpuscles are but trifling. All 
varieties of erythrocytes, small, large, nucleated, and non-nu- 
cleated, may be deformed, so that the size of the poikilocytes 
varies from that of the smallest microcyte to that of the largest 
megalocyte. The kinds of deformity are of infinite variety, but 
it is still possible to designate certain well-defined forms which 
are especially common in this disease, these being the horseshoe 
form (Litten) and the oval form (Cabot), both of which varieties, 
while by no means peculiar to this condition, are found so fre- 
quently and in such abundance in pernicious anemia that their 
presence in the blood is at least highly suggestive. Of these 
two forms, the elongated, oval erythrocyte is found more con- 
stantly, and has been described in but a few other conditions. 
The author has been struck with the predominance of cells of 
this sort in three consecutive cases of purpura haemorrhagica, 
in one of which the deformitv was so marked that scarcelv a 


single normally shaped erythrocyte could be found in certain 
fields of the microscope. Cabot 1 cites Greene as noticing the 
same change in the blood of two patients in whom the tentative 
diagnosis of epidemic dropsy had been made. In addition to 
these well-defined varieties, many cells of other shapes, also met 
with in other severe anemias, are observed, notably those resemb- 
ling the form of a sausage, a spindle, or a club. (See Fig. 52, 
p. 280.) 

In the stained specimen the principal point of interest is the 
presence of nucleated erythrocytes, upon the character of which 
the diagnosis of pernicious anemia must depend. Erythroblasts 
are always to be found in this disease at some stage of its course. 2 
During remissions, however, they may temporarily disappear. 
Megaloblasts are of much greater clinical significance than nor- 
moblasts, and by a differential count will be found always to out- 
number them in every genuine case of pernicious anemia at 
some stage of the disease. This blood picture, which indicates 
a megaloblastic degeneration of the bone marrow, due in all proba- 
bility to the influence of some unknown but specific toxic agency, 
is associated with only two other conditions, namely, nitrobenzol 
poisoning and some cases of high-grade anemia due to Bothrio- 
cephalus latus infection. The predominance of megaloblasts 
over normoblasts in pernicious anemia is well illustrated by Table 
V, which shows that at the first examination the former type of 
cells outnumbered the latter in 26 of the 29 cases here collected. 
The average proportion of megaloblasts to normoblasts in this 
series is somewhat more than 2 to i, and in some cases the 
former were the only kind of erythroblast discovered. The 
total number of erythroblasts of all varieties averaged 220 per of blood, ranging from as low as 3 to as high as 924. 
Regarding this last statement, it should be remembered that it 
is not the actual number of nucleated erythrocytes, but their 
character, which is all important in the diagnosis of this disease. 
In 52 additional cases of pernicious anemia a predominance of 
megaloblasts was found in 50, either at the first or by later examin- 
ation, a total of 76 megaloblastic bloods in the 81 cases studied. 
Of 139 cases reported by Cabot, megaloblasts predominated in 
109 at the first examination, and in all but 3 of the remaining 30 
cases at a subsequent period. 

1 Loc. cit. 

2 In Ehrlich's "aplastic anemia," a fatal type with profound oligocythemia, 
leucopenia, and purpura, the marrow so utterly fails to compensate the blood loss 
that erythroblasts of all varieties are wanting. Such cases are obviously not typical 
pernicious anemia. 











924 693 




840 616 




544 5 12 




47 32 



368 207 

4 6 



336 240 

















250 180 


" 2 35 i?5 



12 224 168 


13 204 148 



14 200 160 



15 192 180 



1 60 96 

6 4 




























2 3 










2 5 




















Average: 220 4- 

146 + 

64 +. 

. 10 

Microblasts are rare in comparison to the other forms of nu- 
cleated erythrocytes ; in some cases they may be relatively nu- 
merous, but in the majority they are absent. They were noted 
in but 9 of the 29 cases tabulated above (Table V), their average 
number for the series being 10 to the In the differential 
count of nucleated erythrocytes microblasts should be totaled 
with normoblasts, of which they are simply degenerate forms, 
more or less stripped of their protoplasm, and hence irregular 
and ragged in outline. 

In addition to the foregoing types of erythroblasts, cells pos- 
sessing the characteristics of both the normoblast and the meg- 
aloblast may be observed in many instances. These atypical 
forms and their clinical significance have been described in a 
previous section. (See p. 192.) In certain corpuscles, both of the 
normoblastic and of the megaloblastic types, division of the nu- 



cleus into several parts may have occurred, and in rare instances 
evidences of true karyokinesis may be seen. Normoblasts show- 
ing complete or partial nuclear extrusion and separation of the 
nucleus into a clover-leaf design are not uncommon, although 
pictures of this sort are found much more frequently in leukemia. 
In many cases of pernicious anemia one cannot but be struck 
with the fact that the majority of these atypical forms appear as 
cells with a megaloblastic protoplasm and a normoblastic nucleus ; 
they are, in the author's experience, much more numerous in this 
disease than cells having a normoblastic protoplasm and a meg- 
aloblastic nucleus, the latter being more common in leukemia. 







































2.8 0.8 










I I.O 






























34.5 o.o 





39.6 o.o 



























77-S o-S 






45.0 8.0 






72.6 2.O 






54.0 i.o 






60.O 7-2 

2 O 





40.0 i.o 






60.8 o.o 











45- J 4-7 

38.3 1.0 










2,000 14.5 










o.o o.o 




12. 1 




























I8. S 




Average: 3,925 + 

26 + 

10 + 

58 + 

1 + 

i + 


Fluctuations in the total number of erythroblasts occur from 
time to time during the progress of the disease, these changes 
sometimes taking place with great abruptness, being of wide 
range and often carrying not the slightest clinical import. A 
marked increase usually but not invariably precedes and accom- 
panies a gain in the number of erythrocytes and in the percentage 
of hemoglobin; and a similar increase, usually associated with 
extreme diminution in the erythrocyte count, is commonly met 
with as a preagonal sign. 

Marked evidences of polychromatophilia are found in many of 
the erythrocytes, both of the nucleated and of the non-nucleated 
varieties. Such cells, when stained with Ehrlich's triacid mix- 
ture, instead of taking the normal orange color of the solution, 
stain some bastard tint, such as slate color, dull purple, or dirty 
gray. Others may show a peculiar, streaked appearance, and 
irregular, pale, unstained areas, while others are scarcely stained at 
all, the greater part of the protoplasm remaining an indefinite shade 
of dead white. Granular degeneration of the protoplasm is distinctly 
evidenced in some of the cells, this process being betrayed by the 
appearance through the stroma of granular areas showing a striking 
affinity for a basic stain, such as methylene-blue. These basophilic 
granules, which have already been described, are not peculiar to 
pernicious anemia, since they have been found in a large number 
of secondary anemias of severe type due to various causes. (See 
p. 194.) 

The same remarks apply to the presence of erythrocytes con- 
taining the "ring bodies" of Cabot. This author 1 reports having 
found them in 9 of 14 cases of pernicious anemia examined during 
the active stages, but he failed to detect them in 4 cases in the 
stages of remission. 




From 10,000-15,000 3 

" 5,000-10,000 25 

Below 5,000 53 

Average, 4527 per 
Maximum, 15,000 " " 
Minimum, 500 " " 

Leucopenia may be counted on in considerably 

LEUCOCYTES, more than one-half of all cases of pernicious 

anemia, a fact which stands in direct contrast to 

the anemias of secondary type, in which an increase in the number 

1 Boston Med. and Surg. Jour., 1904, vol. cl, p. 321. 


of leucocytes is more common. In an occasional case, especially 
in one in which the other blood changes are inconspicuous, the 
number of leucocytes is found to be normal; and, rarely, a mod- 
erate leucocytosis, attributable to some complication, exists. In 
the average case, however, these cells are distinctly below the nor- 
mal standard, and the degree of kucopenia is sometimes extreme, 
the number of cells occasionally falling to below 1000 to the 
cjnm.; in rare instances they may apparently be entirely absent, 
none being found after prolonged search through both the count- 
ing chamber and the stained film. In the 81 cases collected in 
Table VII the number of leucocytes averaged about the mean low 
normal count 4527 being the exact figure), and ranged, in the 
individual case, from 500 to as high as 15,000 to the c-mm. It is 
interesting to note, in connection with the preceding remarks, that 
feucopenia was found in 53 or 65.4 per cent, of these cases. 

The leucocyte count, except in the event of complications, 
roughly parallels that of the erythrocytes, falling cointidentally 
with the oligocythemia and rising again as the erythrocytes in- 
crease. (See chart, p. 279.) An exception to this general rule 
is found in the terminal leucocytosis which not uncommonly 
develops just before the death of the patient. 

Relative lymphocytosis is a common, but not a constant, find- 
ing in the differential count of the stained film. It seems to be 
more frequently associated with low than with high counts, 
although no hard-and-fast rule can be laid down regarding this 
point. In extremely leucopenic blood a noteworthy finding is 
the abnormally high percentage of large mononuclear non-gran- 
ular cells, a change which does not ordinarily take place in con- 
nection with leucocyte counts approaching the normal average. 
The combined percentage of both large and small forms of 
lymphocytes in Table VI averaged 37.8, individual counts vary- 
ing from 9.7 to 85 per cent. The writer has noted the 
frequent occurrence of small lymphocytes containing coarse 
basic granules. Such cells appear to be especially common hi 
this form of anemia. Preagonal rises in die leucocyte count 
are sometimes lymphocytic in character, resembling the blood 
changes seen in lymphatic leukemia, and sometimes purely poly- 
nuclear in type. The relative percentage of potynudear neutro- 
philes averages low (58.6 per cent, in the above series), but iso- 
lated counts show a considerable range; their relative proportion 
to the other forms of leucocytes is largely determined by the 
fluctuations in the percentage of lymphocytes. The eosinophilcs 
are almost invariably decreased, and not infrequently they are 
wholly wanting, a circumstance which was made note of in more 


than 1 8 per cent, of the cases in the present series, in which the 
average percentage of these cells was 1.68. In an occasional 
case their percentage is above normal, as in cases 9, 16, and 19 
in Table VI. 

In no other disease save the myelogenous form of leukemia 
are myelocytes so constantly found, but almost always in relatively 
small percentages. In the cases under consideration these cells 
were absent in only two instances, the average figure for the 31 
cases being 1.82 per cent. In a single case (number 7) the re- 
markably high estimate (for this disease) of 10 per cent, of myelo- 
cytes was made, for in the other cases in which myelocytes oc- 
curred their percentage ranged from 0.4 to 4.3. 

In the stained specimen it is common to find that the leuco- 
cytes, particularly the polynuclear neutrophiles and the myelo- 
cytes, are of smaller size and more deeply stained than they ap- 
pear in normal blood. This peculiarity seems to be more con- 
stant and more striking in pernicious anemia than in any other 

The number of blood plaques is exceedingly 

BLOOD variable, so that it is impossible to make definite 

PLAQUES. statements regarding either the increase or the 

decrease of these bodies. In some cases they 
apparently are greatly increased, as evidenced by the groups of 
agglutinated masses of these cells which are. sometimes seen 
(von Limbeck 1 ), but in other cases it is evident that their number 
is appreciably diminished (Hay em 2 ). Van Emden 3 supports the 
latter view. In one case this observer estimated tKeir* number at 
between 32,000 and 64,000 per 

In a typical case of pernicious anemia the 
DIAGNOSIS, blood picture upon which the diagnosis rests is 

as follows: 
Hemoglobin. Marked absolute decrease, but of relatively higher 

percentage than that of the erythrocytes, this 

giving rise to a high color index. Striking de- 
Erythrocytes. crease, commonly to 1,000,000 or less per 

Counts of about 500,000 are not uncommon 

during the later stages of the disease. Erythro- 

blasts constant, cells of the megaloblastic type 


Megalocytes and microcytes, the former prevailing. 

Poikilocytes, usually numerous and conspicuous. 


1 Loc. cit. * " Lemons sur Ics Maladies du Sang," Paris, 1900. 

3 "Bijd. t. d. ken. v. h. bloed," Leyden, 1896. 


Basophilic stroma degeneration striking in severe 

Leucocytes. Usually decreased; decided leucopenia common. 

Relative lymphocytosis in the majority of cases. 

Small numbers of myelocytes almost invariably 

Eosinophiles few, sometimes absent. 
Plaques. Variable. 

Usually the diagnosis of pernicious anemia presents no difficul- 
ties, and may be made by the examination of the blood alone, 
the association of marked oligocythemia, a high color index, leu- 
copenia, and erythroblasts, chiefly of the megaloblastic variety, 
constituting a typical group of blood changes the significance of 
which is unmistakable. 

It should be borne in mind, however, that these changes are 
not always present in every case when the patient first comes 
under observation, so that repeated and careful examinations of 
the blood are sometimes necessary before a diagnosis is possible. 
Of the above-named changes, the most important, from a clinical 
viewpoint, is the prevalence of nucleated erythrocytes conforming 
to the megaloblastic type. With the two exceptions already noted 
(bothriocephalus anemia and nitrobenzol poisoning) this "mega- 
loblastic blood picture " is seen only in pernicious anemia, and, 
what is more important, it occurs in every true case of this dis- 
ease sooner or later during its course. Inability to detect this 
important characteristic should be regarded rather as a reflection 
upon the thoroughness of the examiner's technic than as a con- 
tradiction of the truth of this statement. Erythroblasts are not 
always numerous in pernicious anemia, and painstaking and pro- 
longed study of several stained films may be necessary before 
this important feature is distinguishable. 

In those cases of doubtful nature in which the typical blood 
changes are not at once evident a tentative diagnosis may be 
made by taking into careful consideration certain other physical 
signs and symptoms which the patient presents. In such in- 
stances attention should be directed to such suspicious points in 
the clinical history as the existence of a severe anemia arising 
either idiopathically or without adequate cause, and pursuing a 
progressively unfavorable course, uninfluenced permanently by 
treatment; the presence of a light lemon-yellow tint of the skin, 
of retinal hemorrhages, of a peculiarly soft, smooth, flabby condi- 
tion of the skin, and sometimes of moderate febrile paroxysms 
and gastric disturbances; and the remarkable preservation of the 
patient's general nutrition and body- weight in comparison with 
the severity of the illness. 


The severe secondary anemias due to hemorrhage, to advanced 
syphilis, and to malignant disease, especially of the stomach, 
sometimes give rise to clinical symptoms which so exactly simu- 
late pernicious anemia that the diagnosis must rest upon the re- 
sult of the blood findings, which are usually well enough marked 
to differentiate the conditions. It is true that in these conditions 
ample proof of sufficient etiological factors for the production of 
the anemia is generally at hand, and this fact should have im- 
portant bearing in ruling out anemia of the pernicious type, but 
it is also equally true that in malignant disease it is sometimes 
impossible to demonstrate the lesion, and that in syphilis the 
clinical history may be obscure, so that the blood examination 
must, after all, often be depended upon for an accurate diag- 
nosis. In secondary anemia from the above causes the oligo- 
cythemia is seldom so excessive as it is in pernicious anemia, the 
erythrocytes rarely falling as low as 1,000,000 per; the 
oligochromemia is likely to be relatively greater than the oligo- 
cythemia, so that a lower color index results; leucocytosis is 
not uncommon; and while deformities of shape and size and 
nucleation of the erythrocytes are frequently present, in some 
instances to as great an extent as in pernicious anemia, megalo- 
cytes do not predominate, nor do megaloblasts ever outnumber 

From chlorosis, which sometimes possesses many clinical mani- 
festations in common with pernicious anemia, the diagnosis may 
usually be readily made by the blood examination, which shows 
decided differences between the two diseases. In a typical case 
of chlorosis the deterioration in the quality of the blood affects 
chiefly the hemoglobin content of the erythrocytes and not the 
cells themselves. Hence it is common to find in this disease 
extreme oligochromemia out of all comparison with the more 
moderate oligocythemia, and consequently a low color index 
just the reverse of the condition found in pernicious, "anemia. 
Deformities in the shape and size of the erythrocytes are not 
uncommon in chlorosis, but they are not likely to be con- 
spicuous; the prevalent change affecting their shape is micro- 
cytosis of a moderate grade, so that a general decrease in the 
diameter of these cells is commonly observed. The most im- 
portant information derived from the blood is, however, of a 
negative character, consisting in the fact that nucleated erythro- 
cytes, should they be present, are chiefly normoblasts. While 
it is true that an occasional megaloblast may be encountered in 
rare instances, no chlorotic bloo.d has ever been known to show 
a predominance of this type of cells. The behavior of the leuco- 


cytes in chlorosis is of no aid in the differentiation of this condi- 
tion from pernicious anemia, for in both diseases the count is 
usually low and relative lymphocytosis common; in the former, 
however, the pronounced leucopenia of the latter condition is not 
often found. Myelocytes, while they may occur in both diseases, 
are much less common in chlorosis. 

In bothriocephalus anemia the expulsion of the parasite by the 
administration of an appropriate vermifuge is soon followed by a 
radical change in the blood picture and other symptoms, the meg- 
aloblasts disappearing, the hemoglobin and erythrocytes quickly 
rising to the normal standard, and the patient's health becoming 
entirely restored. The history of a patient suffering from anemia 
due to nitrobenzol poisoning is sufficiently characteristic to exclude 
true pernicious anemia. The differential diagnosis between this 
disease and splenic anemia is considered in another place. (See 
p. 295.) 

Several reputed instances of the conversion of pernicious 
anemia into leukemia have been reported by Leube and Fleischer, 1 
Waldstein, 2 and Litten. 3 Some such cases probably belong, 
as also do many cases of acute leukemia, to Leube's symptom- 
complex, " leukanemia " high-grade anemia plus either myelemia 
or lymphocytosis, with fever, hemorrhage, stomatitis, and hyper- 
plasia of the spleen and lymphatics. That pernicious anemia is 
ever converted into leukemia is questionable, and it seems reason- 
able to regard such a change as apparent rather than real. For 
example, in a case of leukemia, transient disappearance of the 
myelemia, with persistence of the anemia, such as may occur either 
spontaneously or from treatment, gives a blood picture not 
unlike that of true pernicious anemia, and a case of this sort may 
appear to become converted into leukemia, as the temporarily 
suppressed leukemic blood changes redevelop in course of time. 
A marked leucocytosis engrafted upon pernicious anemia may 
also be mistaken for the transformation of this disease into leu- 
kemia, but it is obvious that this counterfeit can be detected by 
a differential count of the leucocytes. 

1 Virchow's Arch., 1881, vol. Ixxxiii, p. 124. 2 Ibid., 1883, vol. xci, p. 12. 

3 Berlin, klin. Wochenschr., 1877, vol. xiv, p. 256. 

4 Munch, med. Wochenschr., 1900, vol. xlvii, p. 1121. See also Arneth, Deutsch. 
Arch. f. klin. Med., 1901, vol. Ixix, p. 331; Luce, ibid., 1903, vol. Ixxvii, p. 215; 
F. B. Weber, Brit. Med. Jour., 1904, vol. i, p. 1416; Kormoczi, Deutsch. med. 
Wochenschr., 1899, vol. xxv, pp. 238 and 775; Hitschman, Zeitschr. f. Heilk., 
1903, vol. xxiv, p. 190. 



There is nothing distinctive about the appear- 

APPEARANCE ance of the drop of freshly drawn blood, the 

OF THE color and density of which vary with the inten- 

FRESH BLOOD, sity of the anemia present. The author has 

notes of a case of splenic anemia in which it was 

remarked that, from its color and general appearance, the blood 

drop resembled precisely that obtained from a typical case of 

high-grade pernicious anemia; in a second case the color and 

opacity were but slightly below normal. 

No reliable observations have thus far been made regarding 
such minor points as the rate of coagulation, the specific gravity, 
and the reaction of the blood in this form of anemia. 

Decided, often extreme, anemia is the general 
HEMOGLOBIN rule in this disease. In the early stages the 
AND hemoglobin loss is relatively excessive as com- 

ERYTHROCYTES. pared to the decrease in erythrocytes, so that 
the color index is consequently low usually ap- 
proximating the figures found in many cases of high-grade sec- 
ondary anemia, but not averaging so low as in chlorosis. As the 
disease increases in severity, however, the color index tends to 
rise, as in pernicious anemia, this change being illustrated by the 
counts tabulated below. 

In a series of 15 cases reported by Osier 1 the following re- 
sults were obtained: the hemoglobin in 13 cases averaged 47 
per cent., the lowest estimate being 23 and the highest 60 per 
cent.; 42 erythrocyte counts averaged 3,336,357 per, with 
extremes of 2,000,000 and 5,200,000. These figures are very 
closely approximated by the averages of 35 cases collected by 
Lichty 2 : hemoglobin, 47 per cent.; erythrocytes, 3,293,000; and 
leucocytes, 5594. Similar changes were found in series of cases 
reported by Rolleston 3 and by Frederick Taylor. 4 

In a case of splenic anemia in Professor Hare's ward at the 
Jefferson Hospital the writer found the following changes in five 
consecutive counts: 


March 7,1898 45 per cent. 0.82 2,750,000 

March 14, 1898 40 " " 0.73 2,725,000 

March 22, 1898 43 " " 0.76 2,812,000 

March 29, 1898 45 " " 0.69 3,275,000 

April u, 1898 40 " " i.oo 2,000,000 

1 Amer. Jour. Med. Sci., 1900, vol. cxix, p. 54; ibid., 1902, vol. cxxiv, p. 781. 

1 Jour. Amer. Med. Assoc., 1904, vol. xlii, p. 528. 

1 "Splenic Anemia," London, 1902. 

4 Lancet, 1904, vol. i, pp. 1477, 1554, and 1636. 


Deformities affecting the size of the erythrocytes, sometimes 
tending toward striking megalocytosis, may be met with in cases 
characterized by great oligocythemia, such an alteration being also 
associated with a greater or less degree of poikilocytosis, and with 
signs of stroma degeneration. Nucleated erythrocytes, although 
they occur infrequently, may be present in enormous numbers 
in severe cases, creating a blood picture which is distinguishable 
from that of true pernicious anemia only by the fact that normo- 
blasts predominate. Thus, in one of McCrae's counts in a case 
of Osle'r's, in which the hemoglobin was reduced to 20 and the 
erythrocytes to 27.6 per cent., no fewer than 75 erythroblasts 
(of which 21 were normoblasts, 19 megaloblasts, and 35 "inter- 
mediate" forms) were seen while counting 400 leucocytes. In 
the case above summarized the average number of erythroblasts 
per 1000 leucocytes was estimated as 67 (the maximum and 
minimum being 128 and 9, respectively) for the five examinations, 
41 of these cells being normoblasts and 26 megaloblasts. The 
presence of nucleated erythrocytes in such large numbers as were 
found in these two instances must, however, be regarded as most 
exceptional. Polychromatophilia of many of the erythrocytes 
may be a striking feature in advanced cases, but in those of a 
milder grade the phenomenon is absent. Absence of basophilic 
granular degeneration of the cells has been noted by Cohn. 1 

The transition of splenic anemia into lymphatic leukemia has 
been recorded twice in medical literature, according to Zypkin, 2 
who himself encountered such a change. 

Leucopenia, sometimes pronounced, is the 
LEUCOCYTES, characteristic finding, counts of from 2000 to 
4000 cells to the being common; as in 
pernicious anemia, the lowest leucocyte counts are generally 
associated with those cases in which the anemia is most intense. 
In the adult leucocytosis occurs only as the effect of some compli- 
cation, and therefore is but occasionally encountered. In Osier's 
series, above referred to, the number of leucocytes, determined in 
14 cases, averaged 4520 per, ranging from 2000 to 12,497, 
the latter estimate being the only one exceeding 10,000; in 9 of the 
cases the count fell below 5000. In the writer's case the five 
counts averaged 2400, varying from 1000 to 4000. In children, 
also, leucopenia .is the rule, in spite of a child's tendency to 
develop leucocytosis upon the slightest provocation. (See p. 

1 Munch, med. Wochenschr., 1900, vol. xlvii, p. 618. 
3 Wien. klin. Wochenschr., 1903, vol. xvi, p. 577. 


No constant differential changes have been observed, but rela- 
tive lymphocytosis is not infrequent, sometimes involving chiefly 
the large, and sometimes the small, forms of these cells. The pro- 
portion of both combined may be as high as 50 or 60 per cent., 
an increase of this kind bringing about a consequent fall in the 
relative percentage of polynudear neutrophiles. Small numbers 
of myelocytes, rarely in excess of a fraction of one per cent., are 
to be expected in cases with decided oligocythemia. The eosin- 
ophiles remain at about the normal standard. Typical coarsely 
granular mast cells are sometimes found in relatively large numbers 
as high as 5 or 6 per cent, of all forms of leucocytes. 

No special observations concerning the behavior of the blood 
plaques in this disease have been recorded up to the present time. 
It is evident, however, that they are not notably increased in 

To recapitulate, the blood changes which have 
DIAGNOSIS, been most frequently found in splenic anemia 

may be tabulated as follows: 

Hemoglobin. Marked diminution; color index variable. 
Erythrocytes. Usually reduced moderately, sometimes exces- 
sively. Counts between 3,000,000 and 4,000,000 
cells per are most common. 
Deformities of shape, especially megalocytosis, 
and poikilocytosis common in advanced cases. 
Erythroblasts rare, except in cases with decided 
oligocythemia. Normoblasts invariably predom- 

Polychromatophilia in severe cases. 
Leucocytes. Leucopenia the general rule. 

Relative lymphocytosis common. 
Small numbers of myelocytes in advanced cases. 
Relatively large percentages of mast cells *not 
Eosinophiles normal. 
Plaques. Not increased. 

Many writers still hesitate to assign to splenic anemia the r61e 
of a definite clinical entity, choosing to regard the condition 
either as a splenic form of Hodgkin's disease or as high-grade 
secondary anemia with marked splenic hyperplasia. But of late 
the majority of observers lean toward at least a tentative recognition 
of the condition as a distinct although an obscure disease. 1 Osier, 2 

1 For a critical re'sume" of the literature on splenic anemia, Sippy's article in 
the American Journal of the Medical Sciences, 1899, vol. cxviii, p. 570, should be 
consulted. 2 Loc. cit. 


well expresses the consensus of opinion when he remarks that 
"provisionally, until we have further knowledge, it is useful to 
group together . . . cases of idiopathic enlargement of the spleen 
with anemia without lymphatic involvement, and to label the 
condition splenic anemia." Band's disease, or the terminal stage 
of splenic anemia, is characterized by hypertrophic liver cirrhosis 
with jaundice and by ascites. 

It is quite obvious, from a glance at the above synopsis of the 
blood condition, that splenic anemia presents no characteristic 
blood picture by which the diagnosis can be made, so that in 
order to differentiate it from a number of other diseases which it 
more or less closely simulates, careful study of other clinical 
features is essential. 

The onset of splenic anemia is gradual and insidious, its course 
is prolonged often for a number of years, and its termination is 
ultimately fatal. The principal clinical features are the leuco- 
penic anemia, the great splenic tumor, and the absence of en- 
largement of the superficial lymphatics. In some instances the 
anemia develops in advance of the splenic tumor, but it is more 
often the case that the enlargement of the spleen is the earliest 
demonstrable lesion. The anemia is responsible for such symp- 
toms as dyspnea, vertigo, cardiac palpitation, loss of strength 
and appetite, and the occurrence of unexplained, irregular periods 
of fever; and for such signs as hemic heart murmurs, pallor, 
lemon-yellow discoloration of the skin and mucous membranes, 
and sometimes pigmentation of the skin. The enlarged spleen 
may extend as far down as the umbilicus, and sometimes far be- 
low this point, even to the iliac crests; the surface of the organ 
is smooth and free from nodules, its consistence is firm, and its 
shape is unaltered. It may give rise to no symptoms, but occa- 
sionally it is the cause of great pain and of hematemesis, the latter 
being due to simple mechanical congestion. Epistaxis, purpura, 
and hematuria have also been observed. Ascites sometimes de- 
velops, as the result either of the splenic enlargement or of the 
anemia. Enlargement of the liver, usually associated with de- 
cided icterus, occurs as a terminal symptom, and such gastro- 
intestinal disturbances as anorexia, nausea, vomiting, and both 
constipation and diarrhea are extremely common. Splenic anemia 
may prove fatal within six months after the onset of the initial 
symptoms, or it may drag along for as many years, but as a 
general rule its duration does not exceed two or three years. 
In one of Osier's cases the condition probably lasted for at least 
twelve years, and in a case recently treated in the Jefferson 
Hospital the splenic tumor and the anemia existed for six years, 


if not longer. As in pernicious anemia, periods of remission dur- 
ing which the leading symptoms disappear and the quality of the 
blood improves are commonly observed in this condition. 

The myelogenous form of leukemia, pernicious anemia, and 
Hodgkin's disease with splenic enlargement all present clinical 
features counterfeiting more or less faithfully splenic anemia, but 
the differential diagnosis between these conditions does not in- 
volve any great difficulty. The result of the blood examination 
gives the clue to the two diseases first named, the myelocytic 
type of blood in leukemia and the predominance of megaloblasts 
in pernicious anemia being sufficient to fix the identity of these 
conditions. In Hodgkin's disease with enlargement of the spleen 
there is more or less marked enlargement of the superficial lym- 
phatic glands, and the splenic tumor rarely attains the size to 
which it grows in splenic anemia; the blood picture of the two 
conditions, it must be recalled, may be identical. 

Enlargements of the spleen due to such factors as chronic ma- 
larial injection, amyloid disease, malignant growths, echinococcus 
cysts, and hepatic cirrhosis also occasionally require differentiation 
from splenic anemia. A history of previous attacks of malarial 
fever and the detection of the specific parasite or of pigment in 
the blood will serve to distinguish tumors of the spleen of malarial 
nature. In amyloid disease a history of long-standing suppura- 
tion, of tuberculosis, or of syphilis, and the presence of signs in- 
dicating amyloid involvement of other organs, nota'bly the liver, 
kidneys, and intestines, are the chief differentiating features. In 
malignant disease of the spleen the tumor is uneven, .irregular, 
and nodular, evidences elsewhere of malignant lesions generally 
exist, and a well-defined leucocytosis is common. Echino- 
coccus disease of the spleen pursues a protracted course unac- 
companied by signs of anemia, but generally shows decided 
eosinophilia, and, unless secondary infection takes place, is unas- 
sociated with rises in temperature; fluctuation of the tumor can 
frequently be detected, and hooklets can be recognized in the 
fluid obtained from the organ by aspiration. In splenic enlarge- 
ments associated with the different varieties of hepatic cirrho- 
sis, the previous history and the cachexia of the patient, the 
relatively moderate size of the tumor, the signs of portal con- 
gestion, the condition of the liver, and the course of the disease 
should be taken into account. 



An approximate idea of the intensity of the 
APPEARANCE anemia may usually be formed by noting the 
OF THE gross appearance of the fresh blood drop, but it 
FRESH BLOOD, must be remembered that it is only when the 
process has reached a comparatively high grade 
of development that the fact is betrayed by any marked devia- 
tion from normal in the color and density of the blood. In the 
average. case of well-marked secondary anemia the color of the 
drop is but slightly paler than normal, if, indeed, it is visibly al- 
tered; but if the anemia is of decided severity, it may resemble a 
thin, serum-colored liquid streaked with crimson, similar to the 
watery blood drop of typical pernicious anemia. In such cases 
microscopical examination of the fresh film shows that there is 
little or no tendency toward rouleaux formation. 

In general terms it may be stated that the 

COAGULATION, rapidity of coagulation bears a direct relation 

to the grade of the anemia, since it has been 

determined that the greater the oligochromemia and oligocy- 

themia, the more rapid the process of clotting. In secondary 

anemias with erythrocyte counts under 1,000,000 Lenoble 1 found 

that coagulation was, as a rule, at least twice as rapid as normal. 

The specific gravity of the whole blood is re- 

SPECIFIC duced, a change which is dependent chiefly upon 

GRAVITY. the loss of hemoglobin. Sufficient reference has 

already been made to this subject in a previous 

section. (See p. 132.) 

The majority of authors maintain that the al- 

ALKALINITY. kalinity of the blood is decreased in relation to 

the degree of the anemia, and a large number of 

experiments in anemias due to various factors apparently justify 

this general belief. But several careful investigators, notable 

among whom is Low}', 2 have contradicted these reports, having 

found the alkalinity normal or even above normal in numerous 

cases. The author quoted, for example, calculated the alkalinity 

in various cases of secondary anemia at from 360 to 675 mgm. 

NaOH, as compared with his normal standard, 447 to 508 mgm. 

Taking the ordinarily well-developed case of 

HEMOGLOBIN secondary anemia as an example, it is found that 

AND the hemoglobin percentage and number of ery- 

ERYTHROCYTES. throcytes are both decidedly, though not strikingly, 

diminished. As the former usually shows a dis- 

1 Loc. cit. 2 Centralbl. f. d. med. Wissensch., 1894, vol. xxxii, p. 785. 


proportionately greater loss than the latter, subnormal color in- 
dices are the rule, ranging, say, from about 0.75 to 0.85. In 
anemias of severer type, such as those due to gastric cancer and 
to enteric fever, the losses frequently are much more exaggerated, 
and, in so far as the purely quantitative changes in the erythro- 
cytes and their hemoglobin equivalent are concerned, the blood- 
picture of true pernicious anemia may be counterfeited. In the 
anemias of syphilis, of tuberculosis, and of malignant disease in 
general the disproportionate hemoglobin loss may be so decided 
that the blood changes cannot be distinguished from those of 
chlorosis, and to this condition the much-abused term "chloro- 
anemia" has been applied. 

The fact must be emphasized that simply the hemoglobin esti- 
mate and erythrocyte count alone are absolutely uncharacter- 
istic in secondary anemias, for they may range in the individual 
case from slightly subnormal figures to an extreme degree of 
oligochromemia and oligocythemia. In a patient studied by 
von Limbeck, 1 for example, at one time the erythrocytes num- 
bered only 306,000 per, but ultimately perfect recovery 
ensued and the count rose to 4,280,000. But if averages are used 
as a basis for conclusions, it becomes evident that the hemoglobin 
diminution is less marked than in any other blood disease, and 
that the erythrocyte loss is also less than in any other form of 
anemia except chlorosis. Data based upon 200 examinations of 
various types of anemia by the writer give the following results 
regarding these points: 




Secondary anemia 

44.8 per cent-. 

27.1 per cent.* 


14.8 " " 

17.8 " " 


60.6 " " 

4X.4 " * " 

Pernicious anemia 

74."? " " 

76.0 "' " 

Examination of the stained specimen shows a variable degree 
of alteration in the shape, size, and general structure of the cells. 
In mild cases simple pallor of the erythrocytes and perhaps a 
few microcytes and moderately misshapen poikilocytes are the 
only changes to be observed, erythroblasts, polychromatophiles, 
and cells with basophilic stroma degeneration being entirely 
wanting. In severe cases, with excessive oligocythemia, a large 
proportion of the cells are either under- or over-sized, the latter 

1 Loc. cit. 


forms appearing to prevail in relation to the intensity of the ane- 
mic process; poikilocytosis and polychromatophilia are sometimes 
extreme, and evidences of Grawitz's stroma degeneration are 
found, together with a more or less abundance of nucleated eryth- 
rocytes, the majority of which conform to the normoblastic type. 
In most instances normoblasts only are present, but rarely an 
occasional megaloblast, implying a slight tendency toward a fetal 
type of hemogenesis, is also seen. The significance of erythro- 
blasts in anemia and the circumstances under which they are 
found have been discussed in a preceding section. (See p. 187.) 
Typical polynuclear neutrophile leucocytosis 
LEUCOCYTES, is common, but by no means constant, in the sec- 
ondary anemias, independent of their grade, for 
the cellular increase is provoked by a stimulation of the functional 
activities of the marrow, which vary according to the individual 
and to the nature of the exciting cause. The differential changes 
associated with such a leucocytosis (low percentages of lympho- 
cytes and eosinophiles, with, perhaps, a few myelocytes) have 
already been referred to in a preceding section. A moderate 
leucocytosis is especially common in the anemias of children, 
and in those symptomatic of inflammatory and suppurative condi- 
tions and of malignant diseases. While in other anemias, espe- 
cially those of chronic type, a normal leucocyte count or even 
leucopenia may be found, often in association with a relative 
lymphocytosis, as is frequently the case in the anemias of enteric 
fever and of tertiary syphilis. 

The plaques are usually increased, but appa- 

BLOOD rently without any constant relationship to the 

PLAQUES. degree of hemoglobin and erythrocyte loss. In 

some cases these bodies may number more than 

double the maximum normal standard, as in a case of anemia in 

a child with a tumor of the spleen, noted by von Emden, 1 in which 

an estimate of 829,000 to the was made. 

^ The principal blood changes found in second- 

UIAGNOSIS. . p 11 

ary anemia are as follows: 

Hemoglobin. Variable decrease, usually somewhat more marked 
than the erythrocyte loss; color index subnormal. 

Erythrocytes. Variable decrease. 

Erythroblasts, in severe cases; normoblasts out- 
numbering megaloblasts, which are rare. 
Deformities of shape and size, polychromato- 
philia, and basic staining of the stroma in severe 

1 Loc. cit. 


Leucocytes. Commonly increased; rarely leucopenia. 

Polynuclear neutrophiles usually increased, and 
lymphocytes and eosinophiles relatively dimin- 

Lymphocytosis in some cases, usually those of 
severe type and chronic course. 
Small numbers of myelocytes sometimes found. 

Plaques. Usually increased. 


Among the many underlying causes of acute 
ETIOLOGY, post-hemorrhagic anemias may be mentioned 
trauma, abortion, post-partum hemorrhage, epis- 
taxis, pulmonary tuberculosis, peptic ulcer, enteric jever, visceral 
carcinoma, hemorrhagic pancreatitis, and the rupture of an aneu- 
rysm, of a Fallopian tube during ectopic pregnancy, and of a mass 
of extensively varicose veins. Chronic hemorrhages, such as those 
resulting from diseases belonging to the hemorrhagic diathesis, 
from hemorrhoids, or from uterine diseases usually give rise to a 
much less decided blood loss than the first-named conditions, but 
in some instances these factors, if persistent, may be sufficient 
eventually to provoke anemia of great intensity. 

Reduction in the total volume of blood, or. 
EFFECT UPON oligemia, ensues as the immediate effect of an 
THE BLOOD, acute hemorrhage, and a count made* immedi- 
ately after the blood loss may show no reduction 
in the hemoglobin and corpuscular value, since the oligemia af- 
fects the liquid and cellular elements proportionately. As reac- 
tion sets in the system attempts to compensate for the loss of 
blood by the rapid absorption by the capillaries of large amounts 
of liquids from the tissues, so that the blood soon becomes highly 
diluted, or hydremic. This is evidenced by a proportionate dimi- 
nution in the hemoglobin percentage and erythrocyte count, the 
degree of this decrease depending upon the extent of the hem- 
orrhage. It is thought that in many instances this fluid transfer 
from tissue to vessel is inaugurated immediately after or even 
during the hemorrhage, and that the original volume of blood is 
restored within a few hours. A further diminution in hemo- 
globin and erythrocytes occurs after the normal volume of blood 
has been reestablished, so that the minimum decrease is not ob- 
served until some little time has elapsed after the hemorrhage. 
As a rule, the minimum count is seen at some period during the 


first week after the blood loss as early as the first or second day 
in some instances, but as late as the tenth or eleventh day in 
others. This secondary fall is thought to depend upon the in- 
troduction into the circulation of large numbers of immature, 
feebly resistant erythrocytes, which suffer rapid and premature 
destruction, and thus bring about a disturbance in the equilibrium 
between the rate of blood production and blood destruction in 
favor of the latter. As soon as the marrow is able to meet the 
drain in an adequate manner, by the increased production of more 
resistant 'cells, the anemia ceases, and the hemoglobin and eryth- 
rocyte estimates begin to rise. 

Other changes consequent to hemorrhage are a diminution in 
the corpuscular volume in the dry residue of the whole blood, 
and in the proteids of the serum. The serum solids and fibrin- 
forming elements are apparently increased. Despite the loss of 
albumin, Haesslin 1 found a constant fall in the freezing-point of the 

Authorities differ as to the degree of blood loss which man is 
capable of surviving, a difference which is but natural when it is 
remembered that factors other than the actual amount of blood 
lost, conspicuous among which are the age, sex, and resisting 
powers of the patient, are all important in determining the fatality 
of the hemorrhage. According to Immermann, 2 hemorrhages in- 
volving a loss of one-half of the total bulk of blood in the body 
invariably prove fatal. Hayem 3 is authority for the statement 
that, as a general rule, recovery is possible when the total volume 
of blood lost does not exceed one-eighteenth of the individual's 
body- weight. This author has reported the most astonishing 
example on record of post-hemorrhagic cellular decrease, in 
which he observed a diminution in the erythrocytes to n per 
cent, of normal in a case of post-partum hemorrhage, with sub- 
sequent recovery of the patient. Behier 4 has described a case 
of metrorrhagia in which recovery occurred in spite of a reduc- 
tion in the erythrocytes to 19 per cent, of normal. Laache 5 has 
recorded a number of instances in which the corpuscular esti- 
mates fell below 50 per cent, of normal in one case to 32 per cent. 
These last three examples are sufficient to disprove the former 
belief that death inevitably ensues when the corpuscular loss, as 
the result of hemorrhage, falls as low as 50 per cent, of normal. 

Increase in the number of leucocytes, usually of moderate de- 
gree, promptly develops in the great majority of cases and per- 

1 Deutsch. Arch. f. klin. Med., 1902, vol. Ixxiv, p. 577. 

2 Cited by Rieder, loc. cit. 8 Loc. cit. 

* Cited by Laache, loc. cit. 6 "Die Anamie," Christiania, 1888. 


sists for several days. It usually involves an absolute and rela- 
tive gain in the polynuclear neutrophile cells, with a consequent 
decrease in the mononuclear forms, but, rarely, the reverse may 
be noted. In fatal cases this increase may not occur; nor, ac- 
cording to Baumann, 1 does it occur after hemorrhage during the 
adminstration of arsenic, the leucocytes in such instances diminish- 
ing. This investigator also found that the giving of arsenic and 
inorganic iron to animals experimentally bled resulted in slighter 
blood deterioration than when either drug was used alone. 

The maximum count is commonly attained within a few hours 
after the onset of the leucocytosis, the normal being regained 
within a week or less. (See p. 246.) 

The blood plaques are strikingly increased after hemorrhage. 
The coagulability of the blood is abnormally quick, being more 
rapid in profuse than in moderate hemorrhages. 

Following the reestablishment of the normal 
REGENERA- blood volume, regeneration of the erythrocytes 
TION. and hemoglobin and a consequent dissipation of 
the hydremia ensue. The time necessary for 
the completion of this process varies greatly in different individ- 
uals, as the rapidity with which blood regeneration occurs depends 
upon different factors, such as the extent of the original hemor- 
rhage and the age and natural regenerative powers of the patient. 
The latter are at their maximum during the thircl and fourth 
decades of life, at their minimum during infancy and old age, and 
are regarded as more active in women than in men. The exist- 
ence of a well-developed cachexia or an infectious disease, as 
well as the neglect of proper treatment of the hemorrhage, are 
obstacles which retard the regeneration of the blood to its normal 
composition. The process appears to be more active if trans- 
fusion of a normal saline solution has been practised than in un- 
treated cases, the rapidity of the gain being especially striking 
during the latter half of the regeneration period. The transfusion 
of blood hastens regeneration even more decidedly. Otto 2 and 
Hall and Eubank 3 have shown experimentally in animals, bled 
and given transfusions of artificial serum, that regeneration once 
stimulated into activity may carry the blood, quantitatively, con- 
siderably beyond the established normal standard. 

In uncomplicated cases, according to Bierfreund, 4 regeneration 
is effected within four weeks if the hemorrhage produces a hemo- 

1 Jour. Physiol., 1903, vol. xxix, p. 18. 

2 Pfliiger's Arch., 1885, vol. xxxv, p. 57. 

3 Jour. Exper. Med., 1896, vol. i, p. 656. 

4 Langenbeck's Arch., 1890-91, vol. xli, p. i. 


globin loss of 25 per cent., and in about three weeks if the loss 
does not exceed 20 per cent. The latter period may be regarded 
as the average regeneration time in the great majority of instances. 

As regeneration proceeds the hemoglobin and corpuscular 
deficiencies gradually become less conspicuous, but the increase 
in these two constituents does not occur along parallel lines. 
The increase in the number of erythrocytes is much more rapid 
than the gain in the hemoglobin percentage, which usually 
remains subnormal for some time after the normal number of 
corpuscles has been reestablished. Owing to this lagging behind 
of the hemoglobin low color indices are the rule. Faulty hemo- 
genesis, owing to which the great majority of the erythrocytes 
are deficient in hemoglobin and many of them of abnormally small 
size, serves best to explain this slow restitution of the hemoglobin 

The appearance in the blood of normoblasts is common after 
hemorrhage, and in rare instances an occasional megaloblast and 
atypical erythroblast may be observed. According to Ehrlich, 1 
if thorough and systematic search is made, normoblasts may 
be constantly found after the second or third day following the 
blood loss until the regeneration of the blood is complete. The 
transient appearance of large numbers of normoblasts, known 
as "blood crises," has been already described. (See p. 189.) 
Dawson, 2 who has carefully studied the effects of venous hemor- 
rhage in dogs, found no evidence of any close relation between 
the number of erythroblasts and the rapidity and character of 
the regeneration of the hemoglobin and erythrocytes. In severe 
cases polychromatophilia of the erythrocytes may be noted, this 
sign first becoming apparent as early as the first day after the 
hemorrhage, and gradually disappearing as regeneration is effected. 
Deformities in the size and shape of the erythrocytes are not 
uncommon, of which microcytes constitute the most frequent 
example. Large hydropic megalocytes and poikilocytes are met 
with more rarely. 


According to the classification in general vogue 
VARIETIES, at the present time two clinical varieties of leu- 
kemia, the myelogenous or spleno-medullary and 
the lymphatic, are recognized. The myelogenous variety, which 
is almost invariably a chronic process, is associated with a 

1 Loc. cit. 2 Amer. Jour. Physiol., 1900, vol. iv, p. 2. 

5 i 

i B 6 .s a 

^ CO *^ 


^ c M 5 J 


(j tiXJ Q 

2 js u ^ -d 



1 1. 

b S tn 

2 Pi t/3 

> _: . w .S -0 

' J| i fcrtrtS K 
<8 "S- >C1> 4j 


i || 1 1 1 


O *4J 

S 6| cjllf 1 



t3 3 o * O C 




O | o" 

H ^ "^ aj C^'U " 
o O _tn tn J; O "O 

W)-2 -^2 ^2 <n '" 


en 0*3, 9 2 -H 

1 l! 1 ^j 



"cd ** ? 

*J ^u ^j^Pcneo 
coz ^^S^ijnJ^* 


^ tj .+- t .O-^Jc^ 



rti '^* dl 


^3 2>^^^^^i p 


o cd 

^SpSp^ w|(2l|6| 

u C D. 5) > 

3 ^ 

1 tllPll 

i- 1 "i^ot-c 

o u o .5 



io 2 

g S O , i "o 


-a o - i _i 


cJ ^* *n ~ 


o *".S2 i* co 

rt C -15 

" C C 4> 

*-* "ti 55 cd * f 




S, . j.Ji| 

rt Q 3 S 4) 


* M.6. 
4, O. . 

R Z?S t '"^'D, t "'E ; 


b .2 8 

^ 3 2 -" f3 'a -S -^ Sj g ^ 


Q-r-g'o3cn ^ ^ 



4> * O 

*O 2 rt Q j c ^ '"^ i^ *-^ cj 




Marked d 
ing to about 
erythrocyte '. 

sg.ll lilfiil 


| uSn!OoC'So.- | ujs_M 

a u ,0 ,n be u 

about 0.75. 

111 if! 1 !l 

O 'O O u 



60 ' 

<n ' ' i C 

bo -a ' *j e 

.S 2 

^ C oD O Q 

g Q 3 g J> 

'Sb -5 

3 8 C ^ '*-' 

'So rt c & - 

O o - 2 


rt S g TJ 

4* h 

K- ^ ."9 1 e 

S -8 s ^ ^ 1 



O_ ^ g ^jj C C o 


> r? ^ C 



iTo" js ^.S^^ 3 


^ -H ~ c Sj -2 1) 


aT ** ^* 

J CN*_i M <0 *i .2 W g 


j o* 3 S 82 <J 


% . 

g H ^ 3 g|| 8 


c ji S. *3, 5 ! s 


<u u bo 

S ** a "cs < 


o ^'o ^ * <M 


8g. 2 


M " . 

"111 Slil 1 

.>^l^ .iali .f 




f A Hull 


C *J .fj w) 



3 OJ2J2 
G"TJ . 


2 b^ 

cj """-" 



nJ ^1 *- 3 cd 




Marked decrease, averag- 
ing to about 50 per cent. 
Relatively low to erythro- 
cyte loss. 

Moderate decrease, counts 
averaging about 4,000,000. 
Pallor conspicuous. 
Poikilocytosis rarely 
Erythroblasts rare, normo- 
blasts predominating. 
Polychromatophilia rare; 
basophilic stroma degenera- 
tion absent. 
Microcytosis common. 





b . 


Normal or decreased, 
counts averaging about 7000. 
Relative lymphocytosis and 
decrease of polynuclear neu- 
trophiles common. 
Eosinophiles notably de- 
creased; often absent. 
Myelocytes very rare. 

Basophiles not increased. 







I 5 













marked proliferation of myeloid tissue, and is characterized by 
a striking myelemia and generally by conspicuous enlargement 
of the spleen, with little or no involvement of the lymphatic glands. 
The lymphatic form, which may run either an acute or a chronic 
course, more commonly the latter, is a process associated with 
a proliferation of lymphoid tissue, and is characterized by a 
blood picture known as lymphemia, and in the great majority 
of cases by marked enlargement of the lymphatic glands, with 
moderate involvement of the spleen. But these two clinical pic- 
tures, in so far as they relate to the splenic and lymphatic hyper- 
trophy, are by no means constant, for, although cases of mye- 
logenous leukemia always have enlarged spleens, exceptionally 
they may also have decided lymphatic hyperplasia. Furthermore, 
cases of lymphatic leukemia are occasionally encountered in which 
there is a marked splenic tumor without demonstrable signs of 
lymphatic enlargement, as well as those in which the bone marrow 
alone is involved without implication of either the spleen or the 
lymphatic glands. To this group belong the acute leukemias 
with myelogenous lesions described by Miiller, 1 Walz, 2 Pappen- 
heim, 3 Michalis, 4 Dorothy Reed, 5 A. O. J. Kelly, 6 and others 
studies w T hich tend to upset. Ehrlich's theory that lymphatic leu- 
kemia is essentially a lesion of the lymph glands. Because of 
such atypical examples, the gross appearance of the spleen and 
lymphatics must be regarded as a sign of distinctly secondary 
importance to the blood picture, which is alone the tangible diag- 
nostic clue. 

Gerhardt, 7 Fleischer and Penzoldt, 8 and Wey 9 report cases of 
the apparent transition of myelogenous into lymphatic leukemia 
and into pernicious anemia. These so-called conversions of type 
may really represent merely the superposition of a leucocytosis 
sufficiently great temporarily to dominate the blood picture. If 
such were the case, it is obvious that with the disappearance of the 
leucocytosis the true character of the blood changes will become 
apparent in the first instance, a lymphocytosis, and in the second, 
a megaloblastic anemia. 

Of the two forms of the disease, the myelogenous is much the 
commoner. In the series of 42 cases of leukemia which the 

1 Centralbl. f. allg. Path. u. path. Anat.. 1894, vol. v, pp. 553 and 601. 

2 Ibid., 1901, vol. xii, p. 967. 

3 Virchow's Arch., 1899, vol. clvii, p. 19; ibid., 1900, vol. clix, p. 40; ibid., 
1901, vol. clxi, p. 424. 

4 Deutsch. med. Wochenschr., 1901, vol. xxvii, p. 651. 

5 Amer. Jour. Med. Sci., 1902, vol. cxxiv, p. 653. 

6 Trans. Assoc. Amer. Phys., 1903, vol. xviii, p. 481. 

7 Deutsch. Arch. f. klin. Med., 1880, vol. xxvi, p. 368. 

8 Ibid., 1896, vol. Ixvii, p. 300. 8 XV. Cong. f. inn. Med., 1897. 


writer has had the opportunity of studying, 29 were of the myelog- 
enous and 13 of the lymphatic form; while of Cabot's 66 
cases, 1 49 were myelogenous and 17 lymphatic a proportion 
of almost three of the former to one of the latter, for the combined 
series of 108 cases. 

In the present state of our knowledge it is 
PARASITOLOGY. not possible to regard leukemia as a disease of in- 
fectious origin, notwithstanding the suggestive- 
ness of the symptoms shown by many of those cases which run 
an acute course. Within the past few years several investigators, 
notably Delbert, 2 Kelsch and Vaillard, 3 Pallowski, 4 and Lowit 5 
have attempted to ascribe to various micro-organisms specific 
etiological relationship with the condition, but these attempts 
have thus far been unconvincing. Lowit's researches, however, 
are worthy of special attention, if for no other reason than for 
the elaborate and painstaking study which they represent. This 
author believes that two distinct forms of parasites may be dem- 
onstrated in leukemia : the Hcemamceba leukemia magna, thought 
to be the specific cause of the myelogenous form of the disease, 
and the Hcemamceba leukemia parva, which he claims is the 
definite infective principle of the lymphatic form. These so-called 
"specific bodies," which are found both in the blood of the 
peripheral vessels and in the hematopoietic organs,. have either 
a granular or an ameboid appearance, and bear a more or less 
close resemblance to the basophile granules of the leucocytes; 
navicular, segmenting, and vacuolated forms are also. said to 
occur. They are either attached to or lie within the leucocytes, 
especially the small lymphocytes, and more rarely the other varie- 
ties of normal leucocytes and the myelocytes; in an occasional 
instance they are said to be found lying free in the plasma. Al- 
though it is claimed that a leucocytic infection has been produced 
in animals by the injection of blood presumably containing these 
micro-organisms, all attempts to cultivate them on artificial media 
have proved futile. Lowit's amebae are demonstrated only in heat- 
fixed specimens, stained preferably with a steaming hot solution 
of Loffler's methylene-blue, after which they are washed, differ- 
entiated with a 0.3 per cent, solution of hydrochloric acid alcohol, 
again washed, and mounted. In specimens thus treated they stain 
metachromatically, and, if the acid alcohol differentiation has 
been properly effected, are the only elements except the basophile 

1 Loc. cit. t Bull, et me"m. Soc. de chir.. Paris, 1895, vol. xxi, p. 788. 

3 Annal. de 1'Institut Pasteur, 1890, vol. iv, p. 276. 

4 Deutsch. med. Wochenschr., 1892, vol. xviii, p. 641. 

s "Die Leukamie als Protozoeninfektion," Wiesbaden, 1900* 



leucocyte granules which retain the color of the dye. Turk, 1 who 
has followed out Lowit's technic precisely, in investigating this 
author's claims has come to the conclusion that these "specific 
bodies" are in no sense of parasitic nature, but merely artefacts 
resulting from the action of an aqueous solution of a basic dye 
upon the mast cell granules, which causes the partial solution of 
the latter elements and deforms them. Turk claims, furthermore, 
that these so-called amebae can be produced both in the normal 
blood of. man and in the blood of rabbits. 


The drop as it flows from the puncture is, in 

APPEARANCE most instances, of a bright scarlet color, and 

OF THE often has a peculiar and misleading appearance 

FRESH BLOOD, of density. After brief exposure to the air, 

it may become resolved into a serous, scarlet 

fluid, in which are suspended many minute, whitish, fat-like 

masses; the former appears to consist of serum and erythro- 

cytes, and the latter of adhering masses of leucocytes. It was 

probably this striking appearance of the leukemic blood drop 

that led Hughes Bennett erroneously to describe the condition 

as a "suppuration of the blood" before he proposed the more 

suitable term leucocythemia. In some cases the drop is simply 

much darker than normal, but it is difficult to believe that it ever 

resembles the chocolate-brown shade mentioned by some authors 

as occurring in this disease. The blood usually flows very freely 

from the wound, often by fine jets and spurts, especially if slight 

pressure is applied above the site of the puncture. 

Microscopically, the field is found to contain an enormous 
number of leucocytes, the proportion of these cells to the ery- 
throcytes being, by actual count, as great as i to 8 or 6, or even 
greater. Many different varieties of leucocytes may be distin- 
guished in the fresh specimen, the most striking being the large, 
mononuclear, finely granular cells of round or ovoid shape. These 
are the myelocytes which are present in enormous numbers in 
this form of leukemia, of which they form a characteristic blood 
picture. "Fractured" leucocytes, usually eosinophiles, with a 
cloud of escaped granules free in the plasma in the neighborhood 
of the disrupted cell body, may also be observed in variable num- 
bers, although such cells are more numerous in the dried, stained 
film. (See Figs. 54 and 55.) 

The erythrocytes vary in number and in appearance according 

1 XVIII. Cong. f. inn. Med., Wiesbaden, 1900. 



( Triacid Stain.) 

1. Small Lymphocyte. 

2. Large Lymphocyte. 

Contrast this cell with the myelocytes, 10, n, and 12, noting the presence of neutro- 
phile granules in the latter, and their absence in the lymphocyte. The size and 
nuclear characteristics of all these cells are practically the same. 

3. 4. Pol.ynuclear Neutrophiles. 

5. Eosinophile. 

In this "dwarf" eosinophile, ruptured during the preparation of the specimen, the 
granules are peculiarly arranged about the nucleus; no signs of protoplasm are dis- 

6. Eosinophilic Myelocyte. 

Note the irregularity with which the granules are stained. 

7. 8, 9, 10, ii, 12, 13, 14, 15. Myelocytes. (Neulrophilic.) 

These cells vary greatly in size (compare 8 with 9), hut they all have similar distinc- 
tive characteristics a large opalescent nucleus containing a scanty chroinatin net- 
work embedded in a cell body crowded with delicate neutrophile granules, precisely 
like those found in the polynuclear neutrophiles, 3 and 4. The nucleus of 7 is dis- 
tinctly indented and somewhat denser than that of the other myelocytes. This cell 
probably represents a developmental phase of the myelocyte just short of its, transi- 
tion into a typical polynuclear neutrophile. 
16. Normoblast. 

The erythrocytes (stained orange) show many evidences of deformity, an occasional 
megalocyte, many microcytes, and a few poikilocytes being present. Polychromato- 
philia is absent. 

(E. F. FABBR, fee.) 


to the severity of the coexisting anemia; in some instances large 
numbers of poikilocytes, megalocytes, and microcytes, with 
marked pallor of the corpuscles, may be seen, while in others 
the changes affecting the erythrocytes appear to be but trifling. 
In fresh films which have been allowed to dry for some time 
Charcot-Leyden crystals may sometimes be detected. They ap- 
pear as colorless, refractive crystals, shaped like octahedra, having 
long, pointed, sharp angles, and occurring either singly or in 
twos or threes, superimposed at right angles or as collections 
of radiating, crystalline masses. These crystals are not observed 
in the freshly drawn blood, being demonstrable only in films 
which have stood exposed to the air for at least twenty-four hours, 
and only occasionally even under this circumstance. 

On account of the presence in the blood of such large numbers 
of leucocytes a very small drop should be used for making the 
cover-glass spreads for staining, since it is advisable to avoid 
overcrowding the field with these cells. No difficulty will be ex- 
perienced in obtaining thin, evenly distributed spreads if this pre- 
caution is observed, especially if the cover-glasses are slightly 
warmed just before they are used. 

The coagulation of the blood and the for- 
COAGULATION. mation of the fibrin network must be regarded as 
variable. In some cases, especially Jhose with 
great loss of hemoglobin and erythrocytes, both processes are de- 
layed and imperfect, as evidenced by the formation of the "rasp- 
berry-jelly" clots referred to by the German writers. But in 
other cases the coagulation time is unaltered, and the fibrin net- 
work is perfectly normal. 

The alkalinity of the blood is usually de- 
ALKALINITY. creased, and, as in chlorosis, it increases after the * 
patient is given iron, in parallelism with the gain 
in hemoglobin and erythrocytes. In 3 cases studied by Burmin 1 
an average alkalinity equivalent to 146 mgm. was found, Landois' 
method being employed in the investigations. Taylor l found an 
average of 380 mgm. in three tested by the von Limbeck method. 

In cases with severe anemia the density of the 
SPECIFIC blood may fall as low as 1.035 or 1.040. Gra- 
GRAVITY. witz 1 has reported a case in which the figure was 
1.023. The fallacies in leukemia of Hammer- 
schlag's tables of specific gravities and their hemoglobin equiva- 
lents have already been pointed out. (See p. 133.) 

1 Loc. dt. 


Decided hemoglobin and erythrocyte loss is 
HEMOGLOBIN the invariable rule sooner or later during the 
AND course of the disease, the anemia usually being 

ERYTHROCYTES. well denned at the time the patient first comes 
under observation, and becoming acutely marked 
as the termination of the illness approaches. It is generally the 
case that the hemoglobin loss is disproportionately greater than 
the decrease in the erythrocytes, thus producing a moderately 
low color index, but in some cases just the opposite of this is ob- 
served. 1 In the writer's 29 cases, grouped in Table VIII, the 
color index averaged about 0.86. The hemoglobin percentage 
ranged between 24 and 70, averaging 48.6, and the number of 
erythrocytes was as low as 572,000 and as high as 4,200,000 per, the mean average being 2,814,000; the count of these 
cells was diminished to one-half of the normal standard, or below 
this figure, in n of the cases examined. An analysis of Cabot's 
series of 42 cases of myelogenous leukemia 2 shows these average 
findings: hemoglobin, 43 per cent.; erythrocytes, 3,123,000 per; and color index, about 0.68. 




From 60-70 7 From 4,000,000-5,000,000 i 

" 50-60 5 " 3,000,000-4,000,000 15 

" 40-50 5 2,000,0003,000,000 8 

" 30-40 10 " 1,000,000-2,000,000 4 

" 20-30 2 Below 1,000,000 i 

Average, 48.6 per cent. Average, 2,814,000 per 

Maximum 70.0 " Maximum, 4,200,000 " 

Minimum, 24.0 " " Minimum, 572,000 " " 

Fluctuations in the hemoglobin percentage and in the number of 
erythrocytes may or may not accompany variations in the leuco- 
cyte count. Sometimes, as the leucocytes rise, the erythrocytes 
fall, but again they remain practically stationary; or, the leuco- 
cytes may progressively fall to a comparatively moderate count, 
coincidentally with an apparent improvement in the patient's gen- 
eral condition, and yet the erythrocytes do not materially gain in 
numbers. Taylor 3 refers to two such instances which have come 

1 It is to be remembered that hemoglobin estimates in leukemia may be unreli- 
able (except when Dare's instrument is used), for correct readings are sometimes 
impossible, owing to the milkiness of the diluted blood from the presence of such 
immense numbers of leucocytes. About one-half of the hemoglobin figures in the 
accompanying table (Table VIII) were obtained by means of von Fleischl's hem- 
ometer, the remainder being based upon examinations with Oliver's and with 
Dare's instruments. 

2 Loc. cit. s Loc. cit. 



under his observation, in both of which the blood picture at cer- 
tain brief intervals resembled that of pernicious anemia, for under 
the influence of energetic arsenical treatment the leucocytes were 
reduced to normal, while the oligocythemia stubbornly persisted. 
In a case studied for a protracted period it is possible to distin- 
guish a general decrease in the erythrocyte count as the leuco- 
cytes increase, although the reverse may not be true. (See Chart 
II, p. 311.) 

Examination of the stained specimen shows the presence of 
nucleated erythrocytes in practically every case of myelogenous leu- 
kemia, these cells often being many times more numerous than 
in grave cases of pernicious anemia. Normoblasts prevail, 
always being more numerous than megaloblasts ; in some cases 
they are the only type of erythroblast to be observed; in others 
they are associated with a relatively moderate number of typical 
megaloblasts, or, more commonly, with large numbers of atypical 
forms, sharing the characteristics of the typical adult and em- 
bryonic nucleated erythrocytes. In the 9 cases of this variety of 
leukemia in which the writer has made differential erythrocyte 
counts the following estimates were obtained at the first exami- 
nations : 









8 37 6 




9 I 7 8 





























Average: 5,931 



Comparison of the above summary with the table giving the 
number and forms of erythroblasts in pernicious anemia (Table 
V, p. 283) illustrates two striking facts concerning these cells 
in myelogenous leukemia: the immense numbers in which they 
occur and the predominance of normoblasts over megaloblasts. 
Periods of temporary improvement in the patient's general health 
are often heralded by a notable increase in the normoblasts, 
but it is a noteworthy fact that during these remissions, while the 
leucocytes may fall decidedly, the normoblasts tend to persist 
in greater or less numbers. 

Examples of so-called nuclear extrusion, of multinucleation, 


of clover-leaf, dumb-bell, or other irregularly formed nuclei, and 
even, in rare instances, of karyokinesis are observed in many of the 
normoblasts. (See Fig. 49, p. 193.) Such alterations may be more 
conspicuous in highly developed cases of myelogenous leukemia 
than in any other disease of the blood. In an occasional normoblast 
the contracted, glistening, intensely basic nucleus is highly sugges- 
tive of pyknosis. No one who has done much blood work can 
fail to be struck with the obvious avidity displayed by the stroma 
of the erythroblasts for the acid fuchsin of the triple stain, a 
peculiarity which is exhibited in spite of good technic and the 
use of a reliable staining solution. 

Deformities affecting the size and the shape of the erythrocytes 
may be trivial or decided, depending upon the degree of hemo- 
globin and erythrocyte loss present. Polychromatophilia, alone 
or associated with basophilic stroma degeneration, is very com- 
mon in cases with great anemia, these changes affecting both 
the nucleated and the non-nucleated cells. 




Above 1,000,000 i 

From 500,0001,000,000 5 

400,000 500,000 4 

300,000- 400,000 4 

200,000- 300,000 6 

100,000 200,000 6 

50,000 100,000 2 

Below 50,000 i 

Average, 355,119 per 

Maximum, 1,046,000 " 

Minimum, 44,000 " " 

Striking increase in the number of leucocytes 
LEUCOCYTES, is found even during the early stages of the dis- 
ease. Counts in excess of 100,000 are generally 
expected, while in fully 70 per cent, of cases they exceed 300,000 
and in about 20 per cent., 500,000 or higher. In rare instances 
the number of leucocytes may be as high as 1,000,000 per 
One case of the writer's had 1,046,000 leucocytes. The leucocyte 
count may, very infrequently, equal or even exceed that of the 
erythrocytes, and it is easy to see that such a condition is possible, 
should the accompanying oligocythemia be intense. In a patient 
recently admitted to the Jefferson Hospital the leucocytes numbered 
650,000 and the erythrocytes 572,000. In the estimates given 
in Table IX, showing the number of leucocytes at the time the 
patient first applied for treatment, the average count was 355,119 
per, the highest being 1,046,000 and the lowest 44,000. 


Red, Hemoglobin. Black, Erythrocytes. Blue, Leucocytes. 


The number of leucocytes may fluctuate enormously at various 
times during the progress of the disease, a gain or a loss of some 
200,000 cells to the from week to week being a matter of 
common occurrence. Sometimes they are temporarily dimin- 
ished as the result of the administration of arsenic to the point 
of tolerance, or of the vigorous employment of other therapeu- 
tical measures; sometimes the decrease takes place independ- 
ently of the influence of remedial agencies, so far as can be deter- 
mined. When arsenic is withheld, the number of leucocytes 
promptly increases, and in spite of its use they ultimately increase 
as the disease runs its fatal course. The relations of the leuco- 
cytes' fluctuations to the number of erythrocytes have already 
been mentioned. Hayek 1 has drawn attention to the fact that in 
the leukemic individual the morning leucocyte count may be 
greater by more than 100,000 cells per than the estimate 
made during the afternoon, and vice versa. In a case in Pro- 
fessor Wilson's wards at the Jefferson Hospital the writer has 
been able to verify this statement, the counts being as follows: 
ii A.M., 144,000; 6 P.M., 256,000 a difference of 112,000 cells 
within a period of seven hours. The influences of digestion 
and other sources of error were, of course, excluded in making 
this observation. The occurrence of this enormous diurnal 
fluctuation emphasizes the importance of making the blood ex- 
amination of leukemic patients at precisely the same hour each 
day in cases studied for a long period. 



Small lymphocytes 4.1 12.0 . . 0.7 

Large lymphocytes 9.5 28.4 0.3 

Polynuclear neutrophiles 54.3 77.0 34.1 > 

Eosinophiles 5.4 28.7 1.2 

Myelocytes 20.6 44.0 7.0 

Mast cells 2 9.0 28.0 i.o 

The possibility of encountering a case of leukemia during a 
period of remission, when the typical blood changes are absent, 
must be borne in mind, for such instances are observed from time 
to time, although they are very rare. For example, McCrae 3 
reports a case, treated by arsenic, in which twice during a period 
of ten months the blood and general symptoms of the patient 
were typical of myelogenous leukemia, and twice were abso- 
lutely normal. When first examined, this patient's leucocytes 
numbered 584,000 per, three months later they had 

1 Wien. klin. Wochenschr., 1897, vol. x, p. 475. 

2 Estimates in 19 cases. 3 Brit. Med. Jour., 1900, vol. i, p. 760. 



fallen to 9250, two months after this they had risen to 178,000, 
and after a lapse of another five months they again fell to 5000. 
These fluctuations did not depend upon the influence of any in- 
tercurrent infection (see below), and the case is peculiar in that 
the leucocytes, as well as the erythrocytes, not only were normal 
in number, but also normal qualitatively, and in that the patient's 
splenic tumor entirely disappeared during the periods of remis- 
sion. A similar case (myelogenous) is reported by Simon and 
Campbell, 1 in which the leucocytes, after vigorous arsenical treat- 
ment for three weeks, fell from 350,000 to 4000, with disappearance 
of the myelemic blood picture and of the splenic tumor. Twelve 
months later the blood (save for the persistence of mast cells) and 
spleen were still normal. Other cases have been recorded showing 
brief periods of temporary decline to normal in the number of 
leucocytes, but with the persistence of myelocytes, or of the splenic 
enlargement, or of both. 

Here may be noted the astonishing results of Rontgen-ray 
therapy, alone and in conjunction with arsenic, in the treatment 
of myelogenous leukemia. Senn, 2 E. J. Brown, 3 C. H. Weber, 4 
Grosh and Stone, 5 Bryant and Crane, 6 and Aubertin and Beau- 
jard 7 have reported cases thus treated, in which the blood picture 
became normal and the splenic tumor disappeared. Time alone 
can determine whether such cures are symptomatic or radical. 

The following summary illustrates the good effects of arsenic, 
and, in the patient in question, the indifferent results of #-ray 
therapy in myelogenous leukemia, observed in the Jefferson Hos- 
pital : 


Hemoglobin , 60 per cent. 

Erythrocytes 3,520,000 

Leucocytes 341,000 

Small lymphocytes 3 per cent. 

Large lymphocytes i 

Polynuclear neutrophiles. .50 

Eosinophiles 3 

Myelocytes 15 

Mast cells 28 

The presence of myelocytes in large numbers is the hinge upon 
which the diagnosis of myelogenous leukemia must turn, for in 

1 Johns Hopkins Hosp. Bull., 1904, vol. xv, p. 181; also Med. News, 1904, vol. 
Ixxxiv, p. 431. 

2 Med. Rec., 1903, vol. Ixiv, p. 281. 

3 Jour. Amer. Med. Assoc., 1904, vol. xlii, p. 827. 

4 Amer. Med., 1904, vol. vii, p. 824. 

5 Jour. Amer. Med. Assoc., 1904, vol. xliii, p. 18. 

* Med. Rec., 1904, vol. Ixv, p. 574. Presse med., 1904, vol. i, p. 399. 



80 per cent. 
2.5 per cent. 

70 per cent. 
3.0 per cent. 









7- 1 ' 



no other disease are these cells so numerous or so constantly 
present. A high percentage of myelocytes, irrespective of the 
degree of increase in the total number of leucocytes of all forms, 
is as essential for the diagnosis of this variety of leukemia as is a 
predominance of megaloblasts for the recognition of pernicious 
anemia. In most cases they constitute at least 20 per cent, of 
the different forms of leucocytes, and occasionally as high as 50 
per cent, or more. In the 29 cases of the present series (Table 
X) the myelocytes, at the first counts, averaged 20.6 per cent., 
with 7 and 44 per cent, as the minimum and maximum esti- 
mates, respectively. It is the fact, not that myelocytes simply 
occur in this disease, but that they occur in such enormous 
numbers, that is of prime value in the diagnosis, since in no other 
condition in which this type of marrow leucocyte is found in the 
blood are they present in such striking abundance. For example, 

12 3 456 

i, 2, Dwarf forms, with relatively large and deeply stained nuclei situated in a relatively small 
amount of cell body containing neutrophile eranules. 3, "Fractured" myelocyte. 4, Extremely 
large form, with kidney-shaped nucleus. 5, Eosinophilic myelocyte with deeply constricted nucleus. 
6, Myelocyte with an hour-glass constriction of the nucleus. (Ehrlich's triacid stainj 

although myelocytes are very constant in pernicious anemia, they 
are only about one-twentieth as numerous in this disease, on the 
average, as they are in myelogenous leukemia. 

Many of the myelocytes are of very large size, some being 
quite 22 IJL in diameter and occasionally of somewhat larger di- 
mensions; others are dwarfed to no larger than the diameter of 
a small lymphocyte. The nuclei of these larger forms stain 
with relatively less intensity than those of the smaller. In- 
dentation, apparent division, and hour-glass constriction of the 
myelocytes' nuclei are also frequently noted. A very common 
form of this cell in myelogenous leukemia is characterized by 
its large size and its pale, kidney-shaped nucleus, the regularly 
convex border of which lies in intimate contact with fully one- 
half the periphery of the cell body. These and other atypical 
forms of myelocytes are shown in the accompanying illustra- 
tion (Fig. 53). The abnormalities affecting the granules of this 


type of cells, as well as certain degenerative changes, do not differ 
from those which are found in the polynuclear neutrophiles 
described below. 

The relative percentage of polynuclear neutrophiles is low, but 
not especially so, although, of course, the absolute number of 
this type of cells is greatly in excess of the normal standard, as 
may be demonstrated by taking into consideration the high total 
leucocyte count. For instance, in a case having 300,000 leuco- 
cytes per, with, say, 50 per cent, of them polynuclear 
neutrophiles, the actual number of the latter is 150,000 to the, or fifteen times the maximum normal number. The 
cases listed in Table X averaged 54.3 per cent, for this variety 
of leucocytes, with a range between 34.1 and 77.0 per cent, in the 
individual case, but other authors, with more extended series 
of cases as a basis for their statistics, give lower figures than these. 

A feature which at once attracts attention in the examination 
of the stained specimen is the deviation from the normal size of 
a large proportion of these neutrophilic cells. Dwarfed cells, 
often not more than 5 or 6 /i in diameter, and large forms, some 
of them measuring 15 fjt or even more in diameter, are common, 
the nuclei of the former usually staining much more sharply than 
those of the latter, which may exhibit a very feeble reaction to- 
ward the basic dye, and show a more diffuse and delicate chro- 
matin structure than is the rule in normal blood. The nuclei 
tend to exhibit extreme polymorphism and variations in their 
relative size to that of the cell body, and many of the cells are 
deformed in shape, being drawn out into various oblong and 
elliptical designs or into irregular elongated masses. " Frac- 
tured" cells, from which the granules have escaped, are very 
commonly seen. It seems reasonable to attribute the free neu- 
trophile granules sometimes seen in the blood in this disease to 
the rupture of a neutrophilic leucocyte, although the particular 
cell to which they belonged may be difficult to identify; the view 
expressed by some authors that such granules may preexist in 
the plasma is scarcely to be thought of seriously. All these 
deformities are doubtless the result of injuries to the cells in the 
preparation of the cover-glass spreads, and they suggest a lowered 
resistance on the part of the leucocytes. 

The number of granules in the polynuclear neutrophiles varies 
greatly in the individual cells: in some they are densely crowded 
throughout the protoplasm and overrun portions of the nucleus; 
in others they are confined to certain areas of the cell body, es- 
pecially in the neighborhood of the nucleus; while in still others 
they are distributed singly or in twos and threes through the 


protoplasm. Occasionally a cell wholly devoid of granules is 
observed, and, very rarely, one containing both neutrophile and 
a few isolated eosinophile or basophile granules. The neutro- 
phile granules themselves vary greatly in size, being in some 
cells so extremely delicate and fine that they can barely be dis- 
tinguished, while in others they almost equal the size of the 
smaller eosinophile granules. 

Fine and coarse vacuolation of the nucleus and protoplasm, a 
fissured and cracked appearance of the nuclear chromatin, and an 
apparent solution of the protoplasm, with freeing of the nu- 
cleus, are the most prominent degenerative changes affecting the 
polynuclear neutrophiles, as well as the other varieties of leuco- 
cytes, in this disease. 

The relative percentage of lymphocytes, small and large to- 
gether, is decidedly lower than normal, although their total num- 

I 2 3 


i, Cell containing both neutrophile and moderately coarse basophile granules. 2, Polynuclear 
cell with two ovoid nuclei and neutrophile granules, probably representing a later developmental 
stage than 6, Fig. 53. 3, "Fractured" polynuclear neutrophile. (No. i stained with Jenner's 
eosinate of methylene-blue, 2 and 3 with Ehrlich's triacid stain.) * 

ber to the of blood is greatly increased. As shown by 
the cases in Table X, these cells average approximately 14 per 
cent, of all varieties of leucocytes, which represents a diminution 
to about one-half of the proportion found in normal blood. It 
is the small lymphocytes which suffer the greater loss, for their 
proportion in the differential count is sometimes not more than 
a fraction of one per cent., and always greatly below normal; 
the large lymphocytes and "transitional" forms, on the contrary, 
average about normal, and, indeed, may be greatly increased 
in the individual case. Turk's stimulation forms are also met 
with, but these cells, as a rule, are not numerous. 

Atypical forms of lymphocytes are not so common in this form 
of leukemia as they are in the lymphatic variety. Such cells are 
described under the latter disease. (See p. 319.) 

Eosinophilia, as indicated by an increase in the total number of 
eosinophiles, is almost invariably found, and an increase above 


normal in the relative percentage of these cells sometimes, but not 
always, exists. 1 Thus, in the above-mentioned series the eosino- 
philes, which normally do not exceed 500 per, ranged 
from 780 to 129,150 and averaged 14,204 per, these figures 
corresponding to percentages of 1.2, 28.7, and 5.4, respectively. 

Ehrlich's original statement regarding an increase of the eosino- 
philes in this form of leukemia has been contradicted by several 
writers, notably by von Limbeck 2 and by Mtiller and Rieder 3 ; 
but these contradictions are based upon a misconception of Ehr- 
lich's remarks, for he never claimed that an abnormally high 
percentage of eosinophiles was associated with this disease, but 
said simply that their absolute number was increased. 

Marked variation in the size of many of the eosinophiles is com- 
monly observed, dwarf forms, 5 or 6 p in diameter, with densely 
crowded and deeply stained granules, being especially striking 
and apparently more numerous than the larger forms. Eosino- 
philic myelocytes, differing from ordinary 'neutrophilic myelocytes 
only in that they are studded with eosinophile granules, are very 
numerous, and are among the largest forms of the myelocyte 
found in this disease. "Fractured" eosinophiles are common, 
being usually more abundant than neutrophilic cells which have 
thus traumatically suffered. In some of the eosinophiles the 
granules are scanty, and in many their size varies greatly. Un- 
usually large-sized granules are often found, especially in the 
dwarf cells and in the extremely large forms. 

Mast cells with coarse, metachromatic granules are found 
with great constancy, being absent in but a small proportion of 
cases. This cell is especially suggestive of leukemia of this 
variety, since in no other disease does it occur in such large num- 
bers. In 19 cases of the above series the mast cells averaged 
9 per cent, in the differential leucocyte counts. In some 
leukemic bloods the mast cells attain an enormous size, being 
quite the largest cellular elements found in the specimen. They 
may be easily identified by their characteristic reaction toward 
the basic dyes, described in a previous section. (See p. 221.) 

From the above remarks it may be concluded that myelocytes 
and mast cells are present in the circulating blood at 'the expense 
of all the normal varieties of leucocytes except the eosinophiles, 
and that the brunt of this decrease is sustained by the mono- 
nuclear, non-granular forms, chiefly by the small lymphocytes. 

1 Simon's case, in which 13 differential counts were made, consistently showed 
an absence of eosinophiles. (Amer. Jour. Med. Sci., 1903, vol. cxxv, p. 984.) 

J Loc. cit. , f Deutsch. Arch. f. klin. Med., 1891, vol. xlviii, p. 96. 


( Triacid Stain.) 

uuac uwffiji VIIAII LUC lai^ci , 

7, 8, 9, 10, ii. Large Lymphocytes. 

Except in 10, which shows a delicate rim of fuchsin-stained protoplasm, these lympho- 
cytes appear simply as pale chromatin-deficient nuclear structures, lacking cell bodies. 
Compare these cells with the myelocytes, Plate IV. 

12. Transitional Form. 

The upper edge of the nucleus is somewhat indented and the protoplasm is distin- 
guishable; otherwise this cell resembles a large lymphocyte. 

(E. F. FABKR,/<C.) 


These bodies are greatly increased in number 

BLOOD in most cases of this form of leukemia, and may 

PLAQUES. frequently be recognized in the fresh specimen 

and in the diluted blood in the counting chamber 

of the hemocytometer. They are very constantly observed in the 

stained film prepared by the Romano wsky method. 


In most cases the blood drop is watery-looking, 

APPEARANCE pale, and thin, for in this variety of leukemia the 

OF THE anemia is usually very marked. The milky ap- 

FRESH BLOOD, pearance of the drop, frequently observed in the 

myelogenous form of the disease, is not often 

noticed in the lymphatic variety. 

The alterations in the coagulability, alkalinity, and specific 
gravity of the whole blood are similar to those met with in mye- 
logenous leukemia. 

Microscopically, the field is crowded with large numbers of leu- 
cocytes, the vast majority of which are mononuclear cells encir- 
cled by a perfectly hyaline, non-granular protoplasm. They may 
be quite uniformly either of small or of large size, or so many inter- 
mediate sizes may be present that it is impossible to distinguish 
any single predominating type. It is apparent that the leuco- 
cytes do not seem so numerous as in myelogenous leukemia, 
nor are their characteristics so striking, at first glance, because of 
the lack of granulations in their protoplasm. The difference 
between these hyaline cells and the granular leucocytes of the 
last-named disease, even although they may not happen to differ 
greatly in size and shape, is at once patent to the practised eye. 
The changes affecting the size, color, and shape of the eryth- 
rocytes vary with the degree of oligochromemia and oligocy- 
themia present; they are generally quite decided, as in any high- 
grade anemia. 

Marked anemia, characterized by a dispropor- 

HEMOGLOBIN tionate diminution in hemoglobin, is the general 

AND rule in this variety of leukemia, the decrease in 

ERYTHROCYTES. both hemoglobin and erythrocytes, especially in 

the former, being frequently greater than in the 

myelogenous form. In a -series of 13 cases the following figures, 

referring to the initial examinations, were obtained : 





From 40-50 i From 3,000,000-4,000,000 3 

" 30-40 4 2,000,000-3,000,000 4 

" 20-30 4 " 1,000,000-2,000,000 6 

" 10 20 4 

.Average, 38.1 per cent. Average, 3,032,211 per 

Maximum, 47.0 " Maximum, 3,590,000 " 

Minimum, 14.0 " " Minimum, 1,152,000 " " 

In rare instances the number of erythrocytes falls below 
1,000,000, and the hemoglobin so low that it is impossible to 
estimate the percentage at all accurately. Rapidly developing 
and extremely pronounced anemia is generally observed in cases 
which pursue an acute course. 

Nucleated erythrocytes, chiefly of the normoblastic type, are 
commonly found in moderate numbers, but never, except in rare 
instances, usually occurring in children, are they as numerous 
as in the myelogenous form of the disease. As a rule, when 
both normoblasts and mcgaloblasts are present, the former vastly 
outnumber the latter, although occasionally one meets with a 
case in which this predominance of adult-type erythroblasts is 
less pronounced. Thus, in one of the writer's cases the total 
number of erythroblasts was calculated at 10,678 per, of 
which 8512 were normoblasts and 2166 megaloblasts ; such a 
blood picture as this, however, is but seldom found. In general 
terms it may be said that the more acute the form of the disease, 
the more decided the oligochromemia and oligocythemia, and the 
more abundant the erythroblasts, the number and character of 
which appear to depend upon the grade of the anemia present. 
It should not be forgotten that in some cases of typical lymphatic 
leukemia nucleated erythrocytes are so scanty that they are detected 
only after repeated examinations. 

Deformities of size and shape and atypical staining of the eryth- 
rocytes are marked in direct relation to the severity of the anemia. 
The number of leucocytes is largely increased, 
LEUCOCYTES, but usually much less strikingly so than in 
myelogenous leukemia. Counts of 500,000 or 
even of 1,000,000 cells have, it is true, -been reported by a few 
observers, but only as rare examples of the extreme increase which 
it is possible for the leucocytes to attain in this condition. The 
cases in Table XII illustrate the range of the leucocytes in this 
variety of leukemia. 




Above 300,000 2 

From 200,000-300,000 i 

" 150,000-200,000 o 

" 100,000150,000 4 

" 50,000-100,000 3 

Below 50,000 3 

Average, 270,822 per 

Maximum, 958,000 " " 

Minimum, 38,000 " " 

By examination of the stained film the identity of the leuco- 
cytes responsible for the high count is more clearly distinguish- 
able, and it is found that the increase is due to a large absolute 
gain in the lymphocytes, the relative percentage of these cells to 
the other varieties of leucocytes generally being 90, 95, or even 
higher. In the series below summarized (Table XIII) these 
non-granular cells averaged 89.8 per cent, of the leucocytes, and 
equaled or exceeded 90 per cent, in three-fourths of the cases 
examined. In some instances the small lymphocytes are found 
to be in excess, and the field is dotted with small, deeply stained 
cells ranging from about 5 to -10 /* in diameter; in other instances 
the larger forms prevail, so that large, feebly stained cells, from 
about 10 to 15 /* or even larger, are in excess; while in still other 
cases the sizes and staining properties of the cells are so variable 
and atypical that it is impracticable to class them in two definite 
groups, large and small. It is generally- believed that small 
lymphocytes are associated with the more chronic forms of the dis- 
ease, and that the larger varieties are found in excess in the acute 
cases. Many of the larger forms, which possess a relatively large 
nucleus deficient in chromatin and a faintly basic non-granular 
protoplasm, are regarded as the mother-cells of the typical small 
lymphocytes. They are identical with the " lymphogonien " of 
Benda and the " leukoblasts " of Lb'wit, cells resident in the ger- 
minal nests of the lymphatic tissues. 



Total lymphocytes ...89.8 97.7 53.0 

Small lymphocytes 5 1 .4 94.0 21.1 

Large lymphocytes 38.4 76.2 i .o 

Polynuclear neutrophiles 7.6 45.0 1.6 

Eosinophiles 0.6 5.0 o.i 

Myelocytes i .4 4.9 0.3 

Mast cells ' o. i i .o o.o 

1 Estimates in 6 cases. 


Various atypical forms of lymphocytes, the commonest of 
which are pictured below (Fig. 55), are often numerous. With 
basic stains, such as methylene-blue, a ragged, torn condition 
of the basic seam of protoplasm, with so-called "budding," is 
frequently demonstrable, as well as nucleolation of some cells, 
especially of those of large size. Nuclear indentation and division 
and forms characterized by a small lymphocyte's nucleus within 
a large lymphocyte's cell body are also common. 

The relative proportion of polynuclear neutrophiles is markedly 
diminished, commonly to from about 5 or 10 per cent, of the 
total number of leucocytes, and sometimes even to below one per 
cent. These cells do not usually display the abnormal staining, 
the nuclear peculiarities, and the irregularities in size and shape 
that are so often seen in myelogenous leukemia. 

Myelocytes are present in the great majority of cases, but al- 
ways in trifling numbers, as in pernicious anemia; their propor- 
tion rarely exceeds one or two per cent, of all forms of leucocytes. 

I 2 3456 

i, Large lymphocyte with ragged protoplasm. Two small bits of protoplasm, the product of 
"budding," lie free in the plasma beside the cell, a, Large lymphocyte showing a nucleolus. 3, 
Large lymphocyte containing two nuclei. 4, Small lymphocyte containing an indented nucleus. 5, 
Small lymphocyte containing two nuclei. 6, Cell the size of a large lymphocyte with the nucleus of 
a small lymphocyte, (i and 2 are stained with eosin and methylene-blue; 3, 4, 5, and 6 with 
Ehrlich's triacid stain.) 

The percentage of eosinophiles is diminished, usually to a frac- 
tion of one per cent., and in a certain proportion of cases these cells 
are absent from the peripheral blood. It must be remembered, 
however, that even with a low relative percentage figure for the 
eosinophiles true eosinophilia may exist, although usually not to 
so marked a degree as in the myelogenous form of the disease. 
One per cent, of eosinophiles in a leucocyte count of 100,000 means 
looo eosinophiles per of blood, or twice the maximum num- 
ber found in the normal individual. Eosinophilic myelocytes are 
rare, but they occur in small numbers iri an occasional case. 

Increase in the number of basophiles does not occur with great 
frequency, and both the finely granular basophilic leucocytes and 
the typical mast cells are generally conspicuous by their absence, 
in contrast to their abundance in the myelogenous variety of this 


From the above it is evident that in lymphatic leukemia the 
increase in the total number of leucocytes is dependent upon a 
marked absolute gain in the lymphocytes, and that in conse- 
quence of this enormous influx of mononuclear hyaline forms, 
the relative percentages of the other leucocytes, especially of the 
polynuclear neutrophiles, are correspondingly diminished. 

As in the myelogenous form, the number of blood plaques is 
usually much increased. 

This term has been applied to a form of leu- 
ACUTE kemia which pursues a rapid course suggestive 
LEUKEMIA, of an acute infectious process, and ends fatally 
within a few weeks after the onset of the acute 
symptoms. Rapid, progressive enlargement of the lymphatic 
glands and a relatively small splenic tumor, associated with such 
clinical features as rigors and irregular pyrexia, bone pains, ulcer- 
ative stomatitis, and a decided tendency to purpura and to hemor- 
rhages from the mucous membranes, serve to identify most cases 
of this rapidly fatal disease. But, as already stated, there may 
be no evidences of lymphatic or splenic involvement, the lesions 
in such cases being confined to the bone marrow or to the deep 
lymph nodes. Some authors limit the duration of acute leukemia 
to six weeks, but the time limit proposed by Fraenkel, 1 four months, 
is generally accepted as being more appropriate. 

The disease is a rare one, for probably less than 100 authentic 
cases have been recorded up to the present time, although many 
more reputed instances have been published. Ebstein \ collected 
17 cases in 1889; Fraenkel 3 published the statistics of 10 in 1895; 
Bradford and Shaw 4 described, in 1898, 5 cases coming under their 
observation; and Fussell, Jopson and Taylor, 5 in the same year, 
published a collective report, embracing the statistics of- 57 cases 
selected as representing all the true examples of acute leukemia 
reported during the past twenty-one years. Since this report 
about 30 additional cases have been described by various observers. 6 

Beyond stating that lymphemia is the type of blood character- 
istic of acute leukemia, no special description of the condition of 
the blood is necessary. In the majority of cases the leucocyte 
increase may be attributed to a marked gain in the large lympho- 

1 Deutsch. med. Wochenschr., 1895, vol. xxi, p. 639, et seq. 

2 Deutsch. Arch. f. klin. Med., 1889, vol. xliv, p. 343. s Loc. cit. 

4 Medico-Chirurg. Trans., London, 1898, vol. Ixxxf, p. 343. 

5 Trans. Assoc. Amer. Phys., Philadelphia, 1898, vol. xiii, p. 124. 

* For a review of the literature of acute leukemia see (i) Hamman, Amer. Med., 
1904, vol. vii, p. 138; (2) Kelly, Univ. of Penna. Bull., 1903, vol. xvi, p. 270; (3) 
Miller and Hess, Amer. Med., 1904, vol. vii, p. 389; (4) Mixa, Wien. klin. Rund- 
schau, 1901, vol. xv, pp. 655 and 671; (5) Billings and Capps, Amer. Jour. Med. 
Sci., 1903, vol. cxxvi, p. 375. 


cytes, which greatly predominate over the small forms, while the 
polynuclear neutrophiles, myelocytes, and eosinophiles are rela- 
tively few in number. As a rule, the more acute the case, the 
more decided the predominance of the large lymphocytes, which 
usually show well-marked evidences of nuclear and protoplasmic 
degenerative changes. The loss of hemoglobin and erythrocytes 
is generally more marked and the erythroblasts are more numer- 
ous, than in the commoner forms of chronic lymphatic leukemia. 
The blood may coagulate very imperfectly and the plaques are 
often greatly diminished in number. In a case reported by 
Bensaude, 1 total absence of clotting and of serum transudation 
was noted after the expiration of twenty-four hours. 

Acute leukemia may begin as such, or either the chronic lym- 
phatic or the myelogenous form may develop acute symptoms, 
with a coincident change in the condition of the blood, but this 
change in the myelogenous variety is extremely rare. Only nine 
cases of acute myelogenous leukemia are on record. 2 

The development of an acute infectious proc- 

INFLUENCE OF ess in a leukemic individual commonly provokes 

ACUTE INTER- striking changes in the behavior of the leucocytes, 

CURRENT IN- consisting in most instances in a decrease of their 

FECTIONS. total number to the of blood, associated 
sometimes with an increase in the polynuclear 
variety of cells and a relative diminution in the number of myel- 
ocytes. At other times there is practically no alteration in the 
relative proportions of the different forms as they existed in the 
leukemic blood prior to the onset of the complicating infection. 

Among the infectious conditions acting in this manner on the 
leucocytes are abscess, sepsis, pneumonia, influenza, tuberculosis, 
and erysipelas, but it seems that rheumatic fever has no such effect. 
Weil, 3 who has studied the effects of colon, pneumococcus, and 
streptococcus infections in both forms of leukemia, comes to the 
conclusion that the most powerful influence upon the blood pic- 
ture is exerted by streptococcus infections. Malignant disease 
is also capable of bringing about a leucocyte decrease characterized 
by a relative gain in polynuclear neutrophiles at the expense of 
the lymphocytes, but the loss does not appear to be so decided as 
that excited by a specific infectious process. 

In rare instances the occurrence of an, infectious disease fails to 
cause a decrease ift the leucocytes, and thus to destroy the leu- 
kemic picture, but, on the contrary, increases them, by superim- 

1 Sem. med., 1903, vol. xxiii, p. 57. 

J Billings and Capps, loc. cit. 

3 Gaz. hebdom. de med. et de chir., 1900, vol. v, p. 829. 


posing a typical polynuclear neutrophile leucocytosis, which re- 
mains during the existence of the complicating infection the 
conspicuous feature of the blood. The writer has observed a 
typical illustration of such a change in a case of myelogenous 
leukemia, in which, within ten days after the onset of a complicat- 
ing peritonitis, the leucocyte count rose from 245,000 to 400,000, 
and the proportion of polynuclear neutrophiles from 44.5 to 79 
per cent., while the percentage of myelocytes fell from 20.5 
to 8. 

Dock 1 has collected 50 cases of leukemia complicated by 
various intercurrent infections, which in 27 instances were tuber- 
culous and in 23 non-tuberculous. In leukemia plus tuberculosis 
the virulence of the complicating infection appears to govern the 
behavior of the leucocytes, which are not decidedly influenced 
in chronic forms of the disease, but which usually are greatly 
diminished in the acute miliary type. In the 23 cases of inter- 
current infections other than tuberculosis Dock's analysis shows 
that a marked leucocyte decrease occurred in n, a relatively 
slight decrease in 9, and either an increase or no change in the 
remaining 3. Of qualitative changes in these cases there was 
noted a general tendency of the leucocytic blood picture to dis- 
appear, with a decided increase, both absolutely and relatively, 
in the polynuclear neutrophiles. Dock's masterly monograph 
should be consulted for a complete account of this complicated 
topic, which does not lend itself to text-book discussion. 

Coincidentally with the improvement in the condition of the 
blood there is frequently a decrease in the size of the patient's en- 
larged spleen and lymphatics, the period during which the leukemic 
condition is thus bettered, and, so to speak, held in abeyance, 
corresponding to the duration of the complicating infection, for 
the blood gradually regains its leukemic type and the glandular 
and splenic tumors reappear as recovery from the intercurrent 
disease takes place. 

D NO The following blood picture is characteristic 

of the myelogenous variety of leukemia: 
Hemoglobin. Decided loss, averaging about 50 per cent. Color 

index subnormal, or high. 
Erythrocytes. Counts average about 3,000,000 per 

Erythroblasts very numerous, cells of the normo- 

blastic type predominating. 

Deformities of size and shape, polychromato- 

philia, and basophilic stroma degeneration marked 

in cases with severe anemia. 

1 Amer. Jour Med. Sci., 1904, vol. cxxvii, p. 563. 


Leucocytes. Increased to about 350,000 per 

Myelocytes constitute about 20 per cent, of all 


Relative percentage of polynuclear neutrophiles 


Relative percentage of lymphocytes very low. 

Eosinophiles .absolutely, sometimes relatively, in- 

Mast cells average about 10 per cent, of all forms. 

Atypical forms of neutrophiles numerous. 
Plaques. Increased. 

In lymphatic leukemia the blood changes may be briefly ex- 
pressed thus: 
Hemoglobin. Marked loss, averaging about 60 per cent. Color 

index low. 
Erythrocytes. Counts average about 3,000,000 per 

Erythroblasts usually scanty, cells of the normo- 

blastic type predominating. 

Deformities of size and shape and atypical stain- 
ing reaction marked in relation to the degree of 

anemia present. 
Leucocytes. Increased to about 250,000 per, counts 

above this figure being rare. 

Lymphocytes constitute about 90 per cent, of all 


Relative percentage of polynuclear neutrophiles 

strikingly low. 

Relative percentage of eosinophiles diminished; 

rarely, an absolute increase. 

Small numbers of myelocytes frequent. 

Basophiles usually not increased. 

Atypical forms of lymphocytes numerous. 
Plaques. Increased. 

In dealing with the differential diagnosis of leukemia it is nec- 
essary to distinguish the myelogenous from the lymphatic form, 
and also to differentiate both forms of the disease from a number 
of other conditions w r hich may present either somewhat similar 
blood findings or which, apart from the condition of the blood, 
may have closely similar clinical manifestations. Thus, on the 
one hand, leucocytosis and lymphocytosis require differentiation 
because they produce changes in the blood which may be con- 
fused with leukemia; while, on the other hand, one must dis- 
tinguish between leukemia and Hodgkin's disease, splenic anemia, 
and a number of conditions causing enlargement of the spleen, 


neighboring organs, and lymphatic glands, because of the resem- 
blance, even the identity in some instances, of the other clinical 

Myelogenous and lymphatic leukemia can be distinguished 
only by examination of the blood, for the distinction between 
these two forms of the disease cannot be based with any degree 
of certainty upon the gross clinical appearance of the spleen and 
lymphatics. Nothing can be more marked than the contrast be- 
tween the two blood pictures. In the myelogenous form the 
leucocyte count is usually much higher, and is associated with the 
presence of immense numbers of myelocytes, and with an increase 
in the eosinophiles and mast cells; the oligocythemia is not so 
marked, but erythroblasts are exceedingly numerous, and, strangely, 
tend to persist independently of any increase in the erythrocytes 
which may occur from time to time. In the lymphatic form 
the relatively moderate leucocyte increase depends upon an 
excessive gain in the ungranulated cells, or lymphocytes, myelo- 
cytes being either absent or present in trifling numbers, and 
decided increase in the eosinophiles and mast cells being most 
unusual; the oligocythemia is usually decided, but erythroblasts 
are scanty, and stand in relationship to the degree of anemia 
existing. The important points of difference are, therefore, the 
presence of a myelocytic blood in the myelogenous form, .and of 
a lymphocytic blood in the lymphatic variety. 

Pathological leucocytosis may occasionally involve an increase 
in the total number of leucocytes equal to that found in either 
form of leukemia, especially in those cases in whicti a period 
of temporary improvement with a fall in the leucocyte count 
exists. But, aside from the more or less temporary character of 
the increase in leucocytes, the differential count at once shows 
that, unlike leukemia, the gain depends upon a large absolute and 
relative increase in the polynuclear neutrophiles, which constitute 
ordinarily 85 per cent, or more of the several forms of leucocytes. 

Lymphocytosis, which is usually a relative condition, may in 
rare instances become absolute, so that, in addition to the increase in 
the relative percentage of lymphocytes, the total number of leu- 
cocytes in the blood is also decidedly increased. In marked in- 
stances of this sort it is obviously impossible to distinguish the 
blood change from that of lymphatic leukemia, and the aid of 
other clinical symptoms must be invoked to make the diagnosis 
clear. Thus, both an absolute and a relative lymphocytosis, 
closely simulating the lymphatic form of leukemia, have been 
observed in severe cases of chlorosis, in pertussis, in sarcoma of the 
lymphatic structures, and in acute inflammatory processes oc- 


curring in young children. The author recalls an instance of 
marked absolute lymphocytosis in a case of pernicious anemia 
which seemed to justify the tentative diagnosis of lymphatic leu- 
kemia, an error which was later corrected, when the megaloblastic 
blood picture became apparent. In such instances, which are, 
fortunately, of very rare occurrence, it is true that neither the perr 
centage of lymphocytes nor" the count of leucocytes is likely to 
average so high as in lymphatic leukemia, but still the blood 
changes are sometimes very misleading, and should not be relied 
upon to the exclusion of other equally important symptoms. 

The glandular and splenic enlargements of Hodgkin's disease 
form a clinical picture identical with either the myelogenous or 
the lymphatic variety of leukemia, so that these conditions are 
distinguishable only by the result of the blood examination. But 
by this means the diagnosis is made extremely simple, by finding 
in Hodgkin's disease either entirely normal blood or a variable 
degree of anemia. The number of leucocytes is usually nor- 
mal, except in cases in which some complicating inflammatory 
or infectious process causes a moderate increase, typical of a 
polynuclear neutrophile leucocytosis. 

In chloroma the blood changes may be practically those of 
an acute lymphatic leukemia progressive anemia with absolute 
lymphocytosis. In chloroma, however, the clinical picture is 
made up of exophthalmos, deafness, orbital pain, elastic swellings 
of the orbital and temporal regions, and a tendency toward met- 
astases of the "green tumors" in periosteal structures. 

Von Jaksch's multiple periostitis is a symptom-complex 
resembling myelogenous leukemia, in that a myelocytic anemia 
with splenic enlargement and a tendency toward hemorrhage are 
symptoms common to both conditions. In the early stages of the 
disease described by von Jaksch, fever, drenching sweats, painful 
and swollen joints, and thickening of the distal extremities of the 
radius and ulna are observed, these distinctive symptoms being 
succeeded by a preagonal stage marked by a cessation of the bone 
pains and sweating and by an aggravation of the" already marked 
anemia, splenomegaly, and hemorrhagic tendency. 

Still's disease, because of the splenic and lymph-node enlarge- 
ments, may superficially resemble leukemia, but in this form of 
infantile arthritis, besides an aleukemic blood picture, one finds 
multiple arthritis,- especially of the periarticular structures, and 
usually a clear history of rickets. 

The rather close resemblance which certain cases of splenic 
anemia bear to leukemia, together with the points of difference 
between them, have already been described. (See p. 295.) 


Enlargements of the spleen, left kidney, and pancreas may lead 
to the belief that leukemia exists. Thus, splenic tumors due 
to chronic malarial infection, to amyloid disease, to cysts, and to 
malignant neoplasms; enlargements of the left kidney, such as 
can be caused by hydronephrosis, by cysts, and by malignant 
disease; as well as cystic tumors of the pancreas and malignant 
disease of the retroperitoneal glands all may, on physical exami- 
nation, simulate more or less faithfully the leukemic spleen. The 
negative character of the blood findings will at once exclude leu- 
kemia, should one of the above-named conditions be the cause 
of the physical signs suggesting this disease. 

Lymphatic hyperplasia, due to tuberculosis, to syphilis, and to 
malignant disease, may also be mistaken for the glandular in- 
volvement of leukemia, for such enlargements sometimes show 
nothing distinctive. In tuberculous adenitis the blood is either 
normal or anemic, if the cachectic state of the patient is marked; 
or, should there happen to be a secondary infection of the glands 
plus the tuberculous lesions, a simple polynuclear leucocytosis 
is found. In syphilitic adenitis there is often anemia with a mod- 
erate polynuclear leucocytosis, and sometimes with a relative 
lymphocytosis, especially in children. In malignant disease of 
the lymphatics increase in the number of leucocytes may also be 
noted in association with a high-grade anemia; in carcinoma the 
increase involves chiefly the polynuclear neutrophiles, but in sar- 
coma the lymphocytes may be unduly increased, though not to 
the extent found in lymphatic leukemia. 


Nothing characteristic is observed either., in 

APPEARANCE the gross appearance of the fresh blood drop or in 

OF THE the unstained film, microscopically. The blood 

FRESH BLOOD, may appear normal, or it may show changes 

common to any secondary anemia. 

The alkalinity and specific gravity of the whole blood are di- 
minished in relation to the degree of anemia which exists. Coag- 
ulation may take place slowly, and even be as incomplete as it 
is in some cases of leukemia; or it may occur within the normal 
time limit. 

Both the hemoglobin percentage and the 

HEMOGLOBIN number of erythrocytes are normal in the early 

AND stages of the disease, and in slowly progressive 

ERYTHROCYTES. cases the blood may remain unaffected for a long 

period. But sooner or later, as the disease pro- 


gresses and a cachectic condition of the patient develops, anemia 
appears, gradually where the course of the disorder is slow, and 
rapidly in the more acute forms. Counts made when the patient 
first comes under observation usually average 4,000,000 or 5,000,- 
ooo cells per, but in the later stages the number frequently 
falls to one-half this figure pr even less. The loss of hemoglobin 
begins earlier, and in most instances is proportionately somewhat 
greater, than the erythrocyte decrease, so that subnormal color 
indices rule not decidedly low, but yet twenty points or so below 
the normal standard. In cases which develop excessive oligocy- 
themia the index figures may be quite as high as in pernicious 

In the series of 21 cases summarized in Table XIV, the hemo- 
globin averaged about 55 per cent., ranging between 30 and 81 
per cent.; the erythrocyte count ' averaged 3,591,423 per, 
the minimum being 1,300,000 and the maximum 5,225,000, with 
more than one-third of the cases having 4,000,000 cells or more. 




From 80-90 3 Above 5,000,000 2 

70-80 3 From 4,000,000-5,000,000 6 

6070 i 3,000,000-4,000,000 10 

5060 6 2,000,000-3,000,000 i 

40-50 4 " 1,000,0002,000,000 2 

30-40 4 

Average, 55.3 per cent. Average, 3,591,423 per 

Maximum, 81.0 " " Maximum, 5,225,000 " " 

Minimum, 30.0 " " Minimum, 1,300,000 " " 

Qualitative changes affecting the corpuscles occur in relation 
to the intensity of the anemic process, deformities of shape and 
size and atypical staining reaction of the cells being associated 
with cases in which notable hemoglobin and erythrocyte losses 
exist, and being absent when the anemia is moderate. Nucle- 
ated erythrocytes are not common, nor are they numerous when 
present. Usually they are wanting, except in connection with a 
high-grade anemia, under which circumstance a few normoblasts 
may be detected, and in rare instances an occasional megaloblast. 
In the average case the leucocytes are normal, 
LEUCOCYTES, both in number and in the relative percentage of 
different varieties. More rarely, relative lym- 
phocytosis occurs, involving a decrease in the percentage of poly- 
nuclear neutrophiles, but not increasing the total number of 
leucocytes. An instance of this kind has been observed by the 


writer, in which the relative proportion of lymphocytes to other 
forms of leucocytes habitually remained for some months between 
70 and 85 per cent., this change affecting chiefly the large lym- 
phocytes, while the total leucocyte count never exceeded normal. 



Above 20,000 1 

From 15,00020,000 3 

" 10,00015,000 4 

" 5,00010,000 7 

Below 5,000 6 

Average, 8,819 P er 

Maximum, 21,000 " 

Minimum, 1,000 " " 

If secondary infection takes place, it soon becomes evident by 
an increase in the leucocytes to about 20,000 or more, principally 
involving the polynuclear neutrophile cells, at the expense of the 
lymphocytes a picture of typical leucocytosis. In some in- 
stances, however, without any apparent signs of a secondary in- 
fection or of a glandular inflammation, the total count may ex- 
ceed the normal standard by several thousand cells, and yet show 
a normal or even somewhat decreased proportion of neutrophiles. 
These changes in the blood picture must be distinguished from 
those indicating the conversion of Hodgkin's disease into true 
lymphatic leukemia. This transition, although exceedingly rare, 
probably sometimes occurs, as shown by Wende/ Posselt, 2 
Senator, 3 Hosier, 4 Fleischer and Penzoldt, 5 and others. Should 
the anemia be very marked, pronounced leucopenia is commonly 
associated with it. In the present series (Table XV) about one- 
fourth of the cases showed frank leucocytosis, the highest estimate 
being 21,000, the lowest 1000, and the -average 8819. 



Small lymphocytes 16.5 49.0 i.o 

Large lymphocytes 10.8 21.0 0.5 

Polynuclear neutrophiles ...69.6 88.0 46.2 

Eosinophiles 2.2 10.0 o.o 

Myelocytes ._ 0.5 3.0 o.o 

Mast cells * .' .' 0.9 i . r o.o 

1 Amer. Jour. Med. Sci., 1901, vol. cxxii, p. 836. 

2 Wien. klin. Wochenschr., 1895, vol. viii, p. 407. 

3 Berlin, klin. Wochenschr., 1882, vol. xix, p. 533. 

4 Virchow's Arch., 1888, vol. cxiv, p. 461. 

5 Deutsch. Arch. f. klin. Med., 1896, vol. Ixvii, p. 300. 
* Estimates in n cases. 



Small numbers of myelocytes are not uncommon in the ad- 
vanced anemia of Hodgkin's disease, but they are never more 
numerous than in any other condition accompanied by a similar 
deterioration of the blood. These cells were found in 8 of the 21 
cases under consideration. 

Small numbers and low percentages of eosinophiles are the 
rule, in cases both with and without leucocytosis ; it is most un- 
usual for these cells to attain the maximum normal figure, and 
they are sometimes wholly absent. 

Neither the finely granular basophiles nor the typical mast cells 
are increased in this disease. The author found the latter in but 
a single case of n examined. 

As in both forms of leukemia, the number of blood plaques in 
Hodgkin's disease is usually increased. , 

Although no characteristic blood changes 
DIAGNOSIS, occur in this condition, the alterations most com- 
monly observed may be briefly summed up as 
follows : 

Hemoglobin. Normal in the early stages of the disease; later, a 
moderate decrease, estimates averaging about 55 
per cent. Color index commonly subnormal, 
rarely high. 

Erythrocytes. Normal in the early stages ; later, a variable degree 
of oligocythemia, counts averaging about 3,500,- 
ooo per 

Erythroblasts uncommon and scanty when pres- 

Normoblasts prevail almost exclusively, megalo- 
blasts being very rare. 

Deformities of size and shape, polychromato- 
philia, and basic stroma degeneration only in cases 
with high-grade anemia. 
Leucocytes. Normal or moderately increased. 

Either polynuclear neutrophiles pr lymphocytes 
may be relatively increased, more commonly the 

Small numbers of myelocytes in very anemic 

Eosinophiles not increased. 
Basophiles not increased. 
Plaques. Usually increased. 

The absence of characteristic blood changes in Hodgkin's 
disease at once distinguishes the condition from its clinical coun- 
terfeit, leukemia, but, aside from this single disease, the blood ex- 


amination is valueless in the differentiation of other conditions 
having somewhat similar involvement of the glandular structures. 
These conditions, tuberculous and syphilitic adenitis, local lympho- 
matous tumors, and malignant neoplasms of the lymphatics, must 
therefore be distinguished from Hodgkin's disease by other clin- 
ical methods, for all may provoke identical blood changes. 

In reviewing the clinical history of a case of suspected Hodg- 
kin's disease the following symptoms should be given special 
consideration: the gradual onset of a widespread hyperplasia of 
the lymphatic structures, occurring most commonly in males 
under middle age; the progressive character and chronicity in 
most cases of the disorder; the tendency in some cases toward 
the occurrence of unexplained febrile periods, sometimes coincid- 
ing with a rapid and marked increase in the size of the affected 
glands which- disappears as the fever subsides; the cachexia, 
asthenia, and emaciation of the patient, frequently associated with 
gastro-intestinal and circulatory disturbances and with a tendency 
to hemorrhages, such as epistaxis and purpura; the presence of 
pressure symptoms, such as cough, dysphagia, dyspnea, edema, 
and pleural and peritoneal effusions; and the development of 
bronzing of the skin in an occasional case. 

In typical cases the glandular enlargement forms a series of 
distinct, painless, hard tumors, each freely separable from its 
neighbor, and rarely caseating or suppurating. Due weight 
should also be given to the fact that in the majority of cases the 
lesion originates in the superficial lymphatic glands of the cervical 
region, beginning either in the occipital or in the inferior carotid 
triangle. The spleen is moderately enlarged in the majority of 
cases, and in others the liver, kidneys, suprarenals, tonsils, thymus, 
thyroid, and sexual organs may be involved in the lymphoid 

As in leukemia, remarkable improvement in the blood and 
other clinical features of Hodgkin's disease is reported by Senn, 1 
by Steinwald, 2 by Childs, 3 and by Pusey 4 as the result of #-ray 

Tuberculous adenitis usually first involves a group of glands in 
the submaxillary triangle, and tends to produce inflammatory 
adhesions between the tissues and the glandular structure, with 
softening, fusing; caseation, and suppuration of the glands. It is 
of sluggish development, often occurs in the very young, and is 

1 N. Y. Med. Jour., 1903, vol. Ixxvii, p. 665. 

2 Medicine, 1904, vol. x, p. 438. 

S N. Y. Med. Jour., 1904, vol. Ixxx, p. 13. 

4 Jour. Amer. Med. Assoc., 1902, vol. xxxviii, p. 911. 


almost always confined to a single group of glands. Evidences 
of tuberculous lesions in the lungs or in other parts of the body, 
especially of dental caries, cutaneous lesions of the face, and ade- 
noid pharyngeal growths, and the discovery of tubercle bacilli in 
the glandular tissue are valuable evidences of the tuberculous 
nature of the disease, yet they do not positively exclude Hodg- 
kin's disease, since the coexistence of the two conditions in the 
same individual is possible beyond a doubt. Steinberg's belief 1 
that Hodgkin's disease is essentially tuberculosis is still supported 
by Musser 2 and Sailer, 3 although the recent brilliant work of 
Dorothy Reed, 4 of Longcope, 5 and of Simmons 6 is convincing proof 
that the disease is a distinct clinical entity, in no way related 
pathologically to the tubercle bacillus. 

In syphilitic adenitis of the neck the post-cervical groups are 
first affected, the glands being of cartilaginous hardness, painless, 
freely movable, and of small or moderate size. The glandu- 
lar enlargement is often more or less general, but the affected 
groups do not attain a large size. A history of an initial lesion 
in the vicinity of the primary glandular swellings, or of the ap- 
pearance of secondary symptoms, the disappearance of the glan- 
dular tumors after the administration of mercury, and the pres- 
ence of Justus' test will suffice to prove the specific character of 
the hyperplasia. 

A local lymphoma is limited strictly to a single group of glands, 
forming a painless, dense mass, free from inflammatory adhesions, 
caseation, and suppuration. It commonly involves the submaxil- 
lary glands, may attain a large size, and is unassociated with con- 
stitutional symptoms. Such a local lymphatic tumor cannot be 
distinguished from the early stage of Hodgkin's disease, for in 
some cases of the latter the general lymphoid hyperplasia is pre- 
ceded by a period during which the only sign of the condition is 
a localized enlargement of a single group of glands. If, accord- 
ing to Osier, 7 a local tumor of this kind persists^ for over a year 
or eighteen months without involving the glands of the opposite 
side or of the axilla, it is almost certainly a non-malignant 

Sarcoma of the lymphatic tissue forms an immovable tumor, 
early complicated by inflammatory processes which cause inter- 
glandular adhesions and adhesions between the glands and the 

1 Zeitschr. f. Heilk., 1898, vol. xix, p. 21. 2 Amer. Med., 1902, vol. iii, p. 13. 

3 Phila. Med. Jour., 1902, vol. x, pp. 615 and 669. 

4 Johns Hopkins Hosp. Rep., 1902, vol. x, p. 133. 

5 Bull. Ayer Clin. Lab. ? Penna. Hosp., 1903, vol. i, p. 4. 
8 Jour. Med. Research, 1903, vol. ix, p. 378. 

7 Cited by Bramwell, "Anemia," Philadelphia, 1899, p. 203. 


surrounding tissues. The swelling is often red and inflamed, pits 
upon pressure, and resembles an abscess, while the skin over the 
site of the lesion is frequently marked by a maze of tortuous, 
congested, cutaneous veins, and is prone to ulcerate. If nerves 
are entangled in the growth, the tumor is exquisitely painful. 
The adjacent tissues become densely infiltrated by the sarcomatous 
growth, and involvement of distant organs by metastasis is likely 
to occur. If such be the case, microscopical search for sarcoma 
cells in the fresh specimen may give a definite clue. (See " Sar- 
coma," Section VII.) Sarcoma of the lymphatic glands may 
occur at any period of life. 

Carcinoma of the lymphatic glands is secondary to an initial 
growth in some other part of the body, so that in the region of 
the neck search should be made for a primary cancerous lesion 
in the mouth and upper air-passages. The disease is most com- 
monly found during the decline of life. 

Finally, as Tyson so pertinently remarks, 1 it should not be for- 
gotten that all the conditions named as possible to be mistaken 
for Hodgkin's disease are limited to a single group of glands, 
while Hodgkin's disease always extends, and the fact of such 
limitation is of itself sufficient to exclude the disease. This pro- 
gressive involvement of the lymphatic glands, group after group, 
must, after all, be the mainstay in the diagnosis of doubtful- cases. 


Excision of the spleen in man is followed by a 
HEMOGLOBIN diminution in the hemoglobin and erythrocytes, 
AND the degree of which is generally believed to be 

ERYTHROCYTES. more, pronounced than can be accounted for by 
the simple factor of hemorrhage incident to the 
operation. Blood regeneration is slow, especially the restoration 
of the 'hemoglobin, which is prone to increase much less rapidly 
than is the rule in an ordinary secondary anemia. In uncompli- 
cated cases from one to three months' time usually elapses be- 
fore the normal percentages of hemoglobin and erythrocytes are 
attained; in unfavorable cases persistence of the anemia for a 
much longer period is to be observed. Splenectomies attended 
by great loss of blood may excite, in addition to an extreme cel- 
lular decrease, striking qualitative changes, and in such instances 
the blood picture is characterized by the presence of many normo- 

1 "Practice of Medicine," Philadelphia, 1898, p. 606. 



blasts, achromacytes, and corpuscles deformed in shape and size. 
Conspicuous post-operative anemia is especially common in pa- 
tients to whom saline intravenous injections have been admin- 






















i 1 








Before operation. 

Megaloblasts and nor- 

moblasts found. 








2 days after operation. 
















n ' " 

Myelocytes and mast 

cells found. No eryth- 



1 6 days after operation. 








21 " 

Myelocytes, 0.2 per cent. 








27 days after operation. 

No erythroblasts. 








37 days after operation. 








56 " " 

Myelocytes found; no 







99 days after operation. 

Erythrocytes normal. 

2 Z 








Before operation. 







7 days after operation. 








60 " " 








3 years " " 

3 s 








14 days " 





3 2 



27 ," " 








33 " " 








2 years and six months 

after operation. 

4 T 




Before operation. 




3 days after operation. 




6 " 




48 " " 




4 m'ths " 



1 1, 600 

5 years " 

1 Warren, Annals of Surgery, 1901, vol. xxxiii, p. 513. 

2 Hartmann and Vaquez, Compt. rend Soc. biol., Paris, 1897, vol. iv, p. 126. 

3 Ibid. 4 Czerny, cited by Vulpius, Beitrage z. klin. Chir., 1894, vol. xi, p. 633 




S 1 














J o 











13,000 13 















































































W O 













6 days after operation. 
19 "" 
26 " " 
Normoblasts found. 




4 days after operation. 

8 ;; ;: i 

18 " 


30 " " 

52 ' " " 



60. i 




7 days after operation. 

15 " " 
19 " " 

38 " " 
46 " " 
2 m'ths " 

5 " " 









Before operation. 
7 days after operation. 
30 " " 

45 " " - " 
90 " 







60 days after operation. 
91 " 
128 " 
138 " " 
161 " 

212 " 

13 m'ths " " * 
Mast cells ranged from 
0.15 to 0.3 per cent. 

Post-operative leucocytosis of the polynuclear 

LEUCOCYTES, neutrophile type develops promptly, and persists 
in most instances for from four to six weeks, ac- 
cording to Hartmann and Vaquez, 6 but occasionally for a longer 
period. Counts of between 15,000 and 30,000 represent the grade 

1 Tieken, Jour. Amer. Med. Assoc., 1903, vol. xl, p. 887. 

2 Ballance, Practitioner, 1898, vol. Ix, p. 347. 

3 Heaton, Brit. Med. Jour., 1899, vol. ii, p. 476. 

4 Hartmann and Vaquez, cited by Bordet, "Des modifications du sang apres 
la splenectomie," These, Paris, 1897. 

s Staehelin, Deutsch. Arch. f. klin. Med., 1903, vol. Ixxvi, p. 364. 
* Compt. rend. Soc. biol., Paris, 1897, vol. iv, p. 126. 


of leucocytosis ordinarily found, although occasionally the increase 
is far greater 70,000 in a case cited by Czerny, 1 and 75,000 in 
one reported by Hartley. 2 After a number of months, even as 
late as the second or third year after the operation, there is a 
moderate increase in the number of eosinophiles. Small-celled 
lymphocytosis is a frequent but inconstant change, the develop- 
ment and duration of which vary greatly. It is supposed to reflect 
hyperactivity of the lymphatic glands. Eosinophilia commonly 
develops within a few months, as in a case studied by Rauten- 
berg, 3 in which the eosinophiles increased to six times the normal 
percentage by the fourth week' after the operation. In splen- 
ectomized guinea-pigs Kurloff 4 found during the first year after 
the operation a marked lymphocytosis, as high as 60 per cent, 
in some animals, together with a corresponding decrease in the 
number of granular cells, but with no alteration in the num- 
ber of large monoriuclear leucocytes. Eosinophilia became ap- 
parent during the second year, and coincidentally with this change 
a decrease in the lymphocytes to their normal percentage took 
place. In splenectomized dogs Nicholas and Dumoulin 5 noted 
post-operative leucopenia, succeeded by eosinophilia and by 
lymphocytosis, which, after several months, was followed by a 
gradual, progressive lymphocyte decrease. The polynuclear neu- 
trophiles showed relatively high percentages, but were not de- 
cidedly increased. 

The above remarks concerning the differential changes affect- 
ing the leucocytes after splenectomy must be regarded as tenta- 
tive, in view of the fact that sufficient data bearing upon this 
question have not yet accumulated to justify more definite con- 
clusions. Enlargement of the lymph glands, bone pains, and a 
marked susceptibility to various infections are frequent post- 
operative phenomena in splenectomized individuals. 

The table on page 334 shows the condition of the blood in the 
few recorded cases in which thorough blood examinations have 
been carried out. Reports of other cases may be ^ound in Staehe- 
lin's monograph, referred to above. 

It is to be remembered that, aside from the character of the 
splenic lesion, these important factors also determine the degree 
of the post-operative anemia and leucocytosis in this procedure: 
the grade of the preexisting anemia, the amount of hemorrhage 
during the following operation, and the patient's recuperative 
powers. All things being equal, the anemia is least marked and 

1 Loc. cit. 2 Med. News, 1898, vol. Ixxii, p. 417. 

3 Deutsch. Zeitschr. f. Chir., 1902, vol. Ixiv, p. 352. 4 Cited by Ehrlich, loc. cit. 
5 Journ. de phys. et de path, gen., 1903, vol. v, p. 1073. 


^"S-co ~ = 


v ^ c 5 re u 

re c ii 

s . = .? - 


Normal or moderate decre 
averaging to about 60 per ce 
Relatively high or low to 
throcyte loss. 

Normal or moderate decre 
counts averaging about 3,500 
Pallor usually not notably marl 
Poikilocytosis usually abseni 
Erythroblasts extremely rare, 
moblasts predominating. 
Polychromatophilia and b 
stroma degeneration usu 
Changes in diameter variable. 

Subnormal, averaging about o 

Normal or slightly increa 
counts averaging about io,oc 
Polynuclear neutrophiles nor 
or increased, with conseqi 
lymphocyte decrease ; rarely 
increase in lymphocytes. 
Myelocytes rare. 

Eosinoplliles normal or decrea 
Basophiles not increased. 

Usually increased. 

c t^ 

'M g 

J2 s i .i! A u 
~ u o 35 - 
5 o' . = . re O .a 
g o . MJQ <J .S 


1 it - - 


D . 


sf . u 3 '5"2 s re 



.rf 1 1 

O ^ 


1? rr -^ .2 g *~ rt rt v 


V **^- 4^ jj 

<u u 

Ul ^ O 

3 a 3 ^ " ** 



3 0_S 'C 

o w 

re S- 

Si o P ** * ^ "o ? ~ 




<J "rt*> 1 tn Ui ^ O &i "^ 



O*5 ~V 3 o ^ 

-1 D 
U id 


^ M = $ u''" k-re-o " 

- sa i" 5,-S S 

| J 

3 >< 

4>^3 > . 

^ rt 4 m 

ifllilll ict 



C E 5 'in o O* J3tn 
aj ul '5o c w n 



'o o^ o 


a ci u (2 u 

I 1 

7! U 03 




M ^ 

in c i <-> A ; 

^ v 2 '; S 

5 . 8 c B 

CX" S ^2 




g ^ | J8 8 


8 8 U 3 O 2 * 






J ^ .2 Jj'-j . i fe 



aS'^M >> >, S 

frill E =5 & 


ecided decreasi 
o about 60 per c 
slatively very 
tbrocyte loss. 

3 ra rt 2 5~ f J| 

<uog> =-g-= E 
o^a, MOO'S!! -2 


. ^ P tf X *^ ^- r^ 

jw, averaging a 

U 1^8 " | - 

"^ b ? ^ V 

-flsf 28 1"8 If 

! o ^ C" ^ 
ij^^joJ^ s !?* O.S'S 






Q 0.0, U o. u 


S w C/5 W 3 



2 rirz: g ^ i 


flllt I I 

S . 

> en 

re 1? g"-r 3 . C 


"" c >> * *" " 








liSlilli li fl 

't increas 








- ~. S W CQ 


2 a " i> >-c i u c 


_ju re-o g_o g-o - = ' 

lyjf li .^^ i 





fllllJJll i^ 




' ~ * 

4>'i:^jj = < ^O.Q rt J!^ 





U M u 9 v*^ a Ji\E O U 











u JJ u a 









1 ' 











the blood regeneration most prompt in simple wandering spleen 
and in ague cake, while blood deterioration is more marked and 
regeneration slower in rupture of this organ. Splenectomy for 
myelogenous leukemia is almost invariably followed by a pro- 
gressive anemia and leucocytosis, and in nearly all cases by death. 
But five recoveries after removal of the spleen in this disease 
have been reported, 1 the operation having been performed in 
forty-two cases. 

1 Hage'n, Arch. f. klin. Chir., 1900, vol. v, p. 188; also Richardson, cited by 
Warren, loc. cit. 





As a preliminary essential to the intelligent 
FETAL BLOOD, study of the various pathological conditions of 
the blood in children it is necessary briefly to 
refer to certain points of difference in the composition of this tis- 
sue in the child and in the adult. In general terms it may be 
stated that the younger the child, the more unformed are the 
different elements of the blood and the nearer its composition 
resembles the blood of the fetus. 

In fetal blood the specific gravity, both of the whole blood and 
of the serum, is lower than in the adult, and coagulation is very 
slow and imperfect. The erythrocytes vary greatly in size and in 
shape, and are deficient in hemoglobin, which is loosely attached 
to these cells and hence becomes readily dissolved out. Zang- 
meister and Meissl 1 conclude that, in comparison with maternal 
blood, fetal blood is poorer in albumin and nitrogen ahd is less 
active in agglutinative, bactericidal, and immunizing powers. 

Until about the seventh month of intra-uterine life normo- 
blasts constitute the predominating variety of erythrocytes, after 
which period they rapidly diminish in number, until at full term 
few, if any, nucleated red corpuscles are found in the blood. Of 
the different varieties of leucocytes, the mononuclear forms are 
present. in a proportion relatively excessive to the other varieties; 
before the seventh month this lymphocytosis is due to a high 
relative percentage of large lymphocytes, but after this period 
the proportion of small lymphocytes increases, until finally 
they predominate. The percentage of eosinophiles reaches its 
maximum at the seventh month, gradually becoming less and less 
as the end of the intra-uterine life approaches. 

We find, therefore, that in the fetus the blood is characterized 
chiefly by the presence of large numbers of normoblasts, by a 
high relative proportion of mononuclear leucocytes, by a deficiency 

1 Munch, med. Wochenschr., 1903, vol. 1, p. 673. 


of hemoglobin, and by a feeble activity of the serum. The closer 
an infant's blood resembles this picture, the "younger" in point 
of development is such blood considered, and the more strongly is 
it said to revert to a "young" or " embryonal" type. The deficient 
defensive potency of the Blood of the young baby may account for 
the proneness of infants toward infections in general. 

At birth the blood of the full-term infant is of 
THE BLOOD higher specific gravity and richer in hemoglobin 

AT BIRTH, and in corpuscular elements than that of the older 
child or of the adult. 

The specific gravity of the blood of the average healthy infant 
at the time of birth and during the first few weeks of life is, ap- 
proximately, i. 066. For normal children the average, which is 
reached by the beginning of the second year, varies from 1.050 to 
1.058, being slightly higher in boys than in girls. 

The maximum amount of hemoglobin is found at birth, the 
percentage at this time ranging from 100 to 104, according to 
the investigations of Hammerschlag. 1 After birth the amount of 
hemoglobin immediately begins to diminish, the minimum, which 
may be as low as 55 or 60 per cent., being attained by the end 
of the third week of life. It remains at or about this minimum 
for a variable period of time sometimes for as long as six months 
and then gradually begins to increase. 

At birth the number of erythrocytes in the peripheral blood is 
decidedly higher than normal, counts of between 5,500,000 and 
6,000,000 cells per being found at this time, the highest 
figures being observed in those cases in which ligation of the 
umbilical cord has been delayed. During the first twenty- 
four hours of extra-uterine life this polycythemia becomes still 
more marked, so that the number of corpuscles per may 
reach a maximum of from 7,000,000 to 8,000,000, and sometimes 
higher, by the end of the first day. Beginning with the second day, 
a gradual diminution in the number of these" 1 cells is noticed, and 
the normal 5,000,000 per is reached by the end of the first 
week or ten days. Hayem 2 emphasizes the fact that the fluctua- 
tions in the number of erythrocytes during the early days of life 
stand in inverse ratio to the variations* in the weight of the child, 
the maximum number being found at the time of the infant's 
minimum weight, while as the child begins to gain in weight the 
count decreases. .Scruff 3 is inclined to attribute these fluctuations 
to the amount of liquids in the body, the result of feeding, showing 

1 Centralbl. f. klin. Med., 1891, vol. xii, p. 825. 

3 "Du Sang," etc., Paris, 1889. 

8 Zeitschr. f. Heilk., 1890, vol. xi, p. 17. 


that in fasting children the counts are always higher than in those 
fed at frequent intervals. 

Whatever may be the exact manner of their production, it is 
evident that these fluctuations are to be regarded as purely physio- 
logical in character, depending upon concentration and dilution 
of the blood, rather than as an expression of involvement of the 
blood-making organs. 

The erythrocytes vary greatly in size during the first few days 
of post-natal life, the diameter of some cells being as small as 
3.25 ft and of others as large as 10.25 //. Many observers have 
noticed that the small-sized cells, as a rule, predominate. Such a 
microcytosis in the adult would mean well-defined anemia. 

Nucleated, erythrocytes of the normoblastic type may or may 
not be present in the blood of new-born infants; they are com- 
monly found in large numbers in the prematurely-born child, and 
also occur less numerously in many fully developed babies, not- 
withstanding views to the contrary expressed by some observers, 
notably by Hayem 1 and by Fischl. 2 In most cases normoblasts 
disappear from the blood after the first few days of life, and their 
presence after the sixth month should always be regarded as 

The number of leucocytes at birth averages about 20,000 per, the normal average for young infants, 15,000 per, be- 
ing reached by the end of the first week, after numerical fluctua- 
tions similar to those affecting the erythrocytes. From the second 
or third week until the sixth month a count from 10,000 to 14,000 
may be regarded as normal, while for the child of one year of age 
the average is about 10,000. By the sixth year the number of 
leucocytes falls to the number normal for the adult, 7500 per The following excellent table from Rotch 3 shows these 
average counts of erytKrocytes and leucocytes in children from 
birth until the sixth year of age: 


At birth 5,900,000 21,000 (26,000 to 36,000 

after first feeding). 

End of i st day . . . . 7,600,000 to 8,000,000 24,000 

2d " Generally increased. . 30,000 

4th ' 6,000,000 20,000 

" yth " 5,000,000 15,000 

loth day 10,000 to 14,000 

i2th to iSth day 12,000 

ist year .... 10,000 

6th year and upward . 75 

1 Loc. cit. 

2 Zeitschr. f. Heilk., 1892, vol. xiii, p. 277. 

3 "Pediatrics," Philadelphia, 1896, p. 342. 


The influence of the initial feeding in infants produces a 
marked leucocytosis, the increase amounting to from 5000 to 
15,000 per, as shown by the above table. It is probable 
that the habitual leucocytosis of early childhood is largely refer- 
able to a more or less continuous digestion leucocytosis. (For a 
further discussion of this question see " Digestion Leucocytosis," 
p. 231.) 

The blood of infants and of young children differs greatly from 
that of the adult in the relative proportions of the different forms 
of leucocytes, these qualitative differences becoming less and less 
apparent as the child grows older, and not usually persisting 
beyond the tenth year. In' general terms it may be said that 
these dissimilarities are striking in relation to the youth of the 
child. Compared to the adult, a differential count of the leuco- 
cytes in the child shows that the relative percentage of lympho- 
cytes is more than twice as great, and of polynuclear neutrophiles 
one-half as great, while the proportion of eosinophiles is frequently 
much higher. In the following table, based upon data given by 
Gundobin, 1 these points of difference are contrasted: 


Small lymphocytes . .50 to 70 per cent. 20 to 30 per cent. 

Large lymphocytes 
and transitional 
forms 6 " 14 " " 4 " 8 " 

Polynuclear neutro- 
philes 28 " 40 " " 60 " 75 " " 

Eosinophiles 0.5 " 10 " " 0.5 " 5 " " 

It is important to take into account these differences in mak- 
ing blood examinations in children, in whom one must expect to 
find percentages of lymphocytes which in the adult would be 
regarded as abnormally high. 

Many of the lymphocytes differ from corresponding cells in the 
adult, chiefly in being of larger size, in exhibiting a greater variety 
of nuclear figures, and in having a more basic tendency. Karnizki 2 
also professes to find an occasional myelocyte in the blood of the 
healthy child, and to trace a decided morphological resemblance 
between the lymphocytes, the transitional forms, and the neu- 
trophiles. While it is true that myelocytes appear in the infant's 
blood upon slight provocation, one can scarcely regard them as 
normal elements; nor can a relationship between the lymphatic 
and the myelogenous leucocytes be established more clearly in 
the child than in the adult. 

Leucocytosis in children is of extremely common occurrence, 

1 Jahrb. f. Kinderheilk., 1893, vol. xxxv, p. 187. 

2 Arch. f. Kinderheilk., 1903, vol. xxxvi, p. 42. 


often arising from causes of the most trivial character, and de- 
veloping to a greater degree and with much more rapidity than 
in the adult. It is therefore to be regarded with less significance 
than when it occurs in the mature. Usually the polynuclear cells 
are chiefly involved in the increase, but the inclination of the 
blood of children to revert to the embryonic type appears to 
cause, in many cases, a disproportionate increase of lymphocytes 
in relation to the other forms, this peculiarity being especially 
true of the various pathological leucocytoses. Physiological leu- 
cocytosis in children usually affects chiefly the polynuclear neu- 
trophile cells. 


Children, as a class, are peculiarly susceptible 
FREQUENCY, to anemia, for they appear to lack resisting pow- 
ers against the influence of causes tending to pro- 
duce pathological alterations in the blood. Thus, it is found that 
the same factors which in the adult have little or no effect upon 
the blood, are capable of producing profound alterations in its 
composition in the child. Severe anemias may arise in children 
from apparently the most trivial causes; slight hemorrhage from 
the navel, for instance, may light up an anemia of an intensity out 
of all proportion to the actual amount of the blood loss, while 
minor lesions of the gastro-intestinal tract are commonly asso- 
ciated with blood deterioration of a severe type. 

It is a notable fact that in anemic children a 

GENERAL predominant tendency exists toward a reversion 

CHARACTER- of the blood to a less mature histological type, 

ISTICS. such as that found in the blood of the fetus. 

Thus in children anemia of a type which in the 

adult is unattended by qualitative changes in the corpuscles is 

commonly associated with the presence of large numbers of 

nucleated erythrocytes, these cells being far more numerous than 

the severity of the anemia would seem to warrant. Megaloblasts 

are not of the same significance in infancy as in adult life. They 

may occur in anemias of but moderate intensity in fact, they may 

predominate, as Morse has shown, 1 without necessarily constituting 

a sign of fatal blood disease. In Riviere's experience 2 megalo- 

blasts, in ordinary well-marked infantile anemias, may constitute 

between 20 and 50 per cent, of the erythroblasts, the ratio of 

megaloblastic forms apparently standing in no definite parallelism 

1 Boston Med. and Surg. Jour., 1903, vol. cxlviii, p. 573. 

2 Lancet, 1903, vol. ii, p. 1419. 


to the severity of the anemia of which they are symptomatic. 
Poikilocytosis and deformities in the size of the corpuscles also 
occur with far greater frequency than in the adult. Regeneration 
of the blood takes place slpwly. The oligochromemia is relatively 
greater than the oligocythemia in most anemias of children, this 
being due probably to the fact that the hemoglobin is peculiarly 
prone to separate from the corpuscular stroma. Owing to this fact 
low color indices, as in chlorosis, are common, irrespective of the 
degree of corpuscular diminution. 

Myelo'cytes are commonly found in the blood in all the anemias 
of children ; they are present .in larger relative percentages and in 
less severe pathological conditions than in the adult. The writer 
has noted the frequency, almost the constancy, with which typical 
mast cells and the fine basophiles occur in the specific infections in 
children, notably in enteric fever and in tuberculosis. 

Leucocytosis, often lymphocytosis, and enlargement of the 
spleen are frequently associated with all forms of anemia in the 
young ; and although these conditions are likely to coexist, this is by 
no means the invariable rule. Splenic enlargement is especially 
common in the anemias due to syphilis, rachitis, tuberculosis, 
gastro-intestinal disease, malaria, and septic infection. 

To epitomize, in the anemias of infancy and childhood the 
following prominent features of the blood are found: (a) The 
frequency of a low color index; (ft) the common occurrence of 
erythroblasts and of deformities affecting the shape and size of 
the erythrocytes ; (c) a tendency toward leucocytosis and splenic 
enlargement, and (d) the frequency of myelocytes and mast cells. 
It is owing to these peculiarities that the clas- 
CLASSIFICA- sification of the anemias of children is such a dif- 
TION. ficult matter. The older classifications, based 
upon the nature of the causal factors of the an- 
emia and upon the presence or absence of enlargement of the 
spleen, have failed in many respects to prove adequate, so that it 
becomes necessary to adopt a simpler and more comprehensive 
division from which no exceptions need be made in the individual 
case. Such a classification has been suggested by Morse. 1 This 
author assuming, and rightly so, that* chlorosis is a condition 
wholly foreign to infantile life, and that the disease described 
by von Jaksch as "anemia infantum pseudoleukemica " does not 
represent a distinct clinical condition, proposes this excellent 
classification, slightly modified from that of Monti: 
Primary Anemia. Pernicious anemia. 


1 Arch. Pediat., 1898, vol. xv, p. 815. 


Secondary Anemia. Mild anemia. 

Mild anemia with leucocytosis. 
Severe anemia. 

Severe anemia with leucocytosis. 
Pernicious anemia is rare in the young, and 

PRIMARY likely to be mistaken for other forms of severe an- 
ANEMIA. emia secondary to various conditions. It is prob- 
able that many of the reported cases of Biermer's 
anemia in infants were in reality examples of severe secondary 
anemia. Hutchinson, 1 in his Goulstonian lectures, gives the 
data of ii authentic cases, the total number recorded up to the 
present time. The apparent tendency of pernicious anemia in 
children to become transformed into leukemia is doubtless more 
fanciful than real, a remark which is equally true of those few 
reported instances of the conversion of leukemia into pernicious 
anemia. In the first case the erroneous impression may arise from 
such evidence as marked enlargement of the spleen associated with 
a high leucocytosis; in the second, a temporary disappearance of 
the myelogenous blood-picture plus an aggravation of the exist- 
ing anemia may be sufficient to convey the false impression. It 
must be admitted that these atypical blood changes, so common 
in young children, are highly confusing and difficult to interpret 
without the closest observation and the correlation of other clin- 
ical signs. (See " Leukanemia," p. 290.) 

Infantile pernicious anemia is characterized by the same blood 
changes that are found in the adult, in so far as striking oligocy- 
themia and deformities of the erythrocytes are concerned, but 
the blood often fails to show a high color index. A prevalence 
of megaloblasts and 'of megalocytes is not pathognomonic, as 
it is in the adult. 

Splenic anemia, which is none too certain an entity in the adult, 
can rarely be identified in the child. The splenic enlargement 
in most reputed instances of this condition in early life must be 
regarded as symptomatic of nutritive disturbance, rather than as 
a distinctive factor of anemia. Hutchinson 2 uses the terms in- 
fantile splenic anemia 'and infantile pseudoleukemic anemia 
synonymously, and believes that the former is entirely distinct 
from splenic anemia of the adult. It must be admitted, however, 
that rarely the adult type of the disease is met with in children, as 
shown by the cases reported by Williamson, 3 Hunt, 4 Rolleston, 5 

1 Lancet, 1904, vol. i, pp. 1253 and 1325, and 1402. 

2 Loc. cil. * Med. Chronicle, 1893, vol. xviii, p. 103. 
4 Trans. Path. Soc. London, 1899, vol. 1, p. 209. 

6 Clin. Jour., 1902, vol. xix, p. 401. 


Hamill, 1 and others. Such cases show the same type of anemia, 
leucopenia, and lymphocytosis found in the blood of the adult 
suffering from this disease. 

Leukemia in children js uncommon, but instances have been 
reported during all stages of infancy and childhood, even in the 
new-born. Acute forms of the disease are most frequently met 
with, the great majority of cases, according to Holt, 2 proving 
fatal within a year from the appearance of the first symptoms, 
while in many the disease runs its course in a few weeks. Lym- 
phatic leukemia is about five times as common in children as 
myelogenous. Male children are more commonly leukemic than 
female. Conditions such as rachitis, syphilis, and malarial fever 
have been regarded by some authors as possessing a certain 
amount of importance as etiological factors, but in the vast ma- 
jority of cases the cause of the disease is entirely obscure. 

Of the several collected reports of leukemia in children, the 
two articles of Morse, giving a total of 27 cases, are by far the 
most valuable. In his first communication 3 20 cases were re- 
corded, including one of his own, tabulated below, but of this 
series the diagnosis, in the great majority of instances being based 
either upon clinical symptoms or upon inadequate examination 
of the blood, the reporter is led to remark that "it is highly probable 
that not more than half, perhaps not more than a third, of these 
were really cases of leukemia." In Morse's second article, 4 
which deals with the acute form of the disease, seven cases, again 
including one of his own, also recorded below, are reported. 

Although the literature of pediatrics is fairly rich in alleged 
examples of leukemia in children, the cases, with but a few ex- 
ceptions, are reported in so unsatisfactory a manner that they 
cannot be regarded without reserve as typical. Those reported 
prior to the publication of Morse's first article in 1894 must all 
be open to criticism, owing to the general disregard for differen- 
tial counts shown by the various authors, and, strangely enough, 
this criticism must hold true for many cases recorded during the 
past ten years. The author has been able to collect 21 cases 
(Table XVIII), in all of which the differential count of leuco- 
cytes leaves no doubt as to the precise character of the disease. 

In addition to the above cases several others have been reported 
in which the differential count of leucocytes has been either faultily 
made or entirely neglected. Thus, Pollman 5 believes that he has 

1 Arch. Pediat., 1902, vol. xix, p. 641. 

2 "The Diseases of Infancy and Childhood," New York, 1897, p. 806. 

3 Boston Med. and Surg. Jour., 1894, vol. cxxxi, p. 133. 

4 Arch. Pediat., 1898, vol. xv, p. 330. 

5 Munch, med. Wochenschr., 1898, vol. xlv, p. 44. 




tn t/i in uj ) > S 
33. ..3 .33 5 > 

O O <-> U O O OCO O >% 

c c . '.3 '-3 c '.sec ** u k 

owcscjsjf rtoo M e O 

bC OC f 1 t-| r* tUO r^ bC bC M ^*_, W 

J2 -2 O. G. Q. , G.J2 J2 tn 3 
O O C C P *O p O &f ^* g> " < 

' * H , ; o to 




o a 


"S ' 


C 'f- 1 . 

^-^^^"S^S *^a ? &S 

i | & ml | ii i 8 4 

1! ^it 1 *f 1 S ! 

^MJn"^^ ^^^ % . &> 

s-gi i i itisi i j^ i -d s 

5|| 1 1 ||1| |,||8. t & ** 
t Is c g !* H^ lfl-2 ! js "8 II 

s- fi^ s ji s lrji ?| 2 ic 1 li 

g iS&^^^^lx g\a g,^6x.i2^ S ^ o 

.5 j|jlili|{l!i I^Jl I A -2 ^ 

^ < < < S PM S ti)-^ H 3 




3 ^ pQ (J 

ooqqqqio OMOO aiviS'- 3 
6 ON d d d>d dcNpi _ < ^ J" 3 3 '3 
PIVC f t-~.-^- :s_b<oC 
; 5 *^ *^1 ^ ^^ d 





oo .^ .2 " pL-S 

vO O O O *O O O O ID n3 ^ ^^ to ^ 

M WOOO O tO ^1" I ^J ^ _ij rC **-( *5 

i i 





to t^- q q t^ TJ- ^"ppo\ o^^3^ 


MPI PI I-IPI o.'Srt c bCw 



-HAT aoavq 




o o PO w t^ t'^ be ^ 
to P! t^- 06 d M 

00 00 ON M C 





M ooorf oovq c.s~ 
M w vo PI PO f^ to >- J- ^ "^ 

P) O p c 





* O O O O O ' O O tr O 

5588888 ^88 is^o 

j^ d" c? p^ d" 06" o ^ oo~ c 
ySiMpi^- ^'oc! ^."-^2 
~ _ 2 d3?:S.|| 




rto"od*c5'o6'c.r -gcfi^ ^"*-^ ^ -o 2 

DOO^OOO >Ot>* ^ . ^ g g 




i> O O *O O l ^ " C^ i^ U """"^"Ii! 
Q^*" '^P J w'S vf J -"3. 



HNOO ^ ^ ^ M ONto gJ^CWCl 



^ J^^J'I^I 

o o M "2 iii T! 
c p^, j . *** ~ jS y 

rt - ' ' u 3 fi "5 

2"u " "^"M' ^ <? .J ^ rt H '" "^ 
5 fl "S a 2 i'-c * i - w -g 

^^rt^ J=rt3 M3 

U^ S hpq< 58 





a -a 

ca es 


- : 

t n 
: 43 

rt cs rt rt 
C J= J3 -C 









S E E ^ 



A few erythroblasts. 
To 200 leucocytes, 92 normo- 
blasts and 2 megaloblasts 

were found. Polychromato- 

philia and poikilocytosis. 
No erythroblasts. Slight poly- 

One normoblast to 105 leuco- 
















O to 

co r^- 






o q 





q q 

o\ 6 

\O co 


O fO 



o o 



O t^> 

H r** O w 



00 * 


X) 00 

O\ Ox 

O^OO f^5 O 1 * 



( 1 




O O 

o o o o 
o o o o 



to to 

00 O 





0_ 0_ 

M W 









O O O O 
8O O O 

[ ' 



tO IN 

00 M 






00 O IO O 

vO vo 00 O 

to COVO M 


M T) 



M 04 




00 O 

M d 



O O 

tH Tf 

to O ON 
w N co 



00 C 





55 ^ ^ r 

HN ^>to >-> 

co ON w M 



6 >o 




PQ co 





Carpenter * 
Einhorn 7 
Guinon and Jolly 8 
McCaw 9 

Geissler and Japha 10 . 


vc :H 




seen a case of myelogenous leukemia in a new-born infant, the 
count on the fourteenth day after birth being 2,500,000 erythro- 
cytes and 312,500 leucocytes per The latter consisted 
chiefly of "large mononucleated cells, with large, distinct nuclei 
and an abundance of protoplasm." Nucleated forms of ery- 
throcytes were not found. Cassel, 1 in addition to his own case, 
tabulated above, has collected four others occurring in children 
under fourteen years of age. 

Theodor 2 has collected from German literature six cases of 
acute leukemia in children between the ages of two and one-half 
and eight years of age, and also reports one of his own, apparently 
of the lymphatic form, in a boy of four years. No actual numer- 
ical estimates of the corpuscles are given, but the proportion of 
leucocytes to erythrocytes is stated to vary from i : 9 to i 13. The 
differential count of leucocytes is also very inexact. The greater 
percentage apparently consisted of lymphocytes; myelocytes were 
fairly numerous, many of them containing mitotic figures, and 
normoblasts and megaloblasts were present in large numbers. 
Gilbert and Weil, 3 in 1899, published data of five acute lymphatic 
cases in children between the first and tenth years. Charon and 
Gratea 4 report a case of myelogenous leukemia in a child of eight 
years, the percentage of hemoglobin being 39, the erythrocyte 
count 880,000, and the leucocyte count 305,000. Instances of 
acute lymphatic leukemia in children have also been reported, 
with more or less accuracy, by Bradford and Shaw, 5 by Guinon 
and Jolly, 8 by Haushalter and Richon, 7 and by Bloch and 
Hirschfeld. 8 ' 

In the above classification of the secondary 
SECONDARY anemias, under the heading of mild anemia are 

ANEMIA. included those cases characterized by trifling re- 
duction in the 'hemoglobin percentage and num- 
ber of erythrocytes, and by an absence of histological alterations 
in these cells. The color index in these cases is usually i.oo, or 
slightly- below, uncommonly falling to a low figure. 

The term severe anemia includes cases having marked diminu- 
tion of hemoglobin and erythrocytes, associated with deformities 
of shape and of size and nucleation of these cells. The hemo- 

1 Berlin, klin. Wochenschr., 1898, vol. xxxv, p. 76. 

2 Arch. f. Kinderheilk., 1897, vol. xxii, p. 47. 
8 Arch, de med. exper., 1899, vol. ii, p. 157. 

4 Bull. Soc. roy. sc. me'd. et nat., Brussels, 1896, vol. liv, p. 63. 
* Trans. Roy. Med.-chir. Soc., 1898, vol. Ixxxi, p. 343. 
8 Rev. mens. des mal. de 1'Enf., 1899, vol. i, p. 262. 

7 Arch, de me'd. des Enf., 1899, vol. ii, p.. 356. 

8 Zeitschr. f. klin. Med., 1900, vol. xxxix, p. 32. 


globin loss is especially marked, often being only one- quarter or 
one-third of normal, and the color index is low. 

Anemias with leucocytosis, whether of mild or of severe type, 
are generally marked by a greater degree of hemoglobin and 
corpuscular decrease than anemias without leucocytosis. The 
leucocytosis is moderate in the milder forms, but in severe cases 
the increase in the number of leucocytes often appears to be pro- 
gressive, and the relative number of leucocytes to erythrocytes 
occasionally attains the proportion of i to 100. 

Histological changes in the erythrocytes are more striking in 
grave anemias with leucocytosis than in grave anemias pure and 
simple. This is especially true of the changes relating to nuclea- 
tion of the cells, normoblasts and atypical forms being very nu- 
merous in the former class. As already stated, megaloblasts do 
not necessarily mean a fatal nor even an especially intense anemia. 

Regarding the etiological factors of these secondary anemias, 
the following groups of causes are given by Monti 1 : 

1. CONGENITAL Syphilis, tuberculosis, and other infections. 

Hemorrhagic. From navel, from circumcision, etc. 
Purpuric diseases. 

Malnutrition, improper hygiene, etc. , 
Syphilis, rachitis, and tuberculosis. 

2. ACQUIRED Gastro-intestinal diseases. 

Visceral diseases. 

General. Febrile diseases. 

Septic infections. 
Malignant growths. 

Syphilis, either congenital or acquired, is responsible for a 
large proportion of the cases of anemia in children, especially those 
of a severe type, associated with enlargement of the spleen, and 
often also with enlargement of the lymphatic glands. The hemo- 
globin loss is in most instances disproportionately greater than the 
corpuscular decrease, so that low color indices are especially 
common in this disease the misnamed "chlorosis" of syphilis. 
Deformities and nucleation of the .erythrocytes are common in 
the severer types, and in such forms polychromatophilic changes 
and excessive decrease in the number of erythrocytes are usually 
present. A leucocyte increase is present in the secondary stage, 
and is usually associated with the grave anemia of this disease; 
the relative percentage of lymphocytes is increased, and of poly- 
morphous forms decreased; and small numbers of myelocytes 
are common. From the writer's experience, in the average case, 
of moderate severity the percentage of hemoglobin varies from 

1 Wien. med. Wochenschr., 1894, vol. xliv, pp. 401, 464, 516, 560, and 613. 


about 40 to 50, the erythrocytes are reduced to about 3,000,000 to 
3,500,000 per, and the leucocyte count is in the neighbor- 
hood of 20,000; but in severe cases the erythrocytes may be re- 
duced to 1,000,000, and the leucocytes increased to 50,000 or 
more. As in the adult, Justus' test proves of value in the diagnosis 
of many anomalous cases. Labbe and Armand Delille 1 insist that 
in congenital syphilis the blood picture may precisely resemble 
that of von Jaksch's disease (q. v.}, but that the blood promptly 
returns to normal after energetic mercurialization. 

In rachitis there is usually well-marked anemia, accompanied 
by decided enlargement of the spleen, such cases having most 
decided blood changes. The hemoglobin percentage tends to 
range lower than the percentage of corpuscles, so that low color 
indices prevail; but in the individual case neither the oligo- 
chromemia nor the oligocythemia is generally as marked as in 
syphilis. In severe cases, deformed and nucleated erythrocytes 
and a small percentage of myelocytes are commonly found. The 
number of leucocytes is, as a rule, moderately increased, and 
relatively high percentages of lymphocytes are common. 

Uncomplicated tuberculosis, due to pure infections with the 
tubercle bacillus, produces an anemia which varies in degree 
with the severity of the constitutional effects of the disease. The 
hemoglobin and erythrocytes are usually but slightly decreased, 
the former suffering a relatively greater loss, and the number of 
leucocytes does not rise above normal. If to the tuberculous 
process a septic infection is superadded, the anemia becomes 
severer, and leucocytosis,, involving the polymorphous forms of 
cells, occurs. Splenic enlargement is common and sometimes 
marked. The anemia of tuberculosis is in no way referable to % 
the infection itself, but depends upon the drain on the albumins 
of the blood due to the presence of a long- continued cachexia 
and upon secondary infections. 

Gastro-intestinal diseases, especially those of chronic character, 
cause most marked anemia. Chronic inflammations of the intes- 
tines strikingly affect the blood, the percentage of hemoglobin 
frequently falling to one-quarter of normal or even lower, and the 
number of erythrocytes being decreased to one-half of normal or 
lower. Deformities affecting the shape and size of the erythrocytes 
and nucleation of these cells are of frequent occurrence. A leu- 
cocyte increase, involving in many instances the lymphocytes, 
is usually present, and small percentages of myelocytes have 
been observed. Splenic enlargement is frequently a conspicu- 
ous clinical sign. It should be remembered that in acute forms 

1 Sem. med., 1963, vol. xxiii, p. 50. 


of gastro-intestinal disorders, in which profuse diarrhea and vom- 
iting occur, concentration of the blood takes place, causing tem- 
porary polycythemia Avhich may for a time hide the real degree 
of the blood deterioration. The anemias found in this class of 
diseases are apparently to a large extent autointoxicative in char- 
acter, depending to a less degree upon insufficient nutrition. 

In enteric fever the blood picture does not differ essentially 
from that seen in the adult suffering from this affection, absence 
of leucocytosis or leucopenia with progressive anemia being 
found with great constancy. The difference is simply one of 
degree, the changes developing more rapidly but persisting for a 
shorter period in the child than in the adult. Churchill's studies l 
show that the leucopenia is most decided during the second week 
of the fever; and that the anemia is especially apparent during the 
first three weeks, after which the hemoglobin and erythrocytes 
begin steadily to increase, reaching normal by about the fifth 
week. In 12 cases studied by Stengel and White 2 the hemoglobin 
ranged from 68 to 83 per cent., the erythrocytes from 3,320,000 to 
5,200,000, and the leucocytes (in uncomplicated cases) from 3800 
to 12,320. Polynuclear leucocytosis was observed in 3 cases as the 
result of inflammatory complications. Two uncomplicated cases 
showed fractional percentages of myelocytes, but there were no 
other differential changes of any consequence. Head 3 believes that 
in uncomplicated cases the leucocytes never number more than 
10,000 per The writer has found that the alkalinity of the 
blood varies within wide limits in infantile typhoid, tests by 
Engel's method in 6 consecutive cases showing a range of from 
373 to 692 mgm. NaOH. The rapidity of coagulation also was 
found to vary greatly, clotting taking place in as short a time as 
thirty-seven seconds in one instance, and not occurring for four 
minutes and thirty-five seconds in anotlier. Morse 4 concludes 
that the serum test appears earlier, is less marked, and persists for 
a shorter period in children than in adults. In a nursling it 
should be remembered that a positive reaction may be of uncertain 
value, for the reason that the agglutinating power may be trans- 
mitted from mother to child through the milk, both during the 
active stages of the disease and also during and after convales- 

Under the title "anemia infantum pseudoleukemica" von 
Jaksch 5 has described a condition which he regards as a form of 

1 Boston Med. and Surg. Jour., 1903, vol. clviii, p. 692. 

2 Arch. Pediat., 1901, vol. xviii, pp. 241 and 321. 

3 Ibid., 1902, vol. xix, p. 253. 4 Ibid., igoi, vol. xviii, p. 338. 
s Wien. klin. Wochenschr., 1889, vol. ii, p. 435. 


primary anemia peculiar to the young child. The blood changes, 
which are not characteristic of the disease in question, as this 
author admits, consist of (a) marked oligocythemia and oligo- 
chromemia; (6) extensive and persistent leucocyte increase; and 
(c) striking structural alterations in the erythrocytes. Associated 
with these changes in the blood, and of equal importance in diag- 
nosing the disease, constant enlargement of the spleen and, less 
commonly, enlargement of the liver are found. 

The number of erythrocytes is greatly decreased, usually to 
from 2,000,000 to 3,000,000 per, but sometimes falling 
to 1,000,000 or even to a lower figure, as in one of von Jaksch's 
cases, in which the count was only 820,000. The hemoglobin 
loss is also great relatively more so than the corpuscular decrease. 

The leucocyte gain is decided, averaging, in the majority of 
cases, from 30,000 to 50,000 corpuscles per, and in some 
instances exceeding 100,000. In some cases the increase involves 
principally the polynuclear neutrophiles, while in others the lym- 
phocytes are the cells chiefly affected. The cells show a most 
striking dissimilarity of form and of size, and a highly confusing 
variety of forms atypical in size, shape, and nuclear morphology 
is encountered. This "polymorphous" state of the leucocytes 
is a point insisted upon by von Jaksch in his description of the 
condition. The number of eosinophiles varies within wide 
limits, the percentage of these cells being normal, decreased, or 
increased. Small percentages of myelocytes have been observed. 

The histological changes affecting the erythrocytes consist in 
marked poikilocytosis, deformities of size, loss of color, and nu- 
cleation. Poikilocytes, megalocytes, and microcytes occur in 
large numbers, and the pallor of most of the corpuscles is ex- 
treme. Normoblasts are the most common form of erythro- 
blasts observed, but occasionally the occurrence of atypical forms, 
small and large, and- of megaloblasts has been noted. Karyokin- 
etic changes in these cells are not uncommonly seen, and poly- 
chromatophilia is of frequent occurrence. 

The spleen 'is enlarged in all cases, sometimes moderately, but 
often very greatly, so that the organ extends far below the cos- 
tal margin, and occupies the entire upper left part of the abdomi- 
nal cavity. The spleen is extremely indurated, and may show 
capsular thickening from perisplenitis. The increase in the size 
of the organ is due to a hyperplasia. Enlargement of the liver 
is not constant in all cases, and when present does not reach a 
size corresponding to that of the spleen, as is the case in leukemia ; 
the lower border of the liver is not rounded, but distinctly sharp. 
In a certain proportion of cases' the lymphatic glands are slightly 


enlarged, but never to any notable extent. Changes in the bone 
marrow, common to any severe anemia, have been observed in 
some cases., 

The disease occurs most frequently in infants between the ages 
of seven and twelve months, and is rarely met with in children 
over four years old. By some writers it is supposed to be slightly 
more common in children of the male sex. 

In many cases a previous history of rachitis, syphilis, or long- 
standing gastro-intestinal disease is obtained, although von 
Jaksch denies the existence of these etiological factors in his cases. 

The onset of the symptoms is slow and insidious, and the pal- 
lor of the skin, blanching of the mucous membranes, and other 
signs of anemia slowly develop, with the gradual enlargement of 
the spleen, until these clinical manifestations become marked. 
In all cases there is excessive loss of strength, and in a great 
many a high degree of emaciation. 

Von Jaksch's disease, if untreated, tends to pursue a progres- 
sively grave course, ending fatally; but, under suitable treatment, 
the splenic tumor decreases in size, the leucocytosis disappears 
and the hemoglobin and erythrocytes return to normal. 

Pseudoleukemic anemia of infants is not generally considered 
as a separate clinical entity, but is regarded rather as a form of 
severe secondary anemia associated with marked leucocytosis and 
splenic enlargement. It may be due to a number of different 
causes, the most prominent among which are syphilis, rachitis, 
and chronic gastro-intestinal disease. The conflicting reports of 
different authors concerning this disease, and the incompleteness 
with which the leucocytes have been studied in many instances, 
render it probable that in some of the reported cases pernicious 
anemia and leukemia have masqueraded, as typical examples of 
the condition described by von Jaksch. 

Bacleriemia, generally referable to preagonal infections, 
appears to occur with great frequency in the young child during 
the course of many acute diseases. Delestre's careful studies * of 
general blood infections in children tend to show that infants born 
before full term are peculiarly susceptible to this condition. Using 
careful technic, this author examined 40 children, ranging in age 
from a few days to four years, all of whom were believed to be 
suffering from infections which bade fair to end fatally within a 
few days, at the latest. Of the 32 fatal cases of this series, bacteria 
were found in the blood during life in 14, while of the 8 who re- 
covered, but one gave a positive result. The bacterium found 
with greatest frequency was the streptococcus, while staphylococci, 

1 Annal. de gynecol. et .d'obstet., 1901, vol. Iv, p. 51. 


pneumococci, colon bacilli, and influenza bacilli were isolated more 
rarely. It was furthermore shown that premature babies seemed 
especially susceptible to streptococcus and colon infections, and 
that nursing infants several months old were more prone to suffer 
from the effects of the staphylococcus. 

The blood changes occurring in pertussis, pneumonia, diph- 
theria, scarlet lever, measles, varicella, and other infectious diseases 
of childhood are considered in Section VII. 





The rate of coagulation is, as a rule, somewhat 
GENERAL slower than normal. Hyperinosis is conspicuous, 
FEATURES, and under the microscope the fibrin network ap- 
pears abnormally dense and thick. The iodin 
reaction may be detected in the dried blood film by the method 
described in a previous section. (See p. 226.) These remarks, 
as well as those which follow, do not apply to purely tuberculous or 
"cold" abscesses, the effects of which are referred to elsewhere. 
If the absorption of toxic material from an ab- 
HEMOGLOBIN scess is great enough to produce a systemic effect 
AND upon the patient, anemia of an intensity parallel 

ERYTHROCYTES. to the severity of the poisoning sooner or later 
develops. This fact is sufficient to explain why 
the grades of anemia in .purulent conditions vary within such wide 
limits. The size and the site of the abscess do not appear pri- 
marily to determine the degree of the associated blood changes, 
although, other circumstances being equal, a large, deep-seated 
collection of pus is likely to have a more harmful effect than one 
of small size and superficial situation. Chronicity of the lesion 
seems to go hand 'in hand with an increase in the blood deterio- 
ration few persons harboring pus for a protracted period fail to 
show decided signs of anemia. 

In many cases, especially the acute, the only noticeable change 
is a moderate oligochromemia, but in chronic cases different de- 
grees of ordinary secondary anemia are commonly encountered, 
amounting in an exceptional instance to a reduction of hemo- 
globin to as low as 20 or 30 per cent, of the normal standard, and 
to an erythrocyte decrease to between 2,000,000 and 3,000,000 
cells to the Such profound losses are, of course, unusual, 
for in the majority of patients with anemia the hemoglobin is 
above 50, and the corpuscles above 60, per cent, of normal. The 
average color index for 134 German Hospital cases, listed below, 



was 0.77. The condition of the hemoglobin and erythrocytes 
in these patients is shown by the following summary: 




From 90-100 in ..... 4 Above 5,000,000 

80-90 ' ..... 15 From 4,000,000-5,000,000 ". .41 

70-80 < . ...-27 3,000,000-4,000,000 "..68 

" 60-70 " ..... 23 .. 

..... 23 2,000,000-3,000,000".. 9 

" .22 1,000,000-2,000,000".. 6 

" 30-40 " ..... 13 

!" " 20-30 " ..... 7 

Average, 59 per cent. Average, 3,797,470 per 

Maximum, 95 " Maximum, 5,970,000 " " 

Minimum, 20 Minimum, 1,320,000 " 

If marked anemia exists, a variable grade of cell deformity, 
atypical staining, and nucleation is also to be observed. If the 
latter change is evident, it will be found that the great majority, if 
not all, of the nucleated corpuscles belong to the normoblastic class. 
Practically the same influences governing the 
LEUCOCYTES, behavior of the leucocytes in most other infec- 
tions also determine their increase and decrease 
in abscess. Thus, in both trivial and in extensive pus foci the 
number of leucocytes may be normal or even subnormal; in the 
former instance because systemic reaction is not provoked, and 
in the latter because it is overpowered. Leucocytosis may also 
be absent in case toxic absorption is impossible, owing to the com- 
plete walling-off of the abscess. In all other instances save these 
a definite and usually well-marked leucocyfosis occurs, amount- 
ing on the average to a count of about twice the mean normal 
standard, but often greatly exceeding this figure in the individual 
case. The size of the primary abscess cannot be estimated by 
the height of the leucocytosis, but a tendency of the pus to extend 
is almost always accompanied by a distinct increase in the number 
of cells in excess of the figure originally estimated. Complete evac- 
uation of the abscess is soon followed by a disappearance of the 
leucocytosis and iodophilia, but so long as the pus remains inef- 
fectually drained, these signs persist. 

Analysis of the "first counts" in 258 cases of various forms of 
abscess (excluding appendicitis) shows that in 184, or 71.3 per 
cent., the leucocyte count was in excess of 10,000. The general 
range of the leucocytes in abscess is illustrated by the following 


40,000-50,000 2 

30,000-40,000 5 

20,000-30,000 28 

15,000-20,000 59 

10,000-15,000 90 

5,OOO-IO,OOO 70 

Below 5000 4 

Average, 13,931 per 
Highest, 42,000 " 
Lowest, 550 " 

A more detailed study of these cases, directed toward the 
number of the leucocytes in relation to the site of the abscess, may 
be expressed thus in tabular form: 























107 or 



per cent. 






26 " 






24,000 4,800 19 " 






31,800 11,200 10 " 



Lung. . 



17,6^0 8,200 6 " 







23,400 0,300" " 6 " 



Gall-bladder ... 6 


21/200 9,500 5 ' 







18,560 6,800 

5 ' 

8 3 



A polynuclear neutrophile gain generally accounts for the 
increase when leucocytosis is present, and, rarely, this differential 
change .may be found without any increase in the total number of 
cells. In some instances, and these are not so uncommon as 
is generally believed, the increase affects all forms of cells pro- 
portionately. A few myelocytes are not uncommon in cases having 
a decided anemia or a high leucocytosis. 

The presence of leucocytosis, especially if 

DIAGNOSIS, associated with hyperinosis and a positive iodin 

reaction, is suggestive of abscess rather than of 

other lesions, such as aneurisms, gummata, hematomata, and 

benign neoplasms. An absence of one or all of these signs, on the 

other hand, is not sufficient to exclude pus. The distinctions, as 


shown by the blood, between pyogenic and tuberculous abscesses 
and malignant disease are considered under the last-named con- 


The following counts illustrate the blood changes found in two 
cases of this disease, the first showing practically normal blood, 
except for a moderate relative lymphocytosis and an absence of 
eosinophiles, and the second simply a well-marked secondary 
anemia with a high color index. 


Hemoglobin 86 per cent. 60 per cent. 

Color index 0.93 i .04 

Erythrocytes 4,620,000 per 2,880,000 per 

Leucocytes 8,000 per 4,890 per 

Small lymphocytes 31.7 per cent. 21.0 per cent. 

Large lymphocytes 2.1 " 7.0 " 

Polynuclear neutrophiles 66.2 " 71.0 " 

Eosinophiles o.o " i .o " 

Basophiles o.o " o.o " 

Myelocytes o.o " o.o " 

The erythrocytes showed moderate deformities of size and 
shape in the anemic case, but no signs of nucleation nor of baso- 
philic stroma degeneration were observed. Coagulation, fibrin 
formation, and the number of plaques were apparently normal. 


Anemia, marked by a disproportionately great hemoglobin de- 
crease, is sometimes found, and leucocytosis appears to be an 
almost constant change, judging from the small number of reports 
available. In a case of actinomycosis of the arm the writer found 
55 per cent, of hemoglobin, with 4,985,000 erythrocytes and 12,000 
leucocytes per The number of blood plaques was greatly 
increased and the erythrocytes were pale, but not deformed, 
nucleated, nor basophilic. The percentages of the leucocytes 
were: small lymphocytes, 25.5; large lymphocytes, 7.3; polynuclear 
neutrophiles, 60.0; eosinophiles, 2.4; myelocytes, 3.2; basophiles 
(finely granular), i.o; and mast cells, 0.6. 


Erving 1 reports 4 cases with leucocyte counts of 10,000, 12,000, 
21,000, and 36,200; Ewing's case 2 gave 21,500; and Cabot's two 2 
showed leucocytoses of from 20,000 to 31,700. As a rule, actin- 
omycosis excites a higher leucocytosis when deep organs (liver, 
lungs) are involved than when the lesion is situated in superficial 
parts of the body (jaw, elbow, abdominal wall) where free drainage 
is favored. It is probable that the grade of both the anemia and 
the leucocytosis depends largely upon the amount of septic ab- 
sorption originating from the lesion. 


Malignant jaundice appears to be associated with a moderate 
polycythemia, so far as can be determined by the limited number 
of blood counts made in this disease up to the present time. The 
leucocytes are moderately increased in number, but show no 
peculiar differential changes. In two cases, reported by Grawitz 3 
and by Cabot, 4 respectively, the counts of erythrocytes were 
5,150,000 and 5,520,000, and the number of leucocytes 12,000 
and 16,000 per Bacteriological examination of the blood 
has thrown no definite light upon the nature of this apparently 
infectious process. In many cases hemoglobinemia and lipaci- 
demia have been detected. 


Moderate anemia is commonly, and decided 
HEMOGLOBIN anemia occasionally, associated with this condi- 
AND tion, although the prime importance of this 

ERYTHROCYTES. symptom insisted upon by Addison himself ap 
pears to be somewhat exaggerated, in the light 
of our .more accurate methods of blood study. The "anemiated 
eye" of Addison does not always mean anemia. In advanced 
cases the blood picture may be characterized by marked hemo- 
globin and erythrocyte losses, by the presence of numerous poikil- 
ocytes and microcytes, and by small numbers of normoblasts; 
the hemoglobin readings in such instances range between 20 and 
40 per cent., and the erythrocyte counts between 2,000,000 and 
3,000,000 per, or even less. Tschirkoff 5 reports cases in 
which, notwithstanding the coexistence of a notable oligocythe- 

1 Johns Hopkins Hosp. Bull., 1962, vol. xiii, p. 261. 

2 Loc. cit. ' Loc. cit. 

4 Loc. cit. s Zcitschr. f. klin. Med., 1890, vol. xix, p. 87. 


mia, the hemoglobin percentage remained normal or above nor- 
mal, and this peculiar condition he referred to an increase in the 
amount of corpuscular reduced hemoglobin. This author also 
detected the presence of methemoglobin and melanin in the blood 
of patients suffering from Addison's disease. The polycythemia 
which is sometimes .met with in this condition is doubtless to 
be attributed to such factors as vasomotor changes and blood 
inspissation from emesis. Treatment with suprarenal extract 
tends to. improve the anemia, but to what extent and how per- 
manently is undetermined. 

The number of leucocytes is usually normal 
LEUCOCYTES, or below normal, and extreme leucopenia has 
been repeatedly noted. Relative lymphocytosis 
and sometimes a moderate increase in the number of eosinophiles 
are the most familiar differential changes, together, in some in- 
stances, with the presence of a few myelocytes and basophiles. 


'Nothing definite is known of the behavior of the hemoglobin 
and corpuscles in this infection. Only occasionally can the anthrax 
bacillus be isolated from the peripheral blood, since general inva- 
sion of the circulation by this organism is rare. Blumer and 
Young 1 succeeded in finding the organism in the blood of a single 
case of anthrax septicemia, both in ordinary cover-glass speci- 
mens as well as by culturing. 


Fully three-fourths of all cases of appen- 
HEMOGLOBIN dicitis, whatever their character, show a loss of 
AND at least 30 per cent, of hemoglobin, while in 

ERYTHROCYTES. about one case in five the erythrocytes are di- 
minished 1,000,000 or more to the From 
an analysis, of the cases tabulated below it appears that the average 
hemoglobin loss amounts to about 25 per cent., and the average 
decrease in erythrocytes to about 15 per cent, of the normal 
standard. The. anemia, which may usually be attributed to the 
effects of septicemia, is most frequent and most marked in long- 
standing cases of appendicular abscess, in which type of the dis- 
ease the hemoglobin may fall to between 30 and 40 per cent., 
and the corpuscles to between 2,000,000 and 3,000,000 per 

1 Johns Hopkins Hosp. Bull. 1895, vol. vi, p. 127. 


3 6 7 

In such instances the risk, actual or reputed, of operating upon a 
patient having so low a percentage of hemoglobin must be re- 
called by the surgeon. (See p. 164.) Anemia, usually of a more 
moderate grade, is also frequently found in catarrhal cases, and 
in the individual instance it may reach as high a grade as in the 
purulent form of the disease. The blood impoverishment in such 
instances depends probably upon the debilitated state of the pa- 
tient, apart from the appendix inflammation. 

The following table illustrates the range of the hemoglobin and 
erythrocytes, as shown by the initial examinations of 139 cases l 
in Dr. J. B. Deaver's wards at the German Hospital: 


Purulent, Perfora- 
tive, and Gan- 

HEMOGLOBIN PERCENTAGE. Non-purulent. grenous. 

Above 100 i o 

From 90-100 i 4 

" 80-90 9 20 

" 7o-8o 13 31 

60-70 13 23 

" 50-60 6 7 

" 40-50 2 6 

" 30-40 o .3 

Highest 102 per cent. 100 per cent. 

Lowest 45 38 

Average -69.5 " 72.^ " 


Above 5,000,000 .- 6 14 

From 4,000,000-5,000,000 27 60 

" 3,000,000-4,000,000 ii 15 

" 2,000,000-3,000,000 i . 5 

Highestt 5,660,000 per 5,710,000 per 

Lowest 2,050,000 per 2,100,000 per 

Average 4,295,955 per 4,381,234 per 

Qualitative changes in the erythrocytes are neither common 
nor important, occurring only in cases with decided anemia, and 
consisting simply in deformities of shape and of size. Erythro- 

1 Similar figures were found in the examination of more than 700 additional 
rases not here recorded. 


blasts apparently do not occur, although there is no reason why 
they should not, if the anemia happens to be of a type of sufficient 
severity to provoke marrow changes. 

In simple appendicular inflammation uncom- 
LEUCOCYTES. plicated by pus, gangrene, or peritonitis there is, 
as -a rule, little or no increase in the number of 
leucocytes, although in an exceptional case the leucocytosis is 
fairly well denned. Thus, of the 45 cases of this form of the disease 
below referred to, less than 9 per cent, were accompanied by a 
count in excess of 15,000, the maximum estimate being 17,100, and 
the average 8987 per A relatively high count in this 
variety of appendicitis may usually be attributed to a limited 
periappendicular peritonitis. In some instances it is possible that 
the increase may be due to blood inspissation from vomiting and 
purging, or that it may simply represent a blood finding of the 
associated anemia. 

In cases with abscess, gangrene, or general peritonitis a well- 
marked leucocytosis is- the general rule. Few cases of appen- 
dicular abscess fail to increase the leucocyte count to at least 
15,000 or 20,000 to the, although it is to be remembered 
that should the purulent focus happen to be so effectually walled 
off that absorption of toxic material is practically prevented, so 
decided an increase does not develop. A trivial increase, or, 
indeed, an absence of leucocytosis, is also met with in an occa- 
sional grave case (such, for instance, as one complicated by a 
general purulent peritonitis), owing to the prostration of the 
patient from the systemic poison of the infection. As shown 
below, the average count in purulent and gangrenous appendicitis 
is higher than the maximum count in the* catarrhal form of the 

A high leucocytosis suggests either a localized abscess or a 
general peritonitis, for the differentiation of which other clinical 
data are absolutely essential. The belief is current that if a 
marked leucocytosis occurs early in the attack, peritonitis is prob- 
able, and if it occurs after the first week, a local accumulation of pus 
is suggested. While this is undoubtedly true in many instances, 
in many others the condition of the appendicular lesion may be 
wrongly interpreted if too great reliance is placed on the behavior 
of the leucocytes in connection with the period of the attack. 

Increase in the purulent focus and extension of peritonitis are 
betrayed by an increase in the leucocyte count, provided that the 
patient's powers of reaction are not too greatly crippled. In 
operative cases thorough evacuation of the abscess is followed 
within a few days by a decline to normal in the number of leuco- 


cytes. Persistence of the leucocytosis after the third or fourth 
day following the operation may usually be attributed to undrained 
pus pockets or to a general peritonitis. 

In non-operative cases with abscess the leucocytosis, which 
becomes well developed by the fourth or fifth day of the attack, 
persists but does not tend to increase if the lesion remains local- 
ized; it gradually decreases as the pus collection disappears; 
and it suddenly increases if the process extends. 

To sum up, absence of or slight leucocytosis suggests (a) 
simple catarrhal appendicitis; (6) fulminant appendicitis; or (c) 
a localized pus focus from which no absorption occurs. Well- 
marked leucocytosis indicates (a) a local abscess from which 
absorption of toxins occurs; (b) general peritonitis; or (c) gangrene. 

The following table shows the range of the leucocytes in the 
German Hospital cases to which reference has been made: 


LEUCOCYTES PER C.MM. Non-purulent. Purulent, Perforative, and 


Above 50,000 o i 

From 40,00050,000 o o 

" 35,000-40,000 o 2 

" 30,00035,000 o o 

" 25,000-30,000 " o 6 

" 20,000-25,000 o 16 

" 15,000-20,000 .4 . ' 38 > 

" 10,00015,000 10 24 

5,000-10,000 25 7 

Below 5000 : 6 o 

Highest 1 7,100 per 58,500 per 

Lowest i ,600 " " 6,000 " " 

Average 8,987 " " . 17,955 " 

The qualitative changes found in high leucocyte counts are 
those typical of an ordinary polynuclear neutrophile leucocytosis 
a large absolute and relative gain in polynuclear forms at the 
expense of the hyaline mononuclear cells. 

Findings essentially like the above have been reported by 
Cabot, 1 Richardson, 2 Bloodgood, 3 and Joy and Wright, 4 in America; 
by Longridge, 5 Gulland, 8 and French, 7 in Great Britain; by 

1 Loc. cit. * Amer. Jour. Med. Sci., 1899, vol. cxviii, p. 635. 

3 Prog. Med., 1901, vol. iv, p. 216. 4 Med. News, 1902, vol. Ixxx, p. 628. 

5 Brit. Med. Jour., 1902, vol. ii, p. 1511; also Lancet, 1902, vol. ii, p. 74. 
* Scottish Med. and Surg. Jour., 1903, vol. xii, p. 157. 
7 Brit. Med. Jour., 1904, vol. i, p. 1136. 


Curschmann, 1 Wassermann, 2 Dutzmann, 3 Federmann, 4 Gern- 
gross, 5 Coste, 6 Kuhn, 7 and Stadler, 8 in Germany; and by Cazin 
and Gros, 9 in France. 

The detection of iodophilia is a great diagnostic aid. Like 
leucocytosis, it measures the intensity of the toxemia, but it also 
develops in greatly anemic patients, irrespective of their systemic 
reaction to an appendicular lesion. Iodophilia is most marked 
in cases with abscess, gangrene, and local or general peritonitis ; 
in pus cases the reaction persists so long as the pus remains 
pent up, but it rapidly disappears after adequate drainage is 
established usually, in the author's cases, within thirty-six hours 
or so. Thus it appears that the iodin reaction has much the 
same meaning as leucocytosis, but it is even a more sensitive sign 
than the latter, since it frequently betrays a sepsis too slight to 
excite, or so grave as to stifle, a leucocyte increase. Locke, 10 
from a study of 61 cases of appendicitis, lays stress on iodophilia 
as a negative sign; rarely, if ever, does it fail to develop when 
sepsis is present. 

The conditions which may more or less 
DIAGNOSIS, closely simulate an acute attack of appendicitis 
are numerous, and unfortunately it happens that 
just those lesions in which the resemblance is closest often produce 
blood changes identical with those of appendicitis. Thus, leuco- 
cytosis is the rule in pyosalpinx, ovarian abscess, ectopic pregnancy, 
Pyonephrosis, perinephritic abscess, hepatic abscess, empyema of the 
gall-bladder, mesenteric thrombosis, and malignant disease 0} the 
cecum, all of which may be confused with an appendicular abscess. 
Iodophilia also occurs in these conditions, save perhaps in ectopic 

Such a large proportion of cases of hepatic and renal colic are 
accompanied with acute inflammatory complications, giving rise 
to leucocytosis, that these conditions cannot be differentiated 
with any degree of confidence from appendicitis simply by an 
examination of the blood. The same is true of dysmenorrhea, 
in which inflammatory changes in the uterus may constitute the 
factor of a leucocyte increase. Acute gastritis is sometimes ac- 
companied by a well-marked leucocytosis, and sometimes by 

1 Miinch.-med. Wochenschr., 1901, vol. xlviii, pp. 1907 and 1962. 

2 Arch. f. klin. Chir., 1903, vol. Ixix, p. 392. 

3 Sem. med., 1903, vol. xxiii, p. 324. 

4 XXII. Cong. f. Chir., Berlin, 1903; also Sem. med., 1904, vol. xxiv, p. 206. 

5 Munch, med. Wochenschr., 1903, vol. 1, p. 1586. 

8 Ibid., 1902, vol. xlix, p. 2038. 7 Ibid., 1902, vol. xlix, pp. 2033 and 2085. 

8 Jour. Amer. Med. Assoc., 1903, vol. xli, p. 216. 

9 Sem. med., 1903, vol. xxiii, p. 141. 

10 Boston Med. and Surg. Jour., 1902, vol. cxlvii, p. 289. 


none at all, so that the blood count cannot be relied upon as a 
clue in distinguishing this disease from appendicitis. 

Simple enteralgia, gastralgia, and ovarian neuralgia may be 
ruled out if a leucocyte increase is present, as also may be intestinal 
obstruction, provided that the latter is not 'complicated by inflam- 
matory changes, by gangrene, or by malignant disease. In 
lead colic there is often a pronounced leucocytosis, especially in 
patients with acutely toxic symptoms; but granular basophilia of 
the erythrocytes can be detected even in the earliest stages of 
plumbism, while in appendicitis this change is found only in 
highly anemic subjects. 

The presence of a leucocytosis is sufficient to exclude a non- 
inflammatory ovarian cyst and a movable kidney, and the same 
sign is of no little value in ruling out enteric fever if no leuco- 
cyte-raising complications are apparent. 

The simple fact of the presence or absence of a leucocytosis is 
more often misleading than useful in the diagnosis of appendicitis, 
for this sign, to be of any real value, must invariably be corre- 
lated with other more clinical manifestations. Appendicitis 
should never be ruled out because leucocytosis is absent, nor 
should a moderate leucocyte count be considered an indication 
of the benignancy of the lesion. A count in excess of 20,000, 
particularly if it persists or increases, may be relied upon 'as a 
certain sign of pus or its consequences, and is sufficient to 
warrant operative interference if the symptoms point to the 
appendix as the seat of the trouble. Counts of less than 
20,000 cannot be depended upon to reflect the character of the 
local lesion, since an increase to practically this figure may be 
found occasionally in mild catarrhal cases, as well as in those with 
purulent foci. In the writer's experience the behavior of the 
leucocytes throws a much clearer light upon the progress of the 
disease in both operative and non-operative- cases than it does 
upon the. initial diagnosis, which should be determined chiefly by 
other clinical methods. 

The absence of iodophilia is a dependable sign that no active 
inflammation of the appendix exists, although its presence does 
not necessarily mean appendicitis, as remarked above. 


In spite of their pallor, sufferers from arthritis deformans 
seldom have decided anemia. Recent hematological work in this 
disease shows that most cases have normal hemoglobin and erythro- 


cyte values, and that anemia, when it does develop, is generally 
moderate and traceable to causes other than the joint lesions. 
McCrae's report 1 shows an average hemoglobin percentage of 70.6 
in 33 cases, and an average erythrocyte count of 4,468,000 per in 29. Erving's counts 2 in 40 cases show an average 
hemoglobin percentage of 94, with a range between 80 and 100, 
and an erythrocyte count averaging 5,112,000, with a range 
between 4,148,000 and 5,980,000. These figures evidence a much 
less marked anemia than that detailed by earlier writers, for ex- 
ample by Bannatyne, 3 who found that the hemoglobin commonly 
ranged between 40 and 80 per cent., and the erythrocytes between 
3,000,000 and 4,000,000. Histological changes in the erythrocytes 
are conspicuous by their absence. 

Leucocytosis rarely accompanies arthritis deformans, and when 
present, can generally be referred to some complicating lesion. 
Of McCrae's 33 cases, but 9 showed definite leucocytosis, the 
average count being 7600; in Erving's 40 cases the leucocytes 
exceeded 10,000 per in but 5 instances, and averaged 8885. 
Differential changes are trifling and do not occur in all cases. 
They consist in nothing more than a moderate diminution in the 
polynuclear neutrophiles with a proportionate lymphocyte in- 


In many cases there is great difficulty in 

GENERAL obtaining a sufficient quantity of blood for clin- 

FEATURES. ical examination, even from a deep puncture. 

This peculiarity, which has been attributed to 

excessive dryness of the tissues from drains upon the body fluids, 

is most pronounced in the algid stage of the disease. 

The great decrease in the alkalinity of the blood in Asiatic 
cholera, sometimes spoken of as an acid reaction, was first de- 
termined by C. A. Schmidt 4 by a series of elaborate analyses 
made in 1850, since which time similar findings have been noted 
by Cantani, 5 Straus, 8 and others. 

The density of the blood mass is found to be increased, es- 
pecially in those cases in which the blood is highly inspissated; 
in such instances the specific gravity may rise to as high as 1.073. 
The agglutination of cholera vibrios by the blood serum of 

1 Jour. Amer. Med. Assoc., 1904, vol. xlii, p. i. 

2 Amer. Med., 1903, vol. vi, p. 440. 3 Lancet, 1896, vol. ii, p. 1510. 
4 " Charakteristik der epidemischen Cholera gegeniiber venvandten Transsuda- 

tionsanomilien," Leipsic, 1850. 

* Centralbl. f. d. med. Wissenschaft, 1894, vol. xxii, p. 785. 

* Compt. rend. Soc. biol., Paris. 1883, vol. iv, p. 569. 


cholera patients was first applied as a clinical test by Achard 
and Bensaude, 1 these investigators finding that the reaction may 
occur as early as the reputed first day of the illness and as late 
as the fourth week after recovery. Clinically, the test may be 
made either with dried blood or with serum. 

The studies of Biernacki 2 and of Okladnych, 3 

HEMOGLOBIN which together include the investigation of 62 
AND cases, furnish the most complete data con- 

ERYTHROCYTES. cerning the changes affecting these elements. 
Both of these observers found a more or less 
marked polycythemia with a proportionate increase in the hemo- 
globin percentage, the erythrocyte count in many cases being 
between 6,500,000 and 7,500,000, and in one case reaching a 
maximum of 8,000,000. The increase may often be observed 
within a few hours after the onset of the infection. Concentra- 
tion of the blood is to be considered as the cause of these high 
counts, which are, as a rule, highest in cases characterized by 
pronounced emesis and purging. No constant relation between 
the degree of polycythemia and the gravity of the infection can 
be distinguished. 

The above-quoted authors found high-grade 

LEUCOCYTES, leucocytosis to be the almost invariable rule, 
the increase in leucocytes being not parallel 
with, but rather relatively greater than, the accompanying in- 
crease in erythrocytes. It occurs both in mild and in severe 
cases, as early as within twelve hours after the onset of the dis- 
ease, and as late as the third, fourth, or sixth day. It may be 
present both in the algid stage and in the stage of reaction, but is 
likely to be more decided in the former. The degree of leucocy- 
tosis may range from a minimum count of 14,000 to a maximum 
of 60,000 cells per, the case of average severity showing 
an increase to about 25,000 or 30,000. Preagonal leucocytosis 
may be pronounced, counts of 50,000 being not uncommon. 
Biernacki 'states that "all cases which in the algid stage show 
a leucocytosis of 40,000 to 60,000 soon prove fatal." On the 
contrary, an absence of leucocytosis cannot be regarded as a 
surety that the patient will recover. In a trivial infection distinct 
leucopenia has been observed, but this is rare. Rogers, 4 who in- 
variably found leucocytosis in 23 cases, also believes that the 
higher the count, the worse the prognosis. Of 9 of his patients 
with counts of less than 20,000, 4 died; of 14 with counts ranging 

1 Presse med., 1896, vol. xvi, p. 504. 

2 Deutsch. med. Wochenschr., 1895, vol. xxi, p. 795. 

5 Cited by Biernacki, loc. cit. * Lancet, 1902 vol. ii, p. 659. 


between 20,000 and 46,000, n died. Concentration of the blood 
does not altogether account for the leucocytosis of cholera, for 
the influence of the specific infection as a factor is thought to be 
most active. 

The leucocytosis involves chiefly the polynuclear neutrophiles, 
but not so conspicuously as in most other infections, since the per- 
centage of these cells in cholera seldom exceeds 80. Rogers first 
described this peculiarity, and also drew attention to the behavior 
of the lymphocytes. The small lymphocytes, he found, are 
usually outnumbered to the extent of two to one by the large 
mononuclear forms, the latter's increase becoming more marked 
as the disease progresses, and being especially so in fatal cases. 
He finds that an excessive increase in this form of cell is of bad 
prognosis: of 18 cases with counts exceeding 2000 large lym- 
phocytes per.cjnm., 14 died; while of 5 in which these cells num- 
bered less than 2000 but a single patient died. The eosinophiles 
rarely numbered more than a fraction of one per cent. Sherring- 
ton l has found the mast cells notably increased in some instances. 
Leucocytosis is more constant and tends to 
DIAGNOSIS, reach, a higher degree in Asiatic cholera than in 
any other non-choleraic disease with similar 
symptoms. In acute dysentery and in ptomain poisoning, however, 
the counts may be identical with those of cholera, but the character- 
istic lymphocyte formula of the latter is wanting. From the 
viewpoint of prognosis this differential change and the height of 
the total leucocyte count are obviously helpful signs. 


In long-standing cases moderate secondary anemia involving 
chiefly a hemoglobin loss may be found, for in many instances 
the general debility of the patient or the presence of lesions of 
other organs is quite adequate to give rise to such a change. 
In cyanotic patients the anemia may be hidden by the polycy- 
themia arising from circulatory disturbances, this deceptive blood 
concentration being most conspicuous during an asthmatic par- 

Little or no increase above the normal standard in the number 
of leucocytes is the usual condition, although these cells may show 
a considerable increase in cases associated with acute bronchitis, 
and also during an asthmatic attack. Gabritschew r sky, 2 Fink, 3 von 

1 Proc. of the Roy. Soc., London, 1894, vol. Iv, p. 189. 

2 Arch. f. exp. Path. u. Pharm., 1890, vol. xxviii, p. 83 

3 Inaug. Diss., Bonn, 1890. 


Noorden, 1 Billings, 2 and others have called attention to the pres- 
ence of an eosinophile increase in both asthma and emphysema. 
From 10 to 20 per cent, of this type of cells is not an unusual 
proportion, both in cases with and in those without leucocytosis, 
while in one case Billings has reported three consecutive counts 
of 33.9, 38.2, and 53.6 per cent, respectively, with correspond- 
ing total leucocyte estimates of 7600, 7500, and 8300 per 
In true bronchial asthma the eosinophile increase develops shortly 
before the paroxysm, and persists during and for a short time after 
it, disappearing in the interval between the seizures. This sign 
is regarded of value in differentiating true bronchial asthma from 
the dyspnea due to renal and cardiac lesions, since in the latter 
the eosinophiles are never increased, and it is also considered 
of some clinical utility in heralding an impending asthmatic 


With the exception of a slight oligochromemia, which is fre- 
quently present in severe cases with high temperatures, the 
erythrocytes and their hemoglobin content remain practically 
normal in all forms of- bronchial inflammation. 

A cute catarrhal bronchitis of the larger tubes is ordinarily un- 
attejided by leucocytosis, but, unfortunately for diagnostic pur- 
poses, an occasional case shows a marked increase. Thus, in four 
of Cabot's seventeen cases 3 the counts were 17,600, 23,500, 26,000, 
and 41,000 respectively, while in eleven the leucocytes numbered 
more than 10,000 per In chronic bronchitis leucocytosis 
rarely if ever occurs. . Extension of the inflammation to the finer 
tubules and vesicular structure causes a leucocytosis identical 
with that of crpupous pneumonia (q. v.). 

Inflammation of the tracheobronchial glands, according to 
Carriere, 4 sets up a decided mononucleosis. Lichtwitz and 
Sabrazes t found this change in the blood of children with naso- 
pharyngeal adenoids, the mononuclear increase in such cases dis- 
appearing after removal of the growths. 

1 Zeitschr. f. klin. Med., 1892, vol. xx, p. 98. 

2 N. Y. Med. Jour., 1897, vol. Ixv, p. 691. 

3 Loc. cit. 4 Sem. med., 1902, vol. xxii, p. 44. 
5 Arch, de me*d. des Enf., 1901, vol. iv, p. 120. 



Slow, imperfect coagulation of the blood has 
GENERAL been found, and, in virulent cases, absolute non- 
FEATURES. coagulability. This was noted by Alice M. 
Corthorn * in 10 of 12 fatal cases of pest, whose 
blood, kept sealed in Wright's tubes for three weeks and longer, 
showed no sign of clot formation, but resembled simply a dark- 
red, treacle-like fluid. Jennings 2 found that rouleaux formation 
is but feebly exhibited. 

Since 1894, when Kitasato and Yersin, working independently, 
simultaneously discovered the Bacillus pestis bubonicce in the 
circulating blood of patients infected with plague, this organism 
has been repeatedly isolated from the blood by many different ob- 
servers. In a careful bacteriological study of 27 cases Ogata 
also frequently found in the blood, especially in severe infections, 
a micro-organism morphologically similar to FrankePs pneu- 
mococcus, the significance of this unidentified organism being 
undetermined. The same observer calls attention to the fact that 
blood from patients convalescent from nineteen to sixty-five days, 
although giving negative results by cultural methods, when in- 
jected into mice proves rapidly fatal to these animals, in whose 
tissues the plague bacillus may be recovered in pure culture. The 
relatively large number of positive results to be obtained from 
bacteriological blood examinations in this disease, especially in its 
septicemic form, attaches to the procedure no small diagnostic 
value. The German Plague Commission's results 4 43 positive 
findings in a series of 124 cases examined are probably represen- 
tative of the value of blood culturing in this infection. Powell's 
studies 5 show but 15 positive cultures in 117 cases. Cultural 
methods with blood drawn directly from a vein give, of course, 
the most favorable results, but the bacilli may be often detected 
in the stained cover-glass specimen of finger blood, in which they 
appear as short rods, tending to group together in chains or in 
pairs, exhibiting bipolar staining and decolorizing by Gram's 
method. In view of the fact that the peripheral blood contains 
but small numbers of the bacilli, Rees 6 advises making large 
films on -slides rather than cover-glass specimens, should direct 
examination of the stained film be attempted. 

1 Brit. Med. Jour., 1902, vol. i, p. 1143. 2 "Manual of Plague," London, 1903. 

3 Centralbl. f. Bakt. u. Parasit., 1897, vol. xxi, p. 769. 

4 Cited by Now, Amer. Jour. Med. Sci., 1901, vol. cxxii, p. 416. 

5 Indian Med. Gaz., 1904, vol. xxxix, p. 41. 
8 Brit. Med Jour., 1900, vol. ii, p. 1236. 


The agglutination of the plague bacillus by the blood serum 
from plague subjects has been noted by a number of different 
investigators, but thus far no clinical application of the reaction 
has been made. The inconstancy with which the reaction occurs 
for it may frequently be absent in both the mildest and the most 
severe cases and the variable degrees of serum dilution neces- 
sary for its production appear to bar the acceptance of the test as 
a reliable diagnostic sign. The British Indian Plague Commis- 
sion 1 concludes that "no practical value attaches to the method of 
serum diagnosis in the case of plague." 

The striking bactericidal action of plague serum upon the Bacil- 
lus pestis has been taken by Row 2 as the basis of a clinical test. A 
drop of blood serum from a plague patient is mixed with a loopful 
of a saline emulsion of the Bacillus pestis, and from this mixture 
a hanging-drop culture is made and placed in the dark for twenty- 
four hours at laboratory temperature. The cover-glass is then 
fixed, stained with thionin, and examined microscopically. If the 
serum used in the test is from a person infected with plague, the 
tw T enty-four-hour-old culture shows either no growth of the bacillus 
or, at the most, a very few distorted and atypical organisms. In 
control cases, witfr normal serum, the growth is abundant. Row 
considers this test of real value in estimating the protective effects 
of Haffkinization. Its usefulness as a clinical means of diagnosing 
plague is restricted by the time required (twenty-four hours") for 
the completion of the reaction, though in suspected cases with 
poorly developed symptoms the test may prove of value. 

According to Aoyoma's- studies, 3 the hemo- 
HEMOGLOBIN globin and the erythrocytes are both decidedly 
AND increased above normal in the majority of cases. 

ERYTHROCYTES. Of the six cases examined by this writer, five 
showed marked polycythemia, the highest count 
being 8,190,000, and the average 6,976,666. Septic cases may 
develop secondary anemia of variable intensity. Qualitative 
changes hi the erythrocytes, it is to be presumed, are not. con- 
spicuous, since no mention of such alterations is made. 

In two-thirds of the cases just quoted marked 

LEUCOCYTES, increase in the number of leucocytes was found, 

the gain being greater than is ordinarily met with 

in any condition except leukemia; the count exceeded 100,000 

in four instances, and averaged for the six 96,666. In fulminant 

1 Brit. Med. Jour., 1903, vol. i, pp. 1093, 1155, 1220, 1279. 

2 Ibid., 1902, vol. ii, p. 1895. 

3 Mittheilungen aus d. Mod. Fac. d. Kaiserlich-Japanischen Univcrsitat, Tokio, 
1895, vol. iii, p. 115. 


types of plague extreme leucopenia, due to the overpowering 
toxemia, is not unusual, and in very mild cases the number of 
leucocytes may be normal. The increase was due usually to a 
disproportionately large percentage of poly nuclear neutrophiles, 
but in some cases " large and small mononuclear white cells" 
were observed. -The eosinophiles were, as a rule, conspicuous 
by their absence. Rogers, 1 whose leucocyte counts varied from 
20,000 to 60,000, also observed a lymphocytosis, and emphasizes 
the fact that in most cases the polynuclear neutrophiles show little 
or no relative increase. Contrary to Aoyoma's findings, Zinno 2 
has noted, in an occasional case, a great abundance of eosinophiles 
of the myelocytic type, of the ordinary polynuclear variety, and of a 
form transitional between the two; in one of his cases the eosino- 
philic myelocytes numbered 13 per cent, of the total leucocyte 

The blood plaques are in most cases notably increased in 


Rapidly developing and marked polycythemia, equally striking 
leucocytosis, and a plaque increase are the features of clinical 
interest in the blood picture of the severely burned. To the naked 
eye the blood has a deep purplish color and flows very sluggishly. 

Locke's studies 3 of ten cases of severe burns show that within 
a few hours after the accident a marked increase in the number 
of erythrocytes takes place, amounting in favorable cases to be- 
tween 1,000,000 and 2,000,000 cells to the, and in fatal cases 
to between 2,000,000 and 4,000,000. This polycythemia may be 
explained chiefly by venous stasis and partly by loss of the blood 
plasma. Structural changes in the erythrocytes are not con- 
spicuous. The leucocytes rapidly increase, reaching a count of 
from 30,000 to 40,000 per in non-fatal burns, and of 50,000 
or more in patients who die. The percentage of polynuclear 
neutrophiles increases, although not to so high a figure as is 
commonly found in ordinary inflammatory leucocytosis. De- 
generative changes, especially of their protoplasm, involve many 
of these cells, as well as the other granular leucocytes, and are 
marked in parallelism to the severity of the lesion. Myelocytes 
occur in small numbers, particularly in cases having a high 
leucocytosis. The blood plaques are greatly increased in practi- 
cally every instance. 

1 Lancet, 1902, vol. ii, p. 660. 

2 Centralbl. f. allg. Path. u. path. Anat., 1902, vol. xiii, p. 410. 

3 Boston Med. and Surg. Jour., 1902, vol. cxlvii, p. 480. 



The blood changes in this rare disease closely resemble those 
of lymphatic leukemia, namely, marked anemia with absolute 
lymphocytosis. The hemoglobin and erythrocytes progressively 
diminish as the disease runs its course, and eventually may sink 
quite as low as in pernicious anemia. In a case reported by 
Dunlop * the hemoglobin fell from 32 to 12 per cent., and the ery- 
throcytes from 1,800,000 to 850,000 per during a period of 
but four weeks. Estimates not dissimilar from these have been 
recorded by a number of other writers, notably by Rosenblath 
and Risel, 2 by Gumbel, 3 and by Weinberger. 4 Misshapen and 
otherwise degenerate erythrocytes, together with a variable number 
of normoblasts, make their appearance as the anemia increases 
in degree. 

High leucocyte counts, involving absolute lymphocytosis 
(small celled), are the rule in the cases thus far reported. Ex- 
ceptionally there is simply lymphocytosis of the relative form, as in 
a case reported by B ram well 5 with 95 per cent, of lymphocytes and 
a leucocyte count of 8000. Early in the disease the total count is 
not excessive, 20,000 or 30,000, but after a few weeks it may 
attain much higher figures 75,000 to 100,000 or more. Dunlop's 
case at one time showed a leucocyte count of 245,000 per c,rnm. 
Differentially, the count in chloroma differs but slightly from that 
of lymphatic leukemia. The percentage of lymphocytes is rarely 
so high as in the latter, and mast cells are not so numerous, but 
myelocytes are more abundant. Eosinophiles, both pofynuclear 
and myelocytic, were numerous in a case studied by Dock. 8 These, 
however, are minor differences, and do not serve as reliable 
criteria of differentiation. 

Chloroma has a superficial resemblance to exophthalmic 
goiter, and very closely counterfeits acute lymphatic leukemia. 
Indeed, is strongly tempted to agree with Dock in considering 
it a malignant form of leukemia, from which it differs chiefly in 
being more violently neoplastic and in forming greenish infiltrations 
and metastases. Graves' disease is readily differentiated by the 
blood picture. The clinical differences between chloroma and 
leukemia are dealt with under the latter disease. (See p. 326.) 

1 Brit. Med. Jour., 1902, vol. i, p. 1072. 

2 Deutsch. Arch. f. klin. Med., 1902, vol. Ixxii, p. i. 

3 Virchow's Arch., 1903, vol. clxxi, p. 504. 

4 Zeitschr. f. klin. Med., 1903, vol. 1, p. 383. 

5 Lancet, 1902, vol. i, pp. 451 and 520, * Med. News, 

1904, vol. Ixxxiv, p. 955. 



In impacted calculi with jaundice, coagulation 

GENERAL is frequently but not invariably delayed, but in 

FEATURES, gall-stone complicated by phlegmonous cholan- 

gitis or other purulent sequelae, hyperinosis is 

observed, and coagulation is generally more rapid than normal. 

In 28 cases of cholelithiasis the writer found that clotting occurred 

in less, than five minutes in 7, in from five to ten minutes in 16, 

and in longer than ten minutes in 5. In 60 per cent, clotting was 

delayed, the coagulation time for these cases averaging eight and 

one-half minutes. The general effects of bile upon the blood, 

elsewhere noted, may also be detected when marked jaundice 

develops. (See "Cholemia" and "Icterus.") 

Positive results from bacteriological examination of the blood 
have frequently been obtained in cholelithiasis, streptococci having 
been isolated by Netter, 1 staphylococci and colon bacilli by Sitt- 
mann, 2 streptococci and pneumococci by Canon, 3 and various 
bacteria of unknown identity by other investigators. 



From 90-100 12 Above 5,000,000 13 

23 From 4,000,000-5,000,000.58 

4 3 " 3,000,000-4,000,000.34 

" 50-60" ' 8 " 2,000,000-3,000,000.10 

40-50. " " 1,000,000-2,000,000 . i 

" 30-40 i 

" 20-3 3 

Below 20 i 

Average, 73.5 per cent. Average, 4,080,1 17 per 
Maximum, 98.0 Maximum, 5,390,000 " " 

Minimum, 15.0 Minimum, 1,040,000 " " 

Moderate oligochromemia is found in the 

HEMOGLOBIN greater proportion of cases, but a decided loss 

AND either of hemoglobin or of erythrocytes is com- 

ERYTHROCYTES. paratively rare. In general terms it may be 

conservatively stated that the hemoglobin loss 

on the average amounts to about 30 per cent., and that the cel- 

1 Progres med., 1886, vol. xiv, p. 992. 

2 Deutsch. Arch. f. klin. Med., 1894, vol. liii, p. 323. 

8 Deutsch. med. Wochenschr., 1893, vol. xix, p. 1038. 


lular decrease approximates 15 per cent, of the normal standard. 
In occasional instances, notably those in which suppuration or 
sepsis coexists, the anemia is of a more intense grade, and may 
be associated with various changes indicative of cellular degen- 
eration. In 116 cases of cholelithiasis the foregoing estimates of 
the hemoglobin and erythrocytes were obtained at the initial 

Simple gall-stone does not of itself excite the 

LEUCOCYTES, slightest increase in the number of leucocytes, 

but nevertheless leucocytosis, typically polynu- 

clear in type, is a rather common feature of the blood picture in 

this disease, owing to the fact that such a large percentage of 

cases is complicated by acute inflammatory changes. Thirty-three 

of the 116 cases just mentioned had a count of more than 10,000 

cells to the The following resume of the examinations 

illustrates the range of leucocytes in the series: 



From 20,000-30,000 6 

" 15,000-20,060 8 

" 10,000-15,000 ,. 19 

5,000-10,000 74 \ 

Below 5,000 9 

Average, 9,623 per c.rrim. 

Maximum, 26,000 " " 

Minimum, 4,500 ' " " 

The presence of a leucocytosis excludes simple 
DIAGNOSIS, biliary colic, and indicates as the cause of the in- 
crease some other lesion, such, for example, as 
phlegmonous cholangitis or cholecystitis, hepatic abscess, peritonitis, 
or malignant disease. Hepatic and renal colics cannot be differ- 
entiated by the blood count. 

The surgeon should remember that cases with delayed coagu- 
lation may bleed freely, even fatally, when the knife is used, and 
that to such patients remedies which promote clotting should be 
given before operating. 


In this new clinical entity, first described by Saunby and Rus- 
sell, 1 the hemoglobin and erythrocyte values attain extraordinarily 

1 Lancet, 1902, vol. i, p. 515. 


high figures. O^ler, 1 who has studied the condition in detail, 
reports one case with a hemoglobin percentage of 165 and an 
erythrocyte count of 10,200,000 per, while in a case exam- 
ined by J. N. Hall 2 the hemoglobin was 200 per cent, (making 
it necessary to dilute the blood doubly in order to make the test 
with the von Fleischl hemometer), and the erythrocyte count 
reached 9,949,600. Estimates not differing greatly from these 
have also been reported by Cabot, 3 Saunby and Russell, 4 C. G. 
Stockton, 5 Tiirk, 6 Vaquez and Quiserne, 7 and others. The 
blood is thick and tarry, flows sluggishly from the puncture, 
and coagulates with great rapidity. The cause of the enor- 
mous polycythemia is a moot point. Osier 8 proposes as a 
factor hyperviscosity of the blood, in consequence of which the 
intracapillary flow is impeded. Gibson 9 attributes it to peripheral 
stasis dependent upon myocardial weakness. 

Osier's disease, as it may be fittingly termed, occurs in middle 
age, and, aside from the blood findings, is characterized by idio- 
pathic and permanent cyanosis, by splenic and sometimes hepatic 
enlargement, and almost invariably by albuminuria. In the cases 
thus far studied no clinically demonstrable lesion of the heart or 
lungs has accounted for the cyanosis, nor has autopsy revealed 
tuberculosis of the spleen, which in its primary form is associated 
with cyanosis and polycythemia of a moderate degree. 

The leucocytes are usually not increased, although in an occa- 
sional instance they have been found to number about double the 
maximum normal standard. Differentially they are also normal. 


Graham, 10 of Beyrouth, recently announced that dengue is due 
to a protozoon resembling in some respects the malarial parasite. 
He believes that such an organism is constantly present in the dis- 
ease, and that it undergoes an evolution within the erythrocytes at 
their expense. Furthermore, it is held that the parasite is harbored 
and conveyed by a species of mosquito, the Culex fatigans. 
Graham's work still lacks confirmation, and hence cannot be 

1 Amer. Jour. Med. Sci., 1903, vol. cxxvi, p. 187. 

7 Amer. Med., 1903, vol. v, p. 1026. 

3 Boston Med. and Surg. Jour., 1899, vol. cxli, p. 574; ibid., 1900, vol. cxlii, 


4 Loc. cit. 5 Medical News, 1903, vol. Ix.xxii, p. 948. 
' Cited by Osier, loc. cit. 7 Sem. med., 1902, vol. xxii, p. 235. 

8 Brit. Med. Jour., 1904, vol. i, p. 121. 

9 Lancet, 1903, vol. ii, p. 1560. 10 Med. Rec., 1902, vol. Ixi, p. 204. 


considered conclusive. Study of the corpuscles in this disease 
seems to have been generally neglected. 


The alkalinity of the blood, according to the 

GENERAL investigations of Minkowski, 1 is appreciably 

FEATURES, diminished, especially in cases in which coma 

either impends or exists. The change, however, 

cannot be regarded as constant, since in none of the five cases 

lately studied by Golla 2 did the alkalinity figures differ materially 

from normal. 

Orlowsky 3 in two cases of diabetes mellitus appreciably in- 
creased the blood alkalinity by giving warm alkaline enemata, 
the change thus effected being more decided and persisting longer 
than when caused by the administration of similar drugs by the 
mouth. He believes in treating diabetic coma in this manner, 
the alkali being given for some time after urgent symptoms have 
disappeared. Lipemia is not uncommon, the amount of fat in 
some instances being so large as to produce a milky appearance 
of the blood drop, evident to the naked eye, although in most 
instances the condition is recognizable only by the detection of 
fat globules under the microscope.' Not infrequently diabetic 
blood has a peculiar salmon color. Neisser and Berlin 4 report a 
case of diabetes which showed 20 per cent. ( !) of fat ii\ the in- 
spissated blood obtained by venesection. Lipacidemia may 
be detected in diabetic coma. Glyccmia is present, and can be 
demonstrated by the detection of grape-sugar in relatively large 
amounts, even as great as 5.7 parts per thousand, according 
to Grawitz, 5 or 9 per thousand, according to Hoppe-Seyler. 6 (See 
p. 141.) 

Williamson's Test. This reaction, devised by Williamson 7 in 
1896, depends upon the fact that a warm alkaline solution of 
methylene-blue is decolorized when mixed with a minute quantity 
of glucose. Twenty of the suspected blood, obtained 
by puncturing the finger, are measured by means of Gower's 
hemocytometer pipette, and blown out into 40 of distilled 
water contained in a small test-tube. To this mixture are then 

1 Mitth. a. der med. Klinik. z. Konigsberg, 1888. 

2 Lancet, 1903, vol. i, p. 1230. 

3 Russkiy Vrach, 1901, vol. xxii, pp. 1190 and 1222. 

4 Zeitschr. f. klin. Med., 1904, vol. li, p. 428. 

5 Jj)c. cit. * Virchow's Arch., 1858, vol. xiii, p. 104. 
7 Brit. Med. Jour., 1896, vol. ii, p. 730; also Lancet, 1900, vol. ii, p. 320. 


added, in the order given, i c.c. of a i : 6000 aqueous solution of 
methylene-blue and 40 of a six per cent, aqueous solution of 
potassium hydrate. In a second test-tube the same proportions of 
normal blood and reagents are mixed, to be used as a control. 
The color of both mixtures is precisely the same moderately deep 
bluish-green. Both tubes are placed in a beaker filled with boil- 
ing water, in which they are allowed to remain for four minutes, 
at the end of which time the test fluid containing the diabetic 
blood 'will have turned a dingy yellow color, while the color of 
the control mixture remains unchanged. Care must be taken to 
use not more than 20 of blood, since a positive reaction 
may be more or less closely counterfeited with non-diabetic blood 
should three or four times this quantity be employed. It is 
essential, therefore, to measure the blood accurately, and not to 
trust to the approximate method used by some, of simply taking 
two drops of blood as the equivalent of the required 20 

Williamson's reaction is presumably due solely to the action of 
the grape-sugar contained in diabetic blood, and if this proves 
true, it is not unreasonable to predict that the principle of the test 
may be elaborated into a method for estimating the percentage 
of sugar in the blood. Positive reactions occur constantly in 
diabetes, sometimes even after the disappearance of every trace 
of sugar from the urine, and, so far as investigations up to the 
present time have shown, negative results are invariably met with 
in other diseases. 

Brewer's Test, Bremer, 1 having noticed in diabetes mellitus 
peculiar affinities of the erythrocytes for various anilin dyes, has 
devised upon this basis an ingenious test for the recognition of 
diabetic blood. Several thick films from a suspected case, con- 
trolled by the same number of preparations made from normal 
blood, are prepared, preferably on slides, and heated in an oven 
to a temperature of 135 C., after which they are set aside to 
cool. Both sets of films are then stained for about two minutes 
with a one per cent, aqueous solution of Congo-red (mixed freshly 
just before using), thoroughly washed in running water, and 
dried between bits of filter-paper. Thus treated, diabetic blood 
is either colored pale greenish-yellow or is entirely unstained, 
while normal blood stains typically the red color of the dye. 
Using the same method of heat fixation, other anilin dyes may 
be employed to demonstrate this peculiar behavior of diabetic 
blood. For example, with a one per cent, aqueous solution of 
methylene-blue the diabetic specimen stains yellowish-green and 

1 N. Y. Med. Jour., 1896, vol. Ixiii, p. 301 (Lit.}; also Med. Record, 1897, 
vol. xii, p. 495. 


the normal film blue. Diabetic blood, on the contrary, treated 
with a one per cent, aqueous solution of biebrich-scarlet, takes 
the color of the dye in a typical manner, while normal blood re- 
mains practically uncolored. Ehrlich's triacid stain, as well as 
mixtures of methylene-blue and eosin and methyl-green and 
eosin, have also been used to demonstrate the reaction. The 
cause of Bremer's reaction is unknown, but apparently it is not 
due to the effect of glucose; many authors are inclined to attrib- 
ute it to excessive acidity of the blood. Positive results with 
this test cannot be regarded as pathognomonic of diabetes mel- 
litus, since they have been reported with the blood of persons suf- 
fering from exophthalmic goiter, multiple neuritis, leukemia, and 
Hodgkin's disease. 

There are no constant changes to be found 
HEMOGLOBIN in these elements. Normal hemoglobin percent- 
AND ages and erythrocyte counts are observed in 

ERYTHROCYTES. most cases, while in others in which the cachexia 
is pronounced a well-marked secondary anemia 
may exist. James, 1 in a study of 13 cases, found the number 
of erythrocytes over 6,000,000 in 5; 5,000,000 plus in 5; 4,000,- 
ooo plus in 2; and 3,000,000 plus in i. The hemoglobin per- 
centage was over 100 in 3 cases; 6070 in 8; and 5060 in 2. The 
alterations in the hemoglobin and erythrocytes in diabetes have 
been attributed by the older writers 2 chiefly to the effects of blood 
concentration and dilution. Thus, it was believed that in cases 
with excessive polyuria the blood became inspissated ajid the 
count thus increased, while in cases with pronounced glycemia 
the blood became diluted and the count lowered as a consequence 
of the fluid transfer from the tissues into the capillaries provoked 
by the presence in the blood of a large percentage of sugar. It 
is obvious that these influences, if active, are sufficient to render 
the blood count in diabetes of little or no practical value, since, 
on the one -hand, perfectly normal blood, if diluted, may appear 
anemic, while, on the other hand, anemic blood, if concentrated, 
may seem normal. James 3 contends, however, that the poly- 
cythemia is real, and is not dependent upon inspissation, for 
were the increase merely relative, it would naturally be accom- 
panied by an increase in the density of the blood, and this in his 
experience never occurred, the specific gravity figure for his 
series ranging between 1.054 and 1.060. 

1 Edinburgh Med. Jour., 1896, vol. xlii, p. 193. 

2 Lit. cited by Leichtenstern, " Unters. iiber d. Hg-Gehalt d. Blutcs," Leipsic, 

3 Loc. cit. 



High digestion leucocytosis is the most con- 
LEUCOCYTES. stant change affecting these cells, but this is not 
found in every case. Isolated examples of leu- 
cocytosis, apparently independent of this influence, have been re- 
ported, but in 1 the great majority of diabetics the leucocyte count 
is normal. The presence of small numbers of myelocytes in 
cachectic patients is the only qualitative change to which atten- 
tion has been directed. Mahogany-colored granules, either within 
the leucocytes or extracellular, may usually be demonstrated by 
the iodin method. No numerical change in the blood plaques has 
been reported. 

Williamson's reaction is of real value, es- 

DIAGNOSIS. pecially in the recognition of cases with temporary 

disappearance of sugar from the urine and in 

diabetic coma. Bremer's test and the iodin reaction are to be 

regarded as symptomatic, not necessarily of diabetes. The other 

blood findings are without diagnostic value. 


Usually the changes in the hemoglobin and 
HEMOGLOBIN erythrocytes are, at the most, trifling, for in 
AND about two-thirds of all cases these elements are 

ERYTHROCYTES. practically normal, while in the other one-third 
moderate anemia, more marked in severe than 
in mild cases, is found. The anemia does not develop until 
about the middle of the first week of the disease, and is, as a 
rule, characterized by a diminution of hemoglobin roughly pro- 
portionate to the corpuscular loss. Degenerative changes are rare, 
consisting usually of nothing more than occasional polychromato- 
philia; nucleation and deformities of size and of shape are absent. 
Regeneration of the blood takes place slowly, and, as the loss of 
hemoglobin is made up less rapidly than that of the corpuscles, 
the color index, which is approximately normal early in the dis- 
ease, later falls considerably. 

The loss of hemoglobin does not often exceed 15 per cent., 
nor is the decrease of erythrocytes usually greater than from 
500,000 to 750,000 cells to the Concentration of the 
blood, which frequently occurs during the febrile period, may 
cause striking temporary polycythemia. 

Morse, 1 in single examinations of 30 cases treated without 
antitoxin, found the count of erythrocytes above 5,000,000 in 

1 Boston Med. and Surg. Jour., 1895, vol. cxxxii, pp. 228 and 252. 


21, and below 4,000,000 in but a single instance, a woman with 
chronic anemia; several of his counts were in the neighborhood 
of 6,000,000. From this author's monograph 1 the following 
counts reported by other investigators are taken: Bouchut 
and Dubroisay, 2 4,305,000 as the mean average of 93 counts in 
84 cases; Gilbert, 3 an average of 4,500,000 in 58 counts in 22 
cases; Carter, 4 an average of 4,253,000 in 13 cases; and File, 5 an 
average of 4,588,000 in 18 counts in 10 cases, some of which had 
received antitoxin. Billings, 6 in a painstaking study of 7 cases 
untreated with antitoxin, in which 36 counts were made, found a 
moderate but distinct decrease in hemoglobin and erythrocytes 
in 5 cases, the loss first becoming apparent by the third or fourth 
day, and being proportionate to the severity of the infection. 
This author's first counts, all made during the first week of the 
disease, ranged from 5,200,000 to 6,122,000, the average being 
5,611,285, the hemoglobin for the same period ranging from 70 
to 98 per cent, and averaging 90 per cent. Subsequent counts 
in this series showed that the greatest loss of hemoglobin averaged 
12 per cent., ranging from i to 30 per cent., and that the maximum 
corpuscular loss averaged 878,500, varying from 227,000 to 2,040,- 

It is generally observed that in cases treated with antitoxin 
the anemia is decidedly less than in those treated by other 
methods, and, in fact, a majority of cases thus treated suffer no 
decrease at all. 

Well-marked leucocytosis, beginning probably 

LEUCOCYTES, within a few hours after the infection first occurs, 

characterizes the average case of diphtheria of 

moderate severity. An analysis of the statistics derived from 276 

counts made by reliable investigators 7 shows that over 90 per 

cent, of all cases are accompanied by a more or less marked 

increase in the number of leucocytes. 

In the majority of cases the number of leucocytes is not in- 
creased above 30,000 per, but a much greater leucocytosis 
is sometimes encountered. Thus, the maximum counts of several 
authors are as follows: Felsenthal, 148,229"; Ewing, 72,000; 

T ' Med. and Surg. Reports of the Boston City Hospital 1899, tenth series, 
p. 138. 

1 Compt. rend. Soc. biol., Paris, 1877, vol. Ixxxv, p. 158. 

3 Traite" de Me"d. Charcot-Bouchard, vol. ii, p. 485. 

4 Univ. Med. Mag., 1894-95, vol. vii, pp. 17, 81, and 158. 

8 Lo Sperimentale, 1896, vol. 1, p. 284. Med. Record, 1896, vol. xlix, p. 577. 

7 Ewing, Morse, Billings, Carter, Schlesinger, File, Gabritschewsky, Bouchut 
and Dubroisay, Rieder, Felsenthal, and Gilbert. 

8 Arch. f. Kinderheilk., 1893, vol. xv, p.78. 
* N. Y. Med. Jour., 1893, vol: Iviii, p. 713. 


Gabritschewsky^ijOoo 1 ; Morse, 48,ooo 2 ; Carter, 48,28o 3 ; Billings, 
38,600*; and Gilbert, 3i,ooo. 5 

This increase is to be regarded as a rough gage of the reac- 
tion of the individual against the effects of the toxic products of 
the disease; it is, therefore, absent in very mild cases, where little 
or no reaction is excited, and in severe cases in which the patient's 
resisting powers are overwhelmed by the intoxication. 

In favorable cases the maximum leucocytosis is reached coin- 
cidentally with the height of the disease, and the increase gradually 
fades away during convalescence, having in most cases entirely 
ceased by the time the membrane has disappeared, but occa- 
sionally persisting after the subsidence of all local and systemic 
manifestations of the illness. In unfavorable cases high leucocyte 
counts persist until death, or "in somewhat prolonged cases, 
with much septic absorption, there may be an uninterrupted 
decrease of leucocytes continuing up to the fatal termination" 
(Ewing). No constant relation has been determined between the 
leucocytosis and the extent of the local lesion, the degree of ton- 
sillar and glandular swellings, or the height of the fever, although 
in individual cases some authors have suggested that such relation- 
ship may be distinguished. 

The effects of antitoxin upon the leucocytes are well illustrated 
by the conclusions of Ewing, 6 based upon 228 counts made in 53 
cases before and after the injection of the serum. As the result 
of these investigations this author concludes that antitoxin, within 
thirty minutes after its injection, causes a hypoleucocytosis. In 
favorable cases, after the injection, the original height of the 
leucocytosis is not again attained, but in severe and less favor- 
able cases the dose of antitoxin is followed in a few hours by 
a hyperleucocytosis exceeding that found in the primary count. 
In malignant cases the administration of antitoxin may be fol- 
lowed immediately either by rapid hyperleucocytosis or by ex- 
treme hypoleucocytosis and death. Bize 7 finds that in some 
cases the initial serum injection may not affect the leucocytes be- 
cause of the insufficiency of the dose, and that repeated injections 
are sometimes required to modify the count. This investigator 
has also called attention to the pronounced leucocytoses which 
accompany eruptions due to antitoxin. 

The leucocytosis of diphtheria, as a rule, involves the polynu- 
clear neutrophile cells, most cases showing about 80 per cent, 
of this variety, but in an occasional instance there may be a 

1 Annal. de PInstitut Pasteur, 1894, vol. viii, p. 673. 2 Loc. cit. 

3 Loc. cit. . 4 Loc. cit. 5 Loc. cit. 

8 Loc. cit. 1 Arch, de med. des Enf., 1901, vol. iv, p. 102. 


well-marked increase in the mononuclear forms, considerably in 
excess of the percentage found in health. Relative lymphocy- 
tosis has been observed both during convalescence and in fatal 
cases with leucopenia, and absolute lymphocytosis may occur at 
the height of the disease in cases with high total leucocyte counts. 
In two of Ewing's cases l the estimates showed 60 per cent, of 
lymphocytes in a count of 72,000 leucocytes, and 62 per cent, of 
these cells in a count of 22,500. 

Besredka 2 believes that marked polynuclear leucocytosis is a 
good prognostic sign, especially if this form of cells shows a 
strong tendency to increase after injection of antitoxin. On the 
contrary, cases which fail to show such a change he regards as 
grave, usually as fatal. This characteristic of high percentages 
of polynuclear neutrophiles is also regarded by many other authors 
as a favorable clinical sign, and a low percentage as unfavorable. 

Ewing, 3 by staining the leucocytes with gentian-violet (50 c.c. 
of normal salt solution to which one drop of a saturated alcoholic 
solution of gentian- violet is a.dded), has deduced certain conclu- 
sions from the reaction of the leucocytes to this dye, to which he 
is inclined to attribute great prognostic value. He believes that 
the numbers and percentages of poorly 'stained leucocytes, and 
usually of ameboid figures, invariably increase in unfavorable 
cases, without relation to the total -number of cells found in the 
blood. In his experience any considerable increase of poorly 
stained leucocytes, especially if associated with a decrease of the 
well-stained cells, was invariably the forerunner of a grave or 
fatal change in the patient's condition. In favorable cases, after 
treatment with antitoxin, he noted that the polymorphous forms 
show a decidedly increased affinity for gentian-violet, this char- 
acteristic often being observed within twelve hours after the in- 
jection of the serum. Failure of this peculiarity to develop he 
regards as* a very unfavorable prognostic sign. 

The proportion of eosinophiles is exceedingly variable, these 
cells sometimes being absent, and at other times found in large 
numbers four or five per cent. Kucharzewski 4 concludes, from 
experimental work with animals, that a high percentage of eosino- 
philes is of favorable prognosis and that a low percentage is unfavor- 

Engel 5 found variable percentages of myelocytes in both favor- 

1 "Clinical Pathology of the Blood," 2d ed., New York and Philadelphia, 1904. 

2 Annal. de PInstitut Pasteur, 1898, vol. xii, p. 305. 

3 N. Y. Med. Jour., 1893, vol. Iviii, p. 713. 

4 XIV. Internal. Med. Congress, Madrid, Apr. 23-30, 1903; abst., Jour. Amer. 
Med. Assoc., 1903, vol. xl, p. 1673. 

8 Deutsch. med. Wochenschr., 1897, vol. xxiii, pp. 118 and 137. 


able and unfavorable cases, especially in the latter, and he con- 
siders their presence in relatively high percentages (2 per cent, 
or higher) as an unfavorable prognostic indication. In 7 of 
Engel's fatal cases the percentage of myelocytes ranged from 3.6 
to 14.6, but they never exceeded 1.5 per cent, in patients who 
recovered. An absence of myelemia, however, is no guarantee 
of recovery, for this sign is absent in about one out of every four 
fatal cases. 

Examination of the blood in diphtheria gives 

DIAGNOSIS, no information which is not clearly shown 

by other clinical signs, and it must be regarded 

as of no value as an aid to diagnosis. The leucocytosis in this 

disease, if the very benign and the very severe cases are excluded, 

is, as a rule, proportionate to the intensity of the infection. 

From a prognostic point of view it appears that, as in pneu- 
monia, an absence of leucocytosis occurring in obviously severe 
infections is an unfavorable indication. The presence of a large 
percentage of myelocytes has a similar meaning. Pronounced 
lymphocytosis is also regarded as an unfavorable prognostic 


In acute catarrhal enteritis the same changes 
ENTERITIS occur that are found in acute gastritis, namely, 
AND little or no alteration in the hemoglobin and 

DIARRHEA, erythrocytes, and an inconstant, moderate leuco- 
cytosis. Profuse watery dejecta lead, of course, 
to more or less blood concentration, by depletion of the body- 
fluids, and hence polycythemia may be a transient sign. In the 
summer diarrheas oj infants Knox and Warfield 1 found that the 
leucocyte count, although usually increased, varies so widely that 
the mere presence of a high or a low count is of practically no 
definite diagnostic importance. An increase in the relative per- 
centage of polynuclear neutrophiles, even with a normal number 
of leucocytes, suggests the onset of inflammatory intestinal com- 
plications. These findings Zahorsky has fully corroborated. 2 In 
chronic' enteritis and in gastro-enteritis the interference with the 
patient's nutrition plus a drain upon the albuminoids may in 
course of time give rise to a decided anemia. Leucocytosis is not 
a characteristic of such cases. 

1 Johns Hopkins Hosp. Bull., 1902, vol. xiii, p. 167. 

2 N. Y. Med. Jour., 1903, vol. Ixxviii, p. 505. 


In dysentery and in ulcerative and phlegmon- 

DYSENTERY. ous enteritis acute forms of secondary anemia are 

frequently met with, especially in patients who 

pass much blood by the bowel. Leucocytosis, often of high 

degree, is also common in these lesions. 

In 38 cases of uncomplicated amebic dysentery Futcher 1 found 
that the hemoglobin averaged 63 per cent, and the erythrocytes 
4,802,000 per, while in 43 cases the leucocytes averaged 
10,600. In 15 cases complicated by amebic abscess of the liver the 
hemoglobin averaged 66 per cent., the erythrocytes 4,250,000, and 
the leucocytes 18,350. Doubtless in both series the high erythro- 
cyte values may be referred to blood concentration. The value of 
the leucocyte count in the diagnosis of amebic hepatic abscess is 
doubtful. In most abscess cases, such as in the n reported by 
Schlayer, 2 the leucocytosis is high (averaging in this series 25,000 
and ranging between 18,000 and 62,000); but in other cases low 
counts (6000, 9000, and 11,000 in Osier's series 3 ) are not incom- 
patible with pus. In Rogers' experience 4 the count is higher in 
small, deeply seated abscesses than in those superficially situated. 
In differentiating malarial fever from the intermittent pyrexia of 
hepatic abscess the presence of leucocytosis practically excludes the 

As a means of differentiating amebic from bacillary dysentery 
the serum test is of value, for the Bacillus dysenteric is clumped 
by the blood of persons suffering from bacillary (Shiga 5 ) dysentery, 
but not by the blood of those infected with the amebic form of the 
disease. Rogers 6 has used the test extensively in India with great 

On similar grounds it may be presumed, until proof to the 
contrary is shown, that Shiga's organism is unaffected by the 
blood of those infected with so-called dysentery due to the bacilli 
of Ogata, of Lessage, and of Roger, and to the "Balantidium coli. 

Shiga's claim, 7 that the Bacillus dysenterm is never clumped 
by the blood of non-dysenteric diseases, needs revision, in the 
light of the researches of Park,* Pilsbury, 9 and Strong. 10 
These observers have found that dysentery bacilli of the 
Shiga, the Flexner, and the Kruse strains occasionally clumped 
with the blood of persons suffering from such diseases as alcoholic 

1 Jour. Amer. Med. Assoc., 1903, vol. xli, p. 480. 

2 Munch, med. Wochenschr., 1903, vol. 1, p. 1372. 

3 Med. News, 1902, vol. Ixxx, p. 673: 4 Brit. Med. Jour., 1902, vol. ii, p. 850. 
5 Centralbl. f. Bakt. u. Parasit, 1898, vol. xxiii, p. 599. 

8 Brit. Med. Jour., 1903, vol. i, p. 1315. 

7 Loc. cit. 8 jour. Med. Research, 1903, vol. ix, p. 180. 

9 Med. News, 1903, vol. Ixxxiii, p. 1078. 10 Rep. Surg.-Gcn., U. S. A., 1900. 


enteritis, tuberculosis, enteric fever, appendicitis, pernicious 
anemia, and pleurisy even in dilutions as high as i : 30, i : 50, 
and 1:100. Still, it is safe to regard a positive reaction with the 
Bacillus dysenterm as the most valuable single clinical sign of 
acute bacillary dysentery in adults in whom a recent chronic in- 
flammation of 'the intestine can be ruled out. The test, to be 
dependable, should be made with at least a 1:20 dilution, the 
time-limit being not more than two hours. Pilsbury found 
that false positive reactions do not occur with the blood of non- 
dysenteric children under one year of age. 

Attempts to isolate organisms by bacteriologi- 
SPRUE. cal examination of the blood were unsuccessful 
in Goadby's hands, 1 although the most careful 
methods were employed. 

The hemoglobin and erythrocytes are markedly reduced, and 
in a remarkably short time after the onset of the acute symptoms 
the cellular loss becomes so severe as to simulate that of true 
pernicious anemia. Counts of between 1,000,000 and 2,000,000 
were found to be the rule by Bassett-Smith, 2 with correspondingly 
low hemoglobin values. There are also decided structural changes 
in the erythrocytes, many of which are deformed in shape and 
size (especially microcytes) and show unnatural pallor. A few 
normoblasts, but no megaloblasts, may be encountered. The 
blood plaques are scanty. 

The leucocytes, according to the last-named writer, are not 
increased, and in some instances fall to between 1000 and 2000 
per Relative lymphocytosis, small-celled in type, relative 
eosinophilia (in the exceptional case), and fractional percentages 
of myelocytes are the other leucocyte findings in this grave form 
of enteric disease. 

The effects upon the blood of saline purges were 
EFFECTS OF first determined by Brouardel, 3 and later studied 
PURGES. by Grawitz 4 and by Hay. 5 The investigations of 
these authors have shown that the administration 
of a purgative dose of Epsom or Glauber salts is followed within 
about thirty minutes by an appreciable increase in the number of 
erythrocytes, and that within an hour the count of these cells is 
fully 1,500,000 more than before the purge was given; three hours 
after this maximum is reached the count is again normal. When 
a certain degree of concentration is obtained by these means, con- 
tinued administration of the salt produces neither additional con- 

1 Brit. Med. Jour., 1903, vol. ii, p. 644. 

1 Ibid., 1903, vol. ii, p. 641. 3 Union med., 1897, vol. xxii, p. 405. 

4 Loc. cit. 5 Jour. Anat. and Physiol., 1882, vol. xvi, p. 435. 


centration nor further purgation. Common table salt is also a 
most energetic factor of blood density, even more so than either 
Epsom or Glauber salts. Purgative doses of jalap, croton oil, 
and other drugs of this class are also followed by more or less 


The alkalinity of the blood is generally de- 

GENERAL creased, a change which may be due partly to the 

FEATURES, effect of the pyrexia and toxemia and partly to 

the anemia. Dare and Funke 1 found subnormal 

alkalinity figures in 20 of 23 cases of enteric fever, but were unable 

to determine the relationship of this change to the other clinical 

manifestations. Drouin 2 also found similar alkalinity values in 

this infection. 

The coagulation time of the blood is appreciably diminished 
in the early stages of typhoid, sometimes to such a degree as to 
favor intestinal hemorrhage. During convalescence the rapidity 
of clotting is decidedly increased, and this tendency may be re- 
garded as a possible factor of thrombosis. Wright and Knapp 3 
suggest that the increased coagulability is due to the excessive 
amount of calcium salts present' in. the blood at the time of con- 
valescence, and that this excess depends chiefly upon the- large 
quantity of lime salts ingested by a patient kept on a milk diet. 
These authors advise, as a preventive measure against thrombo- 
sis, partial decalcification of the milk by the addition of sodium 
citrate (20 to 40 gr. to the pint) as soon as the danger of intestinal 
hemorrhage is over. 

Enteric fever is practically always a specific 
BACTERIOLOGY, bacteriemia, and the Bacillus typhosus can be 
cultured from the circulating blood in more than 
75 per cent, of all cases. From 200 to 300 c.c. of nutrient broth 
sown with 2 or 3 c.c. of blood gives satisfactory results, although 
some investigators, notably Schottmuller, 4 prefer to mix the blood 
directly with melted agar, at a temperature of 45 C., and to 
plate the inoculation, so as to get some idea of the number of 
colonies grown. 

The typhoid bacteriemia usually develops early in the disease, 
and therefore frequently may be demonstrated some days before 
the appearance of the serum reaction. Cultures made during the 

'Johns Hopkins Ho?p. Bull., 1903, vol. xiv, p. 175. 

2 " Herno-alcalime'trie ct Hemo-acidime'trie," These de Paris, 1892, No. 83 

3 Brit. Med. Jour., 1902, vol. ii, p. 1706; also Lancet, 1902, vol. ii, p. 1531. 

4 Munch, med. Wochenschr., 1902, vol. xlix, p. 1562. 



first week of the fever yield a higher percentage of positive results 
than those made during the second and third weeks. With def- 
ervescence the bacilli disappear from the blood, but they reappear 
with a relapse, although not with a simple recrudescence of the 
fever a significant fact when the question arises of determining 
the presence of 'a true reinfection. It is the general experience that 
the severer the type of the infection, the more abundant the 
bacteria in the blood. The following tabulation illustrates the 
frequency of positive results in cases examined by modern methods : 


Schottmiiller : 212 

Hewlett 2 125 

Courmont; Lesieur 3 .. 57 

Harris; Kerr * 56 

Busquet 6 43 

Warfield 43 

Kuhnau 7 41 

Lesieur 8 36 

Reudiger * 27 

Cole 10 15 

Castellani u 14 

Orlovsky 12 12 

Auerbach; Unger 13 10 

Courmont w 9 

Total: TOO 









86 per cent. 








549 Average: 78.4 per cent. 

Coleman and Buxton's analysis 15 of 453 collected cases shows 
the following results of blood culturing during the first four weeks 
of the fever: 



First 85 79 93 per cent. 

Second 198 151 76 " " 

Third 115 65 56 " " 

Fourth 55 18 33 " " 

During recent years numerous instances of paracolon (para- 
typhoid) infection, clinically counterfeiting typical enteric fever, 

1 Munch, med. Wochenschr., 1902, vol. xlix, p. 1562. 

2 Med. Rec., 1901, vol. Lx, p. 849. * Sem. me 

4 Jour. Amer. Med. Assoc., 1902, vol. xxxix, p. 1000. 

5 Presse med., 1902, vol. 1, p. 593. 

* Bull. Ayer Clin. Lab., Penna. Hosp., 1903, vol. i, p. 77. 

7 Zeitschr. f. Hyg. u. Infectionskr., 1897, vol. xxv, p. 492. 

8 Gaz. hebd. de med. et chir., 1902, vol. xlix, p. 1128. 

* Medicine, 1903, vol. ix, p. 258. 

10 Johns Hopkins Hosp. Bull., 1901, vol. xii, p. 203. 

11 Centralbl. f. Bakt. u. Parasit., 1902, vol. xxxi, p. 477. 

12 Phila. Med. Jour., 1903, vol. xi, p. 959. 

13 Deutsch. med. Wochenschr., 1900, vol. xxvi, p. 796. 

14 Jour, de physiol. et de path, gen., 1902, vol. iv, p. 154. 

15 Med. News, 1904, vol. Ixxxiv, p. 1046. 

med., 1902, vol. xx, p. 408. 


have been reported, in which a paracolon bacillus has been 
isolated from the blood during life. Among those reporting 
such cases may be mentioned Gwyn, 1 Longcope, 2 H. W. Allen, 3 
Hewlett, 4 Johnston, 5 and Libman. 6 Of 60 cases clinically typhoid 
studied by Coleman and Buxton, 7 2 yielded blood cultures of the 
Bacillus coli communis. In the late stages of the disease pyo- 
genic bacteria have been found in the blood. 

Examination 0} the Rose Spots. Formerly the cultivation of 
typhoid bacilli from the rose spots was attended by indifferent 
success, but the favorable results with this procedure recently 
announced by a number of authors must stamp it as a distinct aid 
to the diagnosis of typhoid. Neufeld, 8 basing his experiments 
upon the belief that the bacteria lodge and multiply in the spots 
protected from the bactericidal action of the blood, examined 
these lesions in 14 cases and obtained positive results in 13. His 
findings have been corroborated by the work of Curschmann, 9 
who found the bacilli in 14 of 20 cases; of Richardson, 10 whose 
results were positive in 5 of 6 cases; of Kozarinoff, 11 who found 
positive results in 12 of 17 cases; of Seemann, 12 who reports posi- 
tive findings in 32 of 34 cases; and of Scholz and Krause, 13 who 
found bacilli in the spots in 14 of 16 cases examined. The latter 
investigators emphasize the statement that the bacilli are prone to 
disappear from the spots after from three to five days, and that, 
to insure the best results, the examinations must be made as' soon 
as possible after the appearance of the roseola. The most favor- 
able results from spot culturing yet reported are those of Pollaco 
and Gemelli 14 invariably positive findings in 50 consecutive 
cases. All investigators agree that, in the great majority of in- 
stances, spot cultures give positive results several days before the 
appearance of the serum test. The chief disadvantages to this 
method of diagnosis appear to be the absence of the roseola in 
some cases, its late development in others, and the possibility 
of not always being able to distinguish typhoid spots from other 

The technic used by Richardson 15 is simple, and, judging 

1 Johns Hopkins Hosp. Bull., 1898, vol. ix, p. 54. 

1 Amer. Jour. Med. Sci., 1902, vol. cxxiv, p. 209. 

3 Ibid., 1903, vol. cxxv, p. 96. 4 Ibid., 1902, vol. cxxiv, p. 200. 

5 Ibid., 1902, vol. cxxiv, p. 187. 8 Jour. Med. Research, 1902, vol. viii, p. I. 

7 Loc. cit. 8 Zeitschr. f. Hyg. u. Infectionskr., 1899, vol. xxx, p. 498. 

Munch, med. Wochenschr., 1899, vol. xlvi, p. 1597 

10 Phila. Med. Jour , 1900, vol. v, p. 514. 

11 N. Y. Med. Jour., 1903, vol. Ixxviii, p. 196. 

12 Wien. klin. Wochenschr., 1902, vol. xv, p. 580. 
u Zeitschr. f. klin.Med., 1900, vol. xli, p. 405. 

14 Centralbl. f. inn. Med., 1902, vol. xxiii, p. 121. u Loc. cit. 

39 6 


from his results, trustworthy. After having washed the skin of 
the part with Alcohol and ether, the spot is frozen with an ethyl- 
chlorid spray, after which it is crucially incised. Its substance 
is then removed by scraping with a small skin-curette, and trans- 
ferred to a tube of nutrient bouillon. A second tubs is inoculated 
with the blood which collects as soon as the effects of freezing 
have worn off, both cultures being then incubated and examined 
in the usual manner. At least five or six spots should be thus 
treated in each case, and two tubes, one for the scrapings, the 
other for the blood inoculation, used for each spot. 

If blood serum from a case of enteric fever is 

SERUM TEST, mixed with a broth culture of the Bacillus typho- 

sus and a small drop of this mixture placed upon 

a slide and examined under the microscope, it will be observed 


Large clumps o? motionless bacilli sep- 
arated by open spaces. The few bacteria out- 
side of the clumps are devoid of motility. 

A few small clumps of bacilli having 
impaired motility. Persistent motility of the 
bacteria in other parts of the field. 

that the bacilli, instead of continuing to dart actively to and fro 
across the field, as they do in the pure culture, are attracted to 
each other, lose their power of propulsion, and become grouped 
together in large agglutinated clumps of irregular outline, which, 
after the lapse of a variable length of time, become more and 
more compact and homogeneous. In the typical positive reaction 
the field of the microscope shows islands of clumped bacilli, 
separated from each other by large open spaces, containing per- 
haps a few isolated organisms the motility of which is decidedly 
inhibited at first, and finally entirely lost. If the clumps are of 
very large size, they produce a peculiar grayish mottling of the 
specimen visible to the naked eye, a point to which attention 



was first directed by Greene. 1 In a small proportion of cases 
the clumps undergo a granular change, and then become entirely 
destroyed; in some instances they remain unaltered for several 
days; and in still others they may break up after a few hours, 
the field then becoming refilled with isolated, actively motile 

In a certain percentage of instances the agglutinated, motion- 
less masses of bacilli may be observed as soon as the specimen is 
brought into focus, so that the reaction may be said to have taken 
place immediately. In other instances some little time elapses 
before the character of the test can be determined, and in such 
reactions the formation of the clumps, from their inception' out 
of two or three bacteria to their 
completion, when they consist 
of several hundred organisms 
tightly glued together into a 
densely crowded mass, may be 
studied advantageously. In 
these slower reactions, while 
early clumping may progress 
to some extent, many isolated 
bacteria are seen, the motility 
of which persists for a variable 
period, gradually growing less 
and less, until finally, with 
more or less crippled powers 
of propulsion, the organisms 
are attracted to the clump 
centers with which they are 
ultimately incorporated, either 

becoming adherent at first approach, or, as is usually the case 
with the more active bacilli, circling around the edges of the 
clump for some little time before becoming attached to it. Still 
other reactions are characterized by an almost immediate cessa- 
tion of motility, followed by tardy agglutination, and usually by 
the formation of clumps of smaller size than those noted in a 
prompt and immediate reaction. 

If the reaction is negative, the motility of the bacilli persists 
and the formation of clumps is not observed, regardless of the 
time during which the specimen is watched. Not unless aggluti- 
nation is marked and entire loss of motility occurs may a reaction 
be considered positive ; and pseudo-reactions resulting in the for- 
mation of small masses of more or less motile organisms, together 

1 Med. Record, 1896, vol. 1, pp. 697 and 805. 


The bacilli are actively motile throughout the 



with persistent motility of many unclumped bacteria in other parts 
of the field, cannot be regarded as typical in any sense. Clumping of 
small numbers of bacteria sometimes occurs in the pure culture 
during its growth, and this source of error must be eliminated by 
habitually examining the culture before each test or series of tests. 

Technic. In order to exclude all sources of error, such as 
may arise from the clumping of the typhoid bacillus by non- 
typhoid serum, provided that the latter is sufficiently concentrated 
and is allowed enough time to exert its agglutinative powers, the 
reaction can be considered of diagnostic value only under the 
following two conditions: first, that the blood to be tested must 
always be diluted with at least twenty volumes of the culture; and, 
second, that loss of motility and clump formation must occur 
within an arbitrary time limit of ten minutes. Under these condi- 
tions it has been shown that agglutination of the typhoid bacillus 
is produced only by the blood from a patient who is or who 
has been infected with enteric fever, save in exceptional cases. 
(See p. 403.) In some cases of typhoid the reaction occurs in 
much higher dilutions, frequently with dilutions of i : 50, or 
i : 100, or even higher. Some habitually work with higher 
dilutions than i : 20, but they extend the time limit of the reaction 
proportionately to the degree of the dilution used. 

Cultures from twelve to twenty-four hours old, grown in 
neutral peptone bouillon from a stock agar-agar culture, are best 
adapted for the test. It is advisable to keep all the cultures at 
room temperature, and to transplant the stock agar growths not 
oftener than once a month, since cultures " forced" by incubation 
and by frequent transplanting may give rise to false reactions 
with non-typhoid blood. The cultures should, of course, be ab- 
solutely uncontaminated, and must respond typically to the 
recognized tests for their identification. 

The test may be conducted either microscopically, by the dried 
blood method, or by the use of fluid blood or fluid serum; or 
macroscopically, the method preferred by Widal. 1 

The dried blood method, perfected and popularized by Wyatt 
Johnston, 2 is to be chosen whenever it is necessary to send the 
blood sample any distance for examination, and where the ex- 
aminer finds it convenient to carry with him to the patient's 
bedside the test-tubes required for the methods next to be de- 
scribed. Johnston's method is especially adapted for use by 
health boards, by which bodies it is now extensively employed 
in nearly all the large cities in this country. 

1 Bull. m<?d., 1896, vol. x, pp. 618 and 766. 

2 N. Y. Med. Jour., 1896, vol. Ixiv, p. 573. 


The technic of collecting the blood specimens is exceedingly 
simple. After having punctured the finger or ear in the usual 
manner, several separate drops of blood are collected upon the 
surface of some non-absorbent material, preferably glass, then 
dried, placed in an envelop or other protective covering, and 
tested at the examiner's convenience. Glass slides or slips of 
non-absorbent Bristol board or paper are most commonly used 
for collecting the blood samples, and specimens thus obtained 
may be kept for several months without aseptic precautions and 
still retain their agglutinative powers. 

If the specimen has been collected on glass, one of the crusts 
is moistened with a drop of sterile water and worked up into a 
thin paste with a platinum loop, after which complete solution 
and proper dilution of the blood are effected by adding twenty 
drops of typhoid bouillon and mixing thoroughly. If the sample 
has been dried on a paper or cardb.oard surface, the blood crust 
may be cut out with a pair of scissors and placed to soak, face 
downward, in a watch-glass containing twenty drops of the culture. 

From one of these mixtures of blood and typhoid culture a 
minute portion is transferred to the center of a clean cover-glass, 
which is at once inverted over a " concave slide," sealed with 
cedar oil, and examined as a hanging-drop with a -inch dry 
objective, using dim illumination. A plain glass slip is used by 
many workers instead of a "hollow slide," but either will prove 

If the fluid blood method is used, the dilution is made at the 
patient's bedside, by adding one drop of the whole blood as it 
flows from the puncture to twenty drops of typhoid bouillon con- 
tained in a small test-tube. The mouth of the tube is then closed 
by a cotton plug, and its contents are thoroughly mixed by vigorous 
snaking. At the time of the test, which must not be delayed 
more than a few hours after the dilution is made, a small drop of 
the mixture is removed from the tube and examined microscopic- 
ally in the usual manner. In order to insure accurate dilutions, 
a graduated pipette is needed for measuring the blood and the 
culture, for the drops of both liquids must be of exactly the same 
size. Either the special pipettes devised for serum testing or the 
Thoma-Zeiss leucocytometer will prove satisfactory for this pur- 
pose. In lieu of either of these instruments a graduated pipette 
may readily be made from a bit of glass tubing or an ordinary 
medicine dropper. 

If the liquid serum method is chosen, fifteen or twenty drops of 
blood, drawn by making a rather deep puncture, are allowed to 
flow into a narrow-calibered test-tube, and set aside for a few 


minutes until coagulation has taken place. As soon as the clot 
has formed the nose of the graduated pipette is thrust into the 
test-tube, and one drop of the fluid serum sucked up and diluted 
with twenty drops of typhoid bouillon contained in a second test- 
tube. The preparation for microscopical examination is then 
made from this dilution. If the requisite apparatus is at hand, 
the serum may be obtained by centrifugalization. 

If the macroscopical method is employed, the whole procedure 
must be carried out under the strictest aseptic precautions, for 
otherwise the growth of contaminating bacteria may interfere with 
the reaction, owing to the length of time required for the comple- 
tion of the experiment. 

The blood from which the serum is obtained is aspirated from 
one of the superficial veins of the arm, according to the technic 
employed in bacteriological examinations, and then immediately 
expelled into a sterile test-tube, which is plugged with cotton and 
set aside until clotting occurs. If it is desired to send the speci- 
men any distance, the blood may be drawn up into a glass bulb, 
previously 'sterilized by heat, and then sealed at both ends. Blood 
collected in this manner will preserve its agglutinative properties 
and remain sterile indefinitely. 

Having thus obtained the serum from the whole blood, the 
test is carried out by adding the serum in definite dilutions to 
either a twenty-four-hour-old bouillon culture of the typhoid 
bacillus or to sterile bouillon inoculated with a typhoid culture 
at the time of the test. 

In the first instance, a i : 20 dilution of serum and twenty-four- 
hour typhoid bouillon is made in a sterile test-tube, which is then 
plugged and placed in an incubator, where it remains for about 
twelve hours at a temperature of 37 C. At the expiration of 
this time a positive reaction may be recognized by the presence 
of dense, whitish, flaky masses (composed of large clumps of 
agglutinated bacteria), forming in the typical positive reaction a 
thick precipitate at the bottom of the test-tube, which contrasts 
with the perfectly clear appearance of the supernatant bouillon. 
If the reaction is negative, the tube shows simply the uniform 
cloudiness of an ordinary typhoid bouillon culture. 

In the second instance a i : 20 dilution of serum and sterile 
neutral peptone bouillon is made, the mixture then being inocu- 
lated with a small loopful of typhoid bacilli derived from either a 
bouillon or an agar culture. The contents of the test-tube are 
then thoroughly mixed, and the preparation incubated at 37 C. 
for twenty-four hours. A positive reaction is characterized after 
this length of time by the formation of a similar grayish-white 


precipitate at the bottom of the tube, underlying an unclouded 
layer of fluid. In negative reactions the typical cloudiness of the 
typhoid growth is diffused throughout the bouillon. 

In both of these macroscopical methods control tubes of normal 
serum and typhoid bouillon should invariably be prepared, and 
incubated side by side with the specimen of serum to be tested. 

Both tests may be carried out at ordinary room temperature, 
but more certain and more typical results are obtained when the 
tubes are incubated at a temperature of 37 C. 

The Test with Dead Cultures. Reudiger, 1 modifying Picker's 
technic, 2 has devised a very satisfactory serum test with a solution 
of dead Eberth bacilli, made by adding i c.c. of formalin to 100 c.c. 
of typhoid bouillon. Four drops of the suspected blood are mixed 
with 2 c.c. of a i : 50x3 aqueous solution of formalin, and, after 
laking has occurred, i c.c. of this blood solution is mixed in a 
small test-tube with 4 c.c. of the dead culture. This gives ap- 
proximately a 30 : i dilution of the blood. The test-tube is then 
set aside in a vertical position, with the result that, if the reaction 
is positive, a flocculent precipitate will appear within an hour 
or two, falling to the bottom of the tube as a granular sediment, 
with clearing of the supernatant bouillon within from twelve to 
twenty-four hours. By microscopical examination this sediment 
is found to consist of tightly clumped masses of bacilli. 

In his clinic at the Jefferson Hospital the writer uses this test, 
in connection with the more rapid hanging-drop method with live 
bacilli, in all suspected enterics, and regards it as quite as accurate 
as the older method, although, of course, much slower. Dried 
blood as well as fresh may be employed, if the precaution is taken 
to powder it finely before adding it to the formalin solution. 
Reudiger obtained positive results with test cultures killed a year 

The Choice oj a Method. The choice between the methods 
of serum testing described above must depend largely upon the 
circumstances under which the test is to be made. 

The dried blood method, as already remarked, is best adapted 
to health board necessities, where samples of blood are collected 
by the general practitioner and sent by mail to the laboratory. 
The chief objection to this method is the impossibility in many 
cases of making accurate dilutions, since usually it can only be 
assumed that a given crust of blood represents the same volume 
in its liquid state as the drop of culture with which it is diluted. 
If the examiner collects the specimens himself, quite accurate dilu- 

1 Jour. Infect. Dis., 1904, vol. i. p. 236. 

2 Berlin, klin. Wochenschr., 1903, vol. xl, p. 1021. 


tions may be obtained if a graduated pipette or a platinum loop 
of fixed size is used to measure both the blood and the culture. 
Another drawback is the fact that more or less typical agglutina- 
tion occasionally occurs with non-typhoid blood, although John- 
ston believes that this source of error may always be eliminated 
by using cultures of sufficient attenuation. 

In hospital work either the fluid blood or fluid serum should 
be chosen, for exact dilutions may be made by these methods, and 
if the test is made within a reasonable length of time after the 
dilution, bacterial contamination need not be feared. The writer, 
a warm advocate of Johnston's method in the early days of serum 
testing, has now discarded it wherever possible in favor of the 
more accurate and equally simple test performed with the fluid 
whole blood or serum. 

The macroscopical method with fluid serum is too slow, and 
requires too elaborate bacteriological apparatus ever to be adopted 
for general clinical use. Reudiger's method is obviously of 
great value to the practitioner who has not access to a laboratory. 1 

Value of the Serum Test. Fully 95 per cent, of all typhoids give 
a positive reaction at some period of the disease, usually as early 
as the eighth day, as nearly as it is possible to compute this period. 
An error of about three per cent, must be allowed for, on account 
of the occurrence of positive or misleading results with non-typhoid 
blood. The statistics of the Philadelphia Bureau of Health 2 are 
of interest in demonstrating the usefulness of the test in routine 
public health work. These data, covering a period of six years 
(1897-1902), show that, of 22,521 tests by the dried blood method, 
in 19,080 cases diagnosed as enteric fever, the discrepancy between 
the laboratory , and the clinical diagnosis ranged between 3.9 and 
8.3 per cent. Rosenberger's collection 3 of 17,280 cases from 
other sources show's 16,352 positive results, or 94.6 per cent. 

In some instances repeated negative results are found until con- 
valescence is well established, and, rarely, cases are encountered 

1 A. J. Wolff (Amer. Jour. Med. Sci., 1903, vol. cxxv, p. 661) has devised a 
clever method of serum testing with the patient's blood and feces. Bouillon 
cultures of the patient's feces are incubated for twelve hours, mixed, in definite 
dilution, with his blood, and examined microscopically. If Eberth bacilli exist 
in the culture thus made, they become clumped and immobilized, while the motility 
of the colon organisms also present is unimpaired. If no reaction occurs, the 
patient's blood should be tested in the usual manner as a control. Wolff reports 
uniformly positive results in 35 cases of enteric fever with this modification of the 
Widal test, and claims that reactions occur from the fourth to the seventh day 
in spite of the general impression that the stools in typhoid do not contain the 
Eberth bacillus before the tenth or eleventh day of the disease. 

2 A. C. Abbott, Annual Report of the Division of Bacteriology, Pathology, and 
Disinfection of the Philadelphia Bureau of Health, 1903, p. 112. 

3 Proc. Path. Soc. Phila., 1904, vol. vii, p. 97. 


in which the reaction never occurs at any time during the entire 
course of the illness. The blood may lose its clumping powers 
at about the time of defervescence, or, on the other hand, this 
peculiarity may persist for months or even years after the attack. 
This last source of error in the test may be due to the presence 
of unsuspected foci containing typhoid bacilli, but sometimes it 
is apparently independent of such factors. Before pronouncing 
upon the value of a positive reaction in the individual case the 
occurrence of a previous attack of typhoid and the presence of 
lesions which may be due to the Eberth bacillus (osteomyelitis, 
cystitis, arthritis, cholecystitis, etc.) must be excluded. 

The reaction may be positive on one day of the disease or for 
a series of days, and negative on the next day or succeeding days. 
Its character is apparently uninfluenced by the intensity of the 
infection, for although, as a rule, the reaction is usually prompt 
and marked in severe infections, it may be just as marked in mild 
cases. A certain relationship seems to exist between the height 
of the fever and the intensity of the reaction, for the latter is 
usually most decided at the period of maximum pyrexia. 

Positive Reactions in Non-typhoid Conditions. In a number 
of conditions other than enteric fever the blood may acquire a 
more or less decided agglutinative action toward the Eberth 
bacillus, but, with rare exceptions, such reactions are attributable 
to low dilutions, to a prolonged time limit, and, perhaps, to a 
previous attack of typhoid. Typhus fever, malarial fever, sepsis, 
pneumonia, tuberculosis, acute osteomyelitis, and influenza are 
among the most important diseases thus simulating a true 
typhoid reaction, and a positive test in any of these conditions 
must, to be conclusive, occur in a high dilution i : 50, i : 60, 
or higher. In Weil's disease positive reactions are common, even 
in high dilutions, and this fact tends to corroborate Weil's original 
suggestion, that this disease in reality is nothing more than 
typhoid aborted by the supervention of jaundice. Or, as 
Lubowski and Steinberg suggest, 1 the reaction may sometimes be 
due to. a Bacillus proteus infection, since immune proteus serum 
may agglutinate the Eberth bacillus, in very high dilutions. The 
blood of patients suffering from other forms of jaundice also has 
a moderately strong agglutinative effe.ct upon the typhoid bacillus, 
although bile itself has no such action. With proper technic, 
however, the blood of jaundice cases need not be a source of 
error, for Libman, 2 in a study of 35 such instances, failed to 
find in any one of them greater agglutinative effect than is often 

1 Deutsch. Arch. f. klin. Mcd., 1904, vol. Ixxix, p 396. 

2 Med. News, 1904, vol. Ixxxiv, p. 204. 


seen in normal, blood. Errors of technic and latent or previous 
typhoid infection are the probable explanation of most of the 
positive results with bilious blood and typhoid cultures, reported 
by Koenigstein, Kochler, Greenbaum, and other earlier students 
of this question. 

In paratyphoid fever the blood serum usually does not clump 
the Eberth organism, although in an occasional instance it does 
so, even in high dilutions. In doubtful cases, therefore, the ag- 
glutination reaction with the typhoid bacillus should always be 
supplemented by similar tests with paratyphoid cultures, as well 
as by bacteriological examination of the blood for the isolation 
of a specific bacterium. 

In the light of our present knowledge of the serum test its 
value appears to be less than its first enthusiastic advocates were 
inclined to urge. A positive reaction, obtained by a skilled 
worker whose technic as to dilution, time limit, and culture has 
been exact, is the most valuable single sign of enteric fever, al- 
though it cannot be regarded as absolutely pathognomonic. On 
the other hand, a negative result with the test is no proof of the 
absence of the disease. 

Anemia develops shortly after the beginning 
HEMOGLOBIN of the fever, slowly and progressively increasing 
AND in intensity throughout the course of the disease, 

ERYTHROCYTES. and persisting during the early weeks of defer- 
vescence. During the first week there is little or 
no decrease in the number of erythrocytes, although the hemo- 
globin loss appears to begin coincidentally with the first manifes- 
tations of the infection. Normal erythrocyte counts are the rule 
during the first seven days, but there are few cases of typhoid 
which fail to show a hemoglobin loss amounting to at least 15 
or 20 per cent, during this period. Whether this early oligo- 
chromemia is due to the influence of the fever, or whether it rep- 
resents the actual prefebrile state of the patient's blood, is difficult 
to decide. By the second week the corpuscular decrease becomes 
evident and steadily grows more and more marked as the disease 
progresses, reaching its maximum at about the end of deferves- 
cence.. Thayer 1 distinguishes a slight accentuation of the oligo- 
cythemia between the third and fourth weeks, the decrease con- 
tinuing until the seventh, when a still more decided fall occurs, 
followed in the eighth week by a considerable rise. A slow rise 
in the erythrocyte curve is observed after defervescence, and, 
indeed, sometimes before the end of the febrile stage, until by the 
end of the fourth or fifth week of convalescence the count is again 

1 Johns Hopkins Hosp. Reports, 1900, vol. viii, p. 487. 



normal. The hemoglobin after the first week follows the same 
general course as the erythrocytes, but its decrease during the 
early weeks is relatively greater and its regeneration slower during 
the post-febrile period. These data, as well as those relating to 
the leucocytes, have been confirmed by W. A. Winter. 1 

Thayer's analysis 2 of the blood examinations in enteric fever, 
made in the Johns Hopkins Hospital during a period of eleven 
years, furnishes a striking illustration of the development of the 
anemia during the progress of the disease. Arranged according 
to the week of the fever, the following hemoglobin and ery- 
throcyte averages in uncomplicated cases are shown : 


(165 estimates.) 

ist week, 

21 estimates 

.76.1 per cent. 

6th week, 6 estimates 

.62.1 per cent. 

2d " 

5i " 

.72.8 " " 

7th " 4 

-50-5 " " 

3d " 


.66.2 " " 

8th " 3 


4 th " 


.60.5 " " 

..9th 4 

.47-7 " " 


2O " 

-57-8 " " 

10th 2 " 

.66.5 " " 


(265 counts.) 

ist week, 

32 counts 

. .4, QI3, 312 

7th week, 8 counts 

. . 3,3OQ,I2s 




8th " 7 " 

7, 6^2, 28; 



. .4, 42Q,2O8 

9th " 6 " 

31 coo. 066 



. 4,222,2?6 

loth " i " 

3 .020.000 



. .4,Il8,C.OO 

nth " i " 

. . 2,100,3*3 




. .4,028,428 

Seventy-four typhoid patients, both with and without compli- 
cations, examined in the German and the Jefferson Hospitals 
showed the following averages, the first estimates being taken in 
all cases in which multiple examinations were made : 


ist week, 14 cases 77.4 per cent. 














4,789,285 per 
4,161,233 " 
3,555,000 ' 
3,490,833 " 
2,445,000 ' 
3,165,000 " 
3,335,000 ' 
2,790,000 ' 

As a rule, the degree of a typhoid anemia is parallel to the 
severity of the attack, but this is not invariably true, since a mild 
case may be associated with a most intense anemia. In the series 

1 Dublin Jour. Med. Sci., 1901, vol. cxii p. 249. 2 Loc. cit. 


included in t^e last tabulation the most marked instances of 
anemia showed hemoglobin and erythrocyte estimates of 40 per 
cent, and 1,720,000; 40 per cent, and 1,850,000; and 50 per cent, 
and i ,800,000, respectively. The most striking example of oligo- 
chromemia showed a hemoglobin percentage of 20, with a cor- 
responding erythrocyte count of 2,470,000. Considerably lower 
estimates than these have been reported by a number of other 
observers, but they are uncommon. 

The effects of the cold tub and of excessive diarrhea and 
sweating may cause a temporary polycythemia from concentration 
of the blood, and these sources of high counts must be excluded 
in making examinations during the early weeks of the disease. 
In four cases examined by the writer to determine the effects of 
the cold plunge it was found that the average erythrocyte increase 
after the bath amounted to 813,000 corpuscles per, and 
the hemoglobin gain to 8 per cent. Hemorrhages, if severe, may 
cause an abrupt fall in the erythrocyte count, often succeeded by 
a more or less successful attempt at regeneration, in an effort to 
compensate for the blood loss. 

Qualitatively, the cells show no peculiar changes, poikilocytosis, 
irregular staining affinities, and deformities of size occurring in 
relation to the intensity of the anemia. Erythroblasts are com- 
paratively rare, being absent or few in number in the average case. 
Normoblasts may be found in cases with high-grade anemia and 
as a sequel to hemorrhage. Megaloblasts are very rare, an occa- 
sional cell of this type being observed now and then only in severe 

A steady, slow decrease in the number of leu- 
LEUCOCYTES. cocytes becomes evident after the first week of 
the fever, the lowest counts being found during 
the fifth or sixth week, after which an increase, which may be 
either permanent or transient and followed by a still more decided 
leucopenia, is observed. It appears that the latter change accom- 
panies cases with severe post-febrile anemia, although sufficient 
data are lacking to justify absolutely positive conclusions on this 
point. In uncomplicated cases the normal count becomes re- 
established by about the fourth week of convalescence. The 
leucopenia of typhoid corresponds in a general way to the severity 
of the attack, and although not marked in the average, in the 
individual case it may be striking, counts of from 2000 to 3000 
being not at all uncommon. 

Kast and Gutig, 1 in 105 cases, found the leucocytes below 
7000 in 97, between 7000 and 9000 in 6, and 9000 or higher in 2 

1 Deutsch. Arch. f. klin. Med., 1904, vol., p. 104. 



cases. Contrary to the experience of others (see below), these 
authors found little or no leucocytosis as the effect of non-typhoid 
complications, which in their cases generally caused simply a 
polynuclear increase with no disturbance of the leucopenia. 

Thayer's report of 832 counts in uncomplicated cases shows 
the following range of the leucocytes, according to the week of 
the disease : 

ist week, 119 counts 6442 8th week, 14 counts 6614 


7 th 









543 1 


1 2th 
1 3th 




. . 8000 

The leucocyte estimates of the 74 hospital typhoids referred to 
above averaged: 

ist week, 14 cases 8026 leucocytes per 



7 th 






Disregarding the week of the fever, the number of leucocytes 
in these cases ranged as follows: 

Above 10,000 in 7 cases. 

From 9,000-10,000 in 3 

8,000- 9,000 8 

7,000 8,000 14 

6,000- 7,000 

. 10 

I ! 

5,000- 6,000 
4,000- 5,000 
3,000 4,000 
2,000 3,000 
1,000- 2,000 

Highest, 16,000 per 

Lowest, 1,333 

Average, 6,706 

It appears from these figures that counts in excess of 10,000 
per may be looked for in more than ten per cent of all 
cases, such an increase being due either to the effects of blood 
concentration from diarrhea, sweating, vomiting, or cold tubbing, 
or to some hidden or frank complication. In four of the seven 
relatively high counts above noted the cause was plain croupous 
pneumonia in two, cholecystitis in one, and furunculosis in one. 


In the other three, all of which were made in patients whose fever 
had not yet run seven days, the factors of the increase were unde- 
termined; possibly it was due to physiological blood inspissation. 

Inflammatory complications, such as otitis, abscess, pneumonia, 
severe bronchitis, peritonitis, cystitis, periostitis, and phlebitis, give 
rise to a prompt leucocytosis in patients whose vital powers are 
sufficiently strong to react against the process. Intestinal hemor- 
rhage is usually followed by an increase reaching its maximum 
within twenty-four hours after the blood loss, and disappearing 
within a week. Intestinal perforation may promptly be followed 
by a leucocytosis, the increase developing within a few hours. 
Thayer has observed that in some instances the increase in the 
number of leucocytes succeeding the perforation may tend to 
diminish and disappear with the aggravation of the symptoms, 
and that not infrequently there is a complete absence of leucocy- 
tosis and sometimes a diminution in the number of leucocytes 
after this accident. He also considers that the prospect of relief 
by surgical interference is best in those cases with a leucocytosis, 
the absence or disappearance of this sign following a perforation 
being an indication of the malignancy of the infection or the pros- 
tration of the patient. 

Qualitative changes are absent or inconspicuous during the 
first two weeks of the fever, but during the third week a slow, 
progressive decrease in the relative percentage of polynuclear 
neutrophiles with a consequent increase in the mononuclear un- 
granulated forms, begins, this change becoming most marked at 
about the end of defervescence. In 23 of the writer's cases the 
percentage of polynuclears averaged, according to the week of 
the disease, 75.0 per cent, for 8 cases in the first week; 70.9 per 
cent, for 7 in the second week; 50.2 for 4 in the third week; 60.0 
per cent, for 2 in the fourth week; and 64.0 and 68.0 per cent, 
for a single case in the fifth and sixth weeks, respectively. Higley * 
finds that the polynuclear neutrophiles decrease much earlier in 
the disease, 9 of his cases examined during the first week showing 
an average of 59.4 per cent, for these cells. 

Thayer has found that the mononuclear cells which are most 
markedly increased are "elements containing nuclei not much 
larger than those of lymphocytes, and often presenting the general 
appearance of a lymphocyte nucleus, with the exception of the 
slight affinity for coloring matters. The size of these cells is 
usually about that, or but little larger than that, of the ordinary 
polymorphonuclear neutrophile." The typical small lymphocyte 
and the transitional forms undergo little or no increase. 

1 Med. News, 1903, vol. Ixxxiii, p. 1140. 



The eosinophiles are almost invariably decreased, both abso- 
lutely and relatively, and are often absent during the active 
febrile stages. The relative percentages of these cells in the 
above cases averaged 0.87, rising as high as 5 per cent, in only 
2 cases, and being entirely absent inn. 

Myelocytes in small numbers may be found in severe forms 
of post-typhoid anemia, but they are absent during the active 
period of the infection. Turk's "stimulation forms" are met 
with under the same conditions. 

Thayer's elaborate report of the Johns Hopkins Hospital cases 
includes the following averages of the differential leucocyte counts : 







ist week, 12 counts 

12.9 per cent. 

12.4 per cent. 

74.0 per cent. 

0.5 per cent. 




13-4 ' 






| n.6 ' 





20. i 
















7 th 









1 6.8 

5 6 -9 


According to Hayem, 1 the number of blood plaques is markedly 
decreased during the febrile period of the fever, as in any other 
condition characterized by pyrexia. ' 

The blood examination furnishes four clinical 

DIAGNOSIS, signs of positive value in the diagnosis of enteric 
fever: the serum reaction; a subnormal leuco- 
cyte count or at least an absence of leucocytosis ; bacteriemia; 
and in cases with roseola the detection of the Eb'erth bacillus by 
spot culturing. The influence of complications upon the behavior 
of the leucocytes must, however, always be borne in mind. 

Acute miliary tuberculosis, cerebrospinal meningitis, malarial 
fever, certain atypical cases of pneumonia and influenza, and 
septicemic and pyemic processes, such as ulcerative endocarditis, 
are the diseases most frequently confounded with typhoid, and in 
their differentiation the blood report often gives just the essential 

Acute miliary tuberculosis, if a pure infection, shows a similar 
absence of a leucocyte increase, and in excluding this disease 
reliance must be placed on the Widal test and upon blood cul- 
turing. Influenza is not characterized by leucocytosis, and must 
be differentiated from typhoid by the aid of the same methods of 

1 "Du Sang," etc. Paris, 1889. 


examination. jCerebrospinal meningitis, pneumonia, and septic 
and pyemic conditions may be differentiated by their association 
with a more or less well-marked leucocytosis. In the last-named 
processes bacteriological examination of the blood not infrequently 
gives conclusive results. Paratyphoid fever, often clinically identi- 
cal with genuine enteric fever, shows also a similar anemia and 
leucocyte range. As a rule, paratyphoid blood does not agglutin- 
ate cultures of the Eberth bacillus; if it does so, the reaction is 
less marked than with cultures of the paratyphoid organism. 
Blood culturing is the court of final appeal in distinguishing these 
two closely related infections. The differentiation of malarial 
fever is referred to under this disease. (See p. 472.) 


In severe infections 'Turk 1 has noted a de- 
GENERAL cided increase in the quantity of fibrin and in the 
FEATURES, number of blood plaques, but in the case of average 
severity these changes are not to be observed. 
Drouin 2 has found that the alkalinity of the blood is greatly de- 
creased. Negative results from bacteriological examination of the 
blood are the rule, although streptococci (Fehleisen), diplococci 
(Pfahler), and other pyogenic bacteria invade the blood at and 
near the erysipelatous lesions. 

Moderate anemia, characterized by a some- 
HEHOGLOBIN what disproportionate hemoglobin loss, is com- 
AND mon in the severer forms of the disease, but not 

ERYTHROCYTES. in mild cases. The decreases are not notable, 
amounting on the average to a loss of -not more 
than 10 or 20 per cent, of corpuscles and of about 30 per cent, 
of hemoglobin. Maragliano's degenerative changes of corpus- 
cular structure have occasionally been found. 

Leucocytosis of the polynuclear neutrophile 
LEUCOCYTES, type is the usual finding, but mild cases fre- 
quently run their course without provoking the 
slightest increase. Except in isolated instances, the leucocytosis 
is not -high, the counts usually being about 15,000, and rarely 
more than 20,000, cells to the Von Limbeck 8 and Chante- 
messe and Rey 4 have shown that the leucocyte and temperature 
curves maintain a definite parallelism in the majority of cases, 
and that the diminution in the leucocytosis, as a rule, anticipates 
the fall in temperature. There is, however, no apparent rela- 

1 Loc. cit. 2 Thfcse de Paris, 1892, No. 83, p. 108. 

3 Loc. cit. 4 Presse md., 1899, vol. vi, p. 316. 

FEVER. 411 

tionship between the height of the count and the degree of 
pyrexia, for moderate leucocytoses are not incompatible with 
strikingly high temperatures. It is generally agreed that the 
highest counts are found in the severest cases, provided that the 
patient's resisting powers are acting normally. An extension 
of the lesion is generally accompanied by an increase in the 

With the onset of convalescence, as the leucocytosis disappears, 
the normal percentages of the different forms of corpuscles, dis- 
turbed during the febrile period, are reestablished by a rapid 
increase in the small lymphocytes and eosinophiles, and a decrease 
in the polynuclear neutrophiles ; the percentage of large lympho- 
cytes remains stationary, and the eosinophiles are absent during 
the height of the attack, according to Chantemesse and Rey. 
Small percentages of myelocytes and an occasional "stimulation 
form" are commonly found during the active stages of the leuco- 


There are no characteristic changes in the hemoglobin and 
erythrocytes, although an anemia indistinguishable from typical 
chlorosis is not an infrequent feature. Such cases must be dis- 
tinguished from so-called "thyroid chlorosis," or chlorosis with 
thyroid hypertrophy, by means of other clinical symptoms. In 
nine cases of which the writer has notes the hemoglobin averaged 
78.6 per cent., and the erythrocytes 4,715,571 per c'.mm. It 
seems likely that in cases of Graves' disease characterized by 
excessive diaphoresis, emesis, and diarrhea, the blood concen- 
tration thus produced may be sufficient more or- less effectually 
to obscure the real grade of anemia existing. 

The leucocytes are not increased in number, and leucopenia of 
a decided degree is frequently observed. Of the above cases, the 
leucocyte average was 4698 per, none of the counts exceeding 
10,000. Relative lymphocytosis is a common change, and moderate 
increase in the percentage of eosinophiles an occasional finding. 


It is a well-recognized fact that more or less hemoglobin and 
erythrocyte losses follow pyrexia maintained for any length of 
time, but an attempt to demonstrate the exact cause or group of 
causes of this anemia involves the analysis of a most complex 
problem in physiology, about which the most skilled investigators 


express diametrically opposite opinions. Some maintain that suffi- 
cient actual destruction of the corpuscles occurs as the result of 
fever to account for their decrease in number, while others attrib- 
ute the loss largely, if not wholly, to the influence of vasomotor 
changes. Maragliano l has shown that capillary contraction ac- 
companies the period of active pyrexia, while Reinert 2 suggests 
that the blood is diminished in volume by the excessive drain 
upon the body fluids occurring at this time. In septic fevers, 
furthermore, additional inspissation of the blood is produced by 
the influence of bacteria and their products. These factors, tend- 
ing to inspissate the blood, favor the 1 production of polycythemia, 
which change is to be observed during the stage of active fever. 
But as defervescence sets in the conditions are reversed, for the 
capillaries then dilate, the draining away of the fluid elements of 
the blood ceases, and, consequently, dilution of the blood now 
occurs. Anemia, therefore, develops coincidentally with the dis- 
appearance of the fever. It is undetermined whether this post- 
febrile anemia is the result purely of these physical causes or of 
these causes plus a certain amount of real hematocytolysis due 
to high temperature. It seems reasonable to regard both factors 
as active. 

Coagulation and fibrin behave so erratically that no definite 
statements regarding them are justified. In the incipient stage 
of septic fevers coagulation is much delayed, according to Schmidt, 3 
but during the later stage it occurs more rapidly than normal. 
The leucocytes in this class of fevers are generally increased in 

The alkalinity of the blood undergoes wide* variations in differ- 
ent febrile states, but it cannot be said that these changes, which 
are probably due to complex chemical processes rather than to 
the primary effect of the fever, are constantly parallel, either to 
the degree of pyrexia or to the behavior of the leucocytes. Lb'wy 
and Richter 4 state that increased alkalinity occurs coincidentally 
with the stage of hypoleucocytosis a statement which Strauss, 5 
Lowit, 6 and others have verified. Fodor and Rigler's experi- 
ments 7 have proved that the pyrexia following infection with 
pathogenic bacteria ultimately effects a diminution in the alkalinity 
of the blood, and that this change is sometimes preceded by a dis- 

1 Berlin, klin. Wochenschr., 1887, vol. xxiv, p. 797. 

2 "Die Zahlung der rothen Blutkorperchen," Leipsic, 1891. 

3 Pfliiger's Arch., 1875, vol. xi, pp. 291 and 515. 

4 Deutsch. med. Wochenschr., 1895, vol. xxi, p. 526. 

5 Zeitschr. f. klin. Med., 1896, vol. xxx, p. 315. 
8 "Die Lehre v. Fieber," Jena, 1897. 

7 Centralbl. f. Bakt. u. Parasit., 1897, vol. xxi, p. 134. 


tinct primary increase. The conclusions voiced by von Jaksch, 1 
Krause, 2 and other earlier writers, that decreased alkalinity is a 
constant accompaniment of febrile processes, cannot be unreserv- 
edly accepted, if von Limbeck's 3 later statements to the contrary 
are to be believed. 


Filariasis, the pathological condition depend- 
OCCURRENCE. ing upon the presence in the body of the parental 
and embryonic forms of the Filaria sanguinis 
hominis, is of wide-spread distribution throughout the tropics and 
subtropics, being prevalent in various districts of Africa, India, 
Australia, China, Japan, South America, and the islands of the 
South Pacific and the West Indies, and, as mentioned below, hav- 
ing been found to a limited extent in North America. 

Six distinct species of embryo blood worms, 
PARASITOLOGY. the parental forms of which do not .enter the cir- 
culation, have been demonstrated in the periph- 
eral blood of man, these parasites being known by the general 
term Filaria sanguinis hominis. These different filariae, according 
to the nomenclature suggested by Manson, 4 are distinguished by 
the names Filaria nocturna, 5 Filaria diurna, Filaria perstans f Fila- 
ria demarquaii, Filaria ozzardi, and Filaria magalhdesi. To but a 
single member of this group, the Filaria nocturna, has an undis- 
puted pathological role been assigned, this parasite being regarded 
as the cause of various forms of ulcer, lymphangitis, lymph varices, 
lymph scrotum, tropical elephantiasis Arabum, endemic chyluria, 
chylous ascites, and other tropical diseases of more or less obscure 
nature. The Filaria Persians, Manson conjectures, may possibly 
be the cause of that peculiar tropical disease known as African 
kra-kra, or "craw-craw." Christy 6 couples this worm etiologically 
with the "tick fever" of Uganda, and has shown that a variety 
of tick (known locally as the Bibo) acts as its intermediary host. 
Bastian 7 suggests that the embryonic forms of the Filaria Persians 
are in reality embryos of a species of Tylenchus, which infests the 
roots of the banana. By eating the fruit thus contaminated it is 

1 Zeitschr. f. klin. Med., 1887, vol. xiii, p. 380. 

* Zeitschr. f. Heilk., 1889, vol. x, p. 106. 

3 Centralbl. f. inn. Med., 1895, vol. xvi, p. 649. 

4 "Tropical Diseases," 3d ed., New York, 1903. 

* The author is greatly indebted to Dr. F. P. Henry and to Dr. J. H. Gibbon 
for the opportunity of making repeated blood examinations in two cases of Filaria 
nocturna infection occurring in their respective hospital services. 

8 Thompson Yates and Johnston Lab. Rep., 1903, vol. v, p. 187. 
7 Lancet, 1904, vol. i, p. 286. 



possible that s6me of the worms ingested may bore through the 
gut and reach the mesentery, where they rest and develop, and 
whence they later pass into the general circulation via the lymph 
stream. Low, 1 on the contrary, has pointed out that the distribu- 
tion of the Filaria perstans does not always correspond to that 
of banana cultivation, and contends that this organism is in no 
way related to the genus Tylenchus. This investigator 2 has dis- 
proved the belief, once held, that the Filaria perstans is the 
specific cause of " sleeping sickness." The other filariae (diurna, 
demarquaii, ozzardi, and magalhaesi] possess no interest from 

a diagnostic standpoint, 
since their life history and 
pathological significance 
are still obscure. 3 


THE FILARIA is by far 
NOCTURNA. the most 
ant member of the above- 
named class of blood 
worms, being the one 
most familiarly known 
of all, as well as the one 
of greatest clinical in- 
terest, because of the 
interesting pathological 
lesions which it is cap- 
able of exciting. In this 
country cases of filariasis 
due to the Filaria noc- 
tnrna have been reported by a number of different observers, Gui- 
teras, 4 de Saussure, 5 Mastin, 6 Slaughter, 7 F. P. Henry, 8 Dunn, 9 and 
Lothrop and Pratt 10 having met with the disease. A few of these 
cases have been regarded by their reporters as indigenous, but the 
great majority of them, it is safe to state, were directly imported 
from the tropics. To the writer's knowledge, at least five cases 
have been diagnosed in Philadelphia during the last six years. 

1 Lancet, 1904, vol. i, p. 420. J Brit. Med. Jour., 1903, vol. i, p. 722. 

3 For a complete description of filariasis and of the various forms of the filariae 
the reader should consult Manson's text-book, above mentioned. Davidson's 
" Hygiene and Diseases of Warm Climates " (Edinburgh, 1893) contains an excellent 
account of the histological structure of filariae. 

4 Med. News, 1886, vol. xlvii, p. 399. 6 Ibid., 1890, vol. Ivi, p. 704. 

* Annals of Surg., 1888, vol. viii, p. 321. 7 Med. News, 1891, vol. ii, p. 649. 
8 Ibid., 1896, vol. xviii, p. 477. Trans. Coll. Phys. Phila., 1898, p. 80. 

10 Amer. Jour. Med. Sci., 1900, vol. cxx, p. 525. 


From a photomicrograph of the parasite in a fresh blood 



As may be inferred from the name, the embryos of the Filaria 
nocturna are found in the peripheral blood most abundantly at 
night, the vast majority of the parasites retiring into the deeper 
circulation during the daytime. From late in the afternoon until 
about midnight they make their way into the peripheral vessels 
in progressively increasing numbers, with more or less fluctuation, 
the maximum number being found at the latter time, after which 
they begin to grow less and less numerous, until, by about eight 
o'clock in the morning, they have practically all disappeared 
from the superficial circulation and reentered the deeper vessels, 
in which they remain until the close of the day. This peculiar 
periodicity is well illustrated by a recent series of investigations 
made by Lothrop and Pratt, 1 who have charted the phenomenon 
in one case, showing the approximate number of parasites to the of blood as follows: 4 p. M., 100; 6 p. M., 275; 8 P. M., 1300; 

10 P. M., QOO; 12 M., 1500; 2 A. M., 700; 4 A. M., 900; 6 A. M., 125; 

8 A. M., 125; and 10 A. M., 100. The highest number ever ob- 
served by these authors was 2100 embryos per, "the specimen 
in which this count was made having been taken at midnight. 
This characteristic periodicity, it should also be remarked, is com- 
pletely reversed if the individual harboring the parasite reverses 
his habits of life, sleeping during the day and moving about at 
night. If such should be the case, the worms will appear in the 
peripheral blood during the daytime, the patient's period of rest, 
and seek the deeper circulation at night, the patient's period 
of activity. : * 

The painstaking studies of Manson 2 and Low 3 have shown that 
mosquitos (Culex fatigans and several of the genus Anopheles) are 
the intermediate hosts of this parasite, which may be found alive 
in the stomachs of these insects after they have fed upon a filarious 
individual. Ecdysis takes place in this organ, and the embryos, 
after having cast their sheaths, manage eventually to penetrate the 
thoracic muscles of their host, in which situation they undergo a 
developmental phase. This lasts about eighteen days, after which 
the larvae thus evolved escape from the thorax and ultimately 
reach the insect's labium and proboscis, whence they are directly 
introduced in the blood of man by the bite of the infected mosquito. 
The parasites having been inoculated into a human being in this 
manner, make their way to some part of the lymphatic system, in 
which they lodge, sexually mature, fecundate, and beget the 
innumerable embryo forms which, via the lymph stream, find 
their way into the circulating blood. 

Appearance in Fresh Blood. In the unstained blood film the 

1 Loc. cit. 2 Loc. cit. 3 Brit. Med. Jour., 1902, vol. i, p. 1472. 

4 i6 


parasite appears under the microscope as a long, slender, graceful 
worm possessing a most remarkable degree of activity. It measures 
about -gJ-0- of an inch in length and ^n77r f an mc h m diameter, 
and is of a pearly-gray color, with perhaps the faintest suggestion 
of a yellowish tone in certain lights. Its general appearance 
conveys to one, at first glance, the impression of a thin, transparent 
tube, through which a rapidly flowing stream of liquid is con- 
stantly circulating. The head (cephalic end) is gracefully rounded, 
while the tail (caudal end) gradually tapers for about one-sixth 
the entire length of the animal, and ends in a fine-pointed ex- 
tremity. The worm is cylin- 
drical in shape, of regular out- 
line, and consists of a central 
body enveloped in a distinct, 
loosely fitting, hyaline, struc- 
tureless sheath, which is about 
as much too large for the body 
as the thumb of an adult's 
glove would be for the little 
finger of a child. Thus, that 
part of the sheath temporarily 
unoccupied by the body is 
prone to collapse, folding upon 
itself and trailing after the 
worm at either or both ex- 
tremities as a twisted, whip- 
like ribbon. The greater part 
of the body appears to be of a 
homogeneous structure when 
examined in the freshly pre- 
pared slide, but after the speci- 
men has been kept for several 

hours, coarse granulations begin to stipple its surface, first develop- 
ing in the center and gradually spreading toward the periphery. 
(See Fig. 60.) A series of fine striations, like the milling on a 
coin, may be observed running along both edges of the body at 
right "angles to its long axis. A viscus, appearing as a mass of 
granular material, occupies a part of the central third of the worm's 
body, running parallel to its long axis. Upon careful examination 
with an oil-immersion objective rhythmical dimpling or puckering 
movements may generally be observed at the tip of the cephalic 
end of the embryo; these movements, which occur with more or 
less regularity at the rate of from twenty-five to forty times a 
minute, have been attributed to the act of respiration. As the 


Showing beginning granular degeneration of the 
body of the parasite in a fresh blood film. 


wriggling of the worm becomes less active close observation will 
show that these pouting movements are caused by the alternate 
covering and uncovering of the cephalic end by a delicate, six- 
lipped prepuce. The sudden projection and the equally rapid 
retraction of a filamentous fang or tongue-like organ from the 
worm's uncovered head may also be noted in some instances, but 
this characteristic is so difficult to make out that it may usually 
be looked for in vain. At a point about one-fifth of the entire 
length of the worm posterior to the head it is possible to make 
out a triangular, slightly luminous patch, shaped like the letter V, 
this spot being known as the V-shaped patch, regarded by Man- 
son as a rudimentary generative organ. A second spot, some- 
what similar to it in appearance but smaller in size, may occasion- 
ally be seen at a point just above the tail of the parasite; this 
spot Manson is inclined to regard as the rudimentary anus. 

The movements of the worm are v rapid and violent in the ex- 
treme; so much that they are followed with difficulty with any 
but a low-power dry objective. The parasite is never at rest: 
one moment it may be curled up into a tight bunch, like a coil 
of rope; the next moment it may suddenly straighten out and 
become rigid for an instant, only to resume its incessant con- 
tortions and twistings, which throw it into every conceivable 
shape. If particular attention is paid to the point, it will be 
noticed that, however rapid and complicated may be its move- 
ments, the parasite is never seen to turn completely over laterally. 
The accompanying series of sketches of the Filaria nocturna in 
a fresh blood slide illustrate a few of the different forms which 
this parasite may assume (Fig. 61). The worm seems to move 
about among the blood corpuscles with graceful and quick un- 
dulations of its body and abrupt whip-like strokes of its tail, 
butting its head against the more resisting masses of cells or else 
seeking a less difficult passage around them, always in motion, 
but never, it appears, with any definite aim to its exertions. Con- 
trary to the views expressed by most observers, that the movements 
of the Filaria nocturna are not truly propulsive in character, the 
writer has repeatedly noticed that this worm sometimes travels 
several times the distance of the diameter of the microscope field 
(&-inch objective, one-inch ocular, and 160 mm. tube-length), 
although in most cases its excursions were limited to a measured 
area not exceeding half a dozen square microns. It cannot be denied 
that these apparently progressive movements of the worm may pos- 
sibly be due to the currents in the blood plasma, but they certainly 
seem to have every characteristic of a true locomotive force. 
After -the slide has been kept for a few hours, the movements 


of the worm,' at first so confusingly rapid, gradually become 
slower and slower, and these torpid, more deliberate turnings and 
twistings may be accurately followed under an immersion-lens. 
If the parasite happens to become confined in a little pool of 
plasma surrounded by rouleaux of half-dried erythrocytes, an 
accident which often happens when the drying of the film has 
spread inward some little distance from the edges of the cover- 
glass^ its finer structure and characteristics may be studied with 
great ease and accuracy. 

9.*6 9 -fd 9 -SO 

954 956 9.58 /OOO 


/Q./O /O./2 



The sketches, made at two-minute intervals, all represent the same parasite. 

The phenomenon of ecdysis, or shedding of the worm's sheath, 
with 'the consequent escape of its naked body into the plasma, 
occurs when slides containing the live filariae are kept for some 
hours in a cold (not freezing) place. It commonly happens that 
just before the death of the parasite an occasional erythrocyte or 
leucocyte becomes tightly adherent to the sheath, swinging to 
and fro with the now lazy, torpid movements of the animal. 

Technic of Examination. A rather large drop of finger 
blood, taken from the patient late in the evening, preferably 


toward midnight, is placed between a slightly warmed slide and 
cover-glass, the edges of which are immediately sealed with 
cedar oil or with vaselin. The parasite should be searched for 
with a low-power dry objective, a f-inch lens being most useful 
for this purpose, and the attention of the examiner directed 
especially to portions of the field which may show any unnatural 
agitation of the blood cells. In specimens prepared in this man- 
ner the filariae will usually remain active for several days, gener- 
ally for at least forty-eight hours, and sometimes for a longer 
period, as in Henry's 1 experience, this author having kept them 
alive for ten days in a cold room. 

Staining the Filarm. Films fixed for fifteen minutes in equal 
parts of absolute alcohol and ether and stained with thionin 
give the clearest-cut pictures, the multitude of small nuclei which 
crowd the body of the filariae being sharply differentiated by the 
use of this dye. Fixation by heat or by formalin cannot be em- 
ployed without risk of injuring the finer structure of the embryo. 
Fair results may also be obtained by staining with methylene- 
blue or with Jenner's or Wright's stain, but the definition is not 
nearly so satisfactory with these solutions as it is with thionin. 
The technic suggested by Manson 2 (washing out the hemoglobin of 
the erythrocytes with water, drying, fixing in alcohol, and staining 
with methylene-blue or with hematoxylin) has proved unreliable 
in the writer's hands. The same comment may be made re- 
garding attempts to demonstrate the structure of the worm by 
staining with fuchsin, as has also been recommended. 

The presence in the circulation of the Filaria 
HEMOGLOBIN nocturna does not appear of itself to be a factor 
AND of any conspicuous changes in the erythrocytes 

ERYTHROCYTES. and their hemoglobin content. The high- 
grade anemia sometimes associated with filariasis, 
mentioned by the Bancrofts 3 and by Ehrlich and Lazarus, 4 is due, 
no doubt, to such complications as hematuria, severe chyluria, and 
chronic diarrhea. In two cases the writer found hemoglobin 
percentages of 85 and 88, and erythrpcyte counts of 4,200,000 
and 4,876,000 per, respectively. These figures may be 
taken as representative for the average case, judging from the 
limited data available. 

Neither structural changes, nor irregular staining affinities of 
the cells and the occurrence of nucleated erythrocytes have been 
reported in connection with the disease. 

1 Loc. cit. 2 Loc. at. 

3 Australasian Med. Gaz., 1894, vol. xiii, p. 6. *Loc. cit 


In the early stages of filariasis a distinct 
LEUCOCYTES, leucocytosis is generally found, but as the disease 
progresses the number of leucocytes gradually 
diminishes, and in cases of long standing the count does not exceed 
the physiological limits of health, except as the result of some 
complication. Thus, in the first case quoted above the leucocyte 
count was found to be 41,000 per, but this increase was re- 
garded purely as a post-operative rise, the patient having been 
operated upon for a supposed varicocele less than twenty-four 
hours before the blood examination was made. The count in the 
second case, one of several years' duration, was 8000. 

The relative percentage of mononuclear non-granular leuco- 
cytes is somewhat higher than normal, with a consequent decrease 
in the proportion of polynuclear neutrophiles. The eosinophiles 
either remain at a maximum normal percentage or may be dis- 
tinctly in excess of this figure. This statement, made by the writer 
in 1901, has since been verified by several observers, notably by 
Calvert, 1 Gulland, 2 Coles, 3 Vaquez, 4 Sicard, 5 and Clerc. 8 Calvert 
has shown that the eosinophilia, like the leucocytosis, diminishes 
as the infection becomes chronic, and that its development is 
cyclical, in that it follows by a few hours the periodicity of the 
embryo worms in the peripheral blood. In the author's two cases 
the percentage of eosinophiles ranged from 3.4 to 9.5; in Gulland's 
case, from 3.0 to 12.0; in Calvert's three cases, from 6.0 to 22.0; 
and in Coles' two cases, from 15.0 to 17.0. The three French 
authors mentioned above report eosinophilia varying from 7.5 to 
12 per cent. It may be noted here that the Filaria loa, found in 
the subconjunctival tissues, may also excite eosinophilia 53 per 
cent, in a case reported by Wuntz and Clerc 7 ; and that in patients 
harboring the guinea-worm a similar eosinophile increase de- 
velops as high as 36.6 per cent, in cases studied by Balfour. 8 
In six cases of guinea- worm infection Powell 9 found eosinophile 
percentages of 4.7, 5.5, 7.5, 7.5, 8, and 12.2, respectively. The 
lymphocytes frequently appear as cells having a deeply stained 
eccentric nucleus surrounded by an abnormally large area of 
protoplasm, the general appearance of these cells being similar 
to those in the illustration shown on page 320, Fig. 55. Typical 
coarsely granular mast cells may be found in small numbers 
or they may be entirely absent, as may also be the finely gran- 

1 Johns Hopkins Hosp. Bull., 1902, vol. xiii, pp. 23 and 133; also Jour. Amer. 
Med. Assoc., 1902, vol. xxxix, p. 1523. 

2 Brit. Med. Jour., 1902, vol. i, p. 831. 3 Ibid., 1902, vol. i, p. 1137. 
4 Sem. me'd., 1902, vol. xxii, p. 418. 5 Loc. cit. 

8 Loc. cit. 7 Sem. med., 1903, vol. xxiii, p. 420. 

8 Lancet, 1903, vol. ii, p. 1649. 9 Brit. Med. Jour., 1904, vol. i, p. 73. 


ular forms of basophiles. The presence of myelocytes has not 
been noted. 

The detection of the Filaria nocturna in the 
DIAGNOSIS, blood serves at once to differentiate idiopathic 
from parasitic chyluria, hydrocele from lymph 
scrotum, hernia and other tumors of the groin from parasitic 
inguinal varicosities (Bancroft's "helminthoma elastica"), and 
filarial orchitis from other inflammatory conditions of the testes. 
N on- parasitic lymphedema, affecting, for example, the legs, can 
but rarely be distinguished by the blood findings from true elephan- 
tiasis Arabum, since in the latter disease it is exceptional to find 
filariae in the general circulation. 


Blake, Hubbard, and Cabot 1 conclude, from a study of 38 
cases, that in simple uncomplicated fractures the number of leu- 
cocytes is seldom increased to any extent, a statement which applies 
also to complicated fractures in the great majority of instances. 
Of 23 simple fractures examined by these authors, in but 10 was 
the count higher than 10,50x3 per, and of these, only 6 
exceeded 12,000. The highest estimate was 15,400, in a fracture 
of the pelvis, and the next highest, 14,800, in a broken leg. Of 
15 complicated fractures, but 2 showed any decided increase 
in the number of leucocytes, namely, a fracture of the tibia and 
fibula, with symptoms suggestive of fat embolism, in which the 
count was 15,600; and a case of fractured ribs with injury of the 
lung, in which the leucocytes numbered 14,900 two days after 
the accident. An estimate of 5400 cells was made in a compound 
fracture of the leg two hours after the accident. 

Lipemia is occasionally met with in fractures of the long bones 
involving injury of the fatty marrow. 


In the acute form there is no deviation from 
HEMOGLOBIN normal in the number and hemoglobin value of 
AND the erythrocytes, except in the event of hyper- 

ERYTHROCYTES. emesis, which, through concentration of the blood, 
may cause a transient polycythemia. In the 
chronic form secondary anemia frequently develops, and occasion- 
ally reaches an extreme grade, should the gastric lesion be suffi- 

1 Annals of Surg., 1901, vol. xxxiv, p. 361. 



cient to interfere radically with the digestion and absorption of 
food. In instances of this sort the quantitative changes may 
simulate those of true pernicious anemia, but the qualitative 
changes typical of this disease are invariably wanting. In pass- 
ing, it seems pertinent to recall the reputed etiological relationship, 
distinguished by some authorities, between gastric tubule atrophy 
and pernicious anemia. Well-defined secondary anemia (without 
erythroblasts) was found by Einhorn 1 in but 4 of 15 cases of gastric 
achylia, but in none were the qualitative blood changes of perni- 
cious anemia detected. In cases associated with gastrectasis 
and hyperacidity, blood inspissation from emesis is a common 

A synopsis of J. A. Lichty's studies 2 of the hemoglobin and 
erythrocytes in 98 cases of various gastric disorders shows the 
following average values: 






Hyperchlorhydria . . 39 



Hypochlorhydria 13 



Gastric achylia .... 6 



Gastric dilatation . . 1 1 



Gastric neurasthenia 




Chronic gastritis. . . . 


91.0 5,498,000 

From other investigations, Lichty also determined that in the 
above-named diseases there is no definite relationship between 
the condition of the blood, the urine, and the gastric contents. 

In gastroptosis the blood remains normal in the great majority 

of patients; in small proportion either a mild secondary anemia 

or a chlorotic blood picture develops. Francine 3 found the latter 

change in 3 of 100 cases. Of 55 patients studied by Steele 

and Francine, 4 the color index averaged 0.95 in 24, 0.9 in 20, and 

fell as low as 0.50 in but a single instance. The theory of Meinert, 5 

that chlorosis is a factor of gastroptosis, has no foundation in fact. 

. : In acute gastritis leucocytosis of the poly- 

LEUCOCYTES. nuclear neutrophile type is common, although 

not constant; the increase is most notable in the 

severest cases, but even in these the count seldom exceeds 15,000 

or 20,000. Hyperinosis also usually exists. In chronic cases 

1 Med. Rec., 1903, vol. Ixiii, p. 321. 2 Phila. Med. Jour., 1899, vol. iii, p. 326. 

3 Proc. Phila. Co. Med. Soc., 1902, vol. xxiii, p. 447. 

4 Jour. Amer. Med. Assoc., 1902, vol. xxxix, p. 1173. 

5 "Zur Aetiologie der Chlorose," Wiesbaden, 1894. 


an absence of leucocytosis is the rule, while leucopenia, resulting 
from defective absorption, is an occasional finding. Relative 
lymphocytosis is commonly associated with leucopenia and 
sometimes with normal leucocyte counts. In a small proportion 
of cases digestion leucocytosis is either delayed or absent. 

The presence of a leucocytosis is a valuable 

DIAGNOSIS, sign in ruling out enteric fever, should the diag- 
nosis lie between this disease and acute febrile 
gastritis. This sign, however, cannot be employed to differentiate 
other acute infections, such, for instance, as appendicitis. 

The blood furnishes no sure means of differentiating chronic 
gastritis from gastric cancer, although a persistent leucocytosis is 
very suggestive of the latter; unfortunately, digestion leucocytosis 
is neither constantly absent in cancer nor invariably present in gas- 

Certain cases of chronic gastric catarrh, with atrophy of the 
stomach tubules, in course of time "develop a clinical picture very 
like that of true pernicious anemia, since they present not only a 
similar cachexia, but also a very striking diminution in hemo- 
globin and erythrocytes. But pernicious anemia is characterized 
by the presence of nucleated erythrocytes the majority of which 
are megaloblasts, while in the secondary anemia of gastric catarrh 
erythroblasts are uncommon, and if present, show a predominance 
of cells of the normoblastic type. 


The average case shows a loss of approxi- 

HEMOGLOBIN mately 40 per cent, of hemoglobin and of 

AND 1,250,00x3 erythrocytes to the, and, owing 

ERYTHROCYTES. to this prevalence of a disproportionately large 

oligochromemia, low color indices are the rule. 

The individual case may show a much greater degree of anemia, 

but no matter how marked the cellular decrease, it is always far 

outstripped by the diminution in the percentage of hemoglobin. 

In fact, in some instances the latter alone is subnormal, the blood 

condition of chlorosis being thus faithfully counterfeited. The 

average index for the cases tabulated below was 0.75. 

Profuse hemorrhage may provoke a very marked anemia, while 
protracted emesis tends to concentrate the blood, thus masking 
its real condition. 

The several degenerative changes affecting the erythrocytes 
common to any severe anemia may be present if the blood de- 


terioration is sufficiently profound. After a severe hemorrhage 
a few normoblasts not infrequently appear in the blood tempo- 
rarily, and an occasional cell of this type may be found at other 
times in cases with marked cachexia. 

The following summary illustrates the hemoglobin and eryth- 
rocyte ranges in 33 cases: 



From 80-90 5 Above 5,000,0000 2 

" 70-80 5 From 4,000,000-5,000,000. . . 16 

" 60-70 7 " 3,000,000-4,000,000... 7 

" 50-60 i " 2,000,000-3,000,000 7 

" 40-50 6 " 1,000,000-2,000,000... i 

" 3-4o 5 

20-30 4 

Average, 57 per cent. Average, 3,798,000 per 

Maximum, 89 " " Maximum, 5,200,000 " " 

Minimum, 20 " " Minimum, 1,090,000 " " 

Futcher l gives the following data of 82 cases : the hemoglobin 
percentage in 42 cases averaged 58, ranging from 12 to 105; 
the erythrocyte count in 44 cases averaged 4,071,000, or from 
1,012,000 to 4,071,000; and the leucocyte count in 45 cases 
averaged 7500, varying from noo to 40,000. 

Greenough and Joslin 2 report hemoglobin estimates in 73 cases, 
of which 34 were below 50 per cent, and 64 below 80 per cent. 
Of their 43 erythrocyte counts, 24 were below 4,000,000 per, the color index for this series averaging 0.67, and ranging 
from 0.35 to 1.41. 

Absence of leucocytosis is the rule, for an 
LEUCOCYTES, increase occurs only after taking food or in the 
event of some complication, such as hemor- 
rhage or perforation. But the fact must be recalled that hemor- 
rhage by no means invariably raises the count; for example, hem- 
atemesis is a symptom in fully 50 per cent, of patients suffering 
from ulcer of the stomach, yet in not more than 30 per cent, of 
all cases, both those with and those without this symptom, does 
the number of leucocytes exceed 10,000 to the Perforation 
always excites leucocytosis, except when the patient is over- 
whelmingly toxic. 

1 Amer. Med., 1904, vol. viii, p. 53. 

2 Amer. Jour. Med. Sci., 1899, vol. cxviii, p. 167. 


The behavior of the leucocytes in the above-mentioned series 
of cases may be expressed thus: 


Above 20,000 2 

From 15,000-20,000 2 

" 10,000-15,000 6 

" 5,000-10,000 18 

Below 5,000 5 

Average, 8,778 per 

Maximum, 29,400 " 

Minimum, 2,400 " 

In cases with leucocytosis the increase affects chiefly the poly- 
nuclear neutrophiles ; in those without leucocytosis minimum nor- 
mal or distinctly subnormal percentages of these cells are not 
uncommon, and a total absence of eosinophiles is th6 general rule, 
these changes being counterbalanced by a proportionate increase 
in the small lymphocytes. 

Hematology gives no aid in distinguishing gas- 

DIAGNOSIS. trie ulcer from gastf algid, duodenal ulcer, and 

simple gall-stone colic, in all of which leucocytosis 

is absent. A well-defined leucocytosis points to acute gastritis 

rather than to ulcer. The differences in the blood pictures of 

gastric ulcer and cancer are referred to under the latter disease. 

(See " Malignant Disease.") > 


Data are wanting regarding the condition of the hemoglobin and 
erythrocytes in human glanders, but it is known that leucocytosis is 
the rule. The Bacillus mallei has been obtained by antemortem 
blood culturing and by Duval 1 and by von Jaksch. 2 Heanley 3 
claims that, with a dilution of i 12500 and a time limit of twelve 
hours, a specific serum reaction occurs with the bacillus of glan- 
ders and glanders blood serum. In lower dilutions similar results 
may be obtained with the serum of patients suffering from variola 
and scarlatina. 

1 Arch, de m<5d. exper., 1896, vol. viii, p. 361. 
1 "Clinical Diagnosis," 4th ed., London, 1899. 
8 Lancet, 1904, vol. i, p. 364. 



The hemoglobin and erythrocytes are unaltered, but the acute 
febrile stage of specific urethritis is usually accompanied by a 
moderate polynuclear leucocytosis, which, in the event of any of 
the inflammatory complications of clap, may be much aggravated. 
Sabraze*s 1 found that the increase usually does not exceed double 
the mean average normal count. Giorgi, 2 contrary to general 
opinion, found an absence of leucocytosis, sometimes leucopenia, 
the rule, together with a relative mononucleosis, at the expense 
of the polynuclear neutrophiles. Some authors formerly claimed 
that circulatory eosinophilia was a feature of this disease, but the 
investigations of Vorbach 3 have shown that such a change, while 
occurring sometimes, is by no means constant; he found in 20 
cases that the percentage of eosinophiles ranged from as low as 
0.05 to as high as 11.5. Bettmann 4 believes that eosinophilia is 
especially frequent in posterior urethritis, an observation which 
thus far is unique. 

In gonorrheal endocarditis and in gonorrheal arthritis the 
gonococcus has been repeatedly cultured from the blood during 
life. In distinguishing gonorrheal arthritis from rheumatic fever a 
positive iodin reaction is suggestive of the former. 


Garrod's 5 earlier teachings regarding the lowered alkalinity of 
the blood in acute gout have been contradicted, apparently with 
ample proof, by the later researches of Levy, 6 who failed to find 
a diminution in any of the 17 cases which he investigated by the 
most approved methods. Still more recently Levy's conclusions 
have been corroborated by Watson. 7 During an acute gouty 
seizure hyperinosis is an almost invariable finding. 

It is questionable whether or not the amount of uric acid in the 
blood is greater during the acute stages than in the interval be- 
tween them, but it is nevertheless a fact that in many gouty persons 
uric acid crystals can be demonstrated in the blood by the " thread- 
test" a reaction by no means peculiar to this disease, as already 
pointed out. (Seep. 144.) 

1 Sem. med., 1902, vol. xxii, p. 435. 

2 Ibid., 1904, vol. xxiv, p. 39. 3 Inaug. Dissert., Wurzburg, 1895. 
4 Arch. f. Dermat. u. Syph., 1899, vol. xxxix, p. 227. 

1 "Gout and Rheumatic Gout," London, 1876, p. 80. 
8 Zeitschr. f. klin. Med., 1898, vol. xxxvi, p. 336. 
7 Brit. Med. Jour., 1900, vol. i, p. 10. 


The cellular elements show no characteristic alterations, and 
are normal, except in long-standing cases, in which an ordinary 
secondary anemia may develop in the course of time. During 
the acute attack a moderate increase in the number of leucocytes, 
affecting chiefly the polynuclear neutrophiles, may or may not be 
found. A relative increase in the eosinophiles is also sometimes 
encountered, in cases both with and without an increase in the 
leucocyte count. In one case of the writer's the blood examined 
during the height of a severe paroxysm showed 100 per cent, of 
hemoglobin, 7,125,000 erythrocytes, and 14,000 leucocytes per, the only peculiar differential change being the presence