THE LIBRARY OF THE UNIVERSITY OF CALIFORNIA PRESENTED BY PROF. CHARLES A. KOFOID AND MRS. PRUDENCE W. KOFOID A GUIDE TO THE CLINICAL EXAMINATION OF THE BLOOD FOR DIAGNOSTIC PURPOSES BY EICHAED C. CABOT, M.D. WITH COLORED PLATES AND ENGRAVINGS Ubirfc 1Rex>fsefc Edition NEW YORK WILLIAM WOOD AND COMPANY 1898 COPYRIGHT, 1898, BY WILLIAM WOOD AND COMPANY. TO WILLIAM SIDNEY THAYEE, M.D., ASSOCIATE PROFESSOR OF MEDICINE IN JOHNS HOPKINS UNIVERSITY, IN GRATEFUL RECOGNITION OF THE STANDARD OF THOROUGH WORK ESTABLISHED BY HIM. PREFACE TO THE THIRD EDITION. ABOUT forty pages of new matter have been added, but the book is no larger than before, as a corresponding number of pages have been omitted from the bibliography. A complete bibliography of the subject would now need a separate volume, and it has therefore seemed best to omit all but the most im- portant references. The principal additions to the book include an account of Professor Oliver's tintometer and haemoglobinometer (which are the only new instruments of importance), new matter in the chapter on the primary anaemias, and on leukaemia, and a de- scription of Miiller's "blood-dust" (the newly discovered con- stituent of normal and abnormal blood). Bremer's and Wil- liamson's tests for diabetic blood and the iodine reaction in the blood during acute suppurative processes are described, and blood exminations are recorded in Malta fever, yellow fever, epidemic dropsy, beri-beri, relapsing fever, tetanus, chicken- pox, whooping-cough, and epidemic cerebro-spinal meningitis, diseases not included in previous editions. New observations on poisoning by alcohol, opium, corrosives, and ptomains, on aneurisms, on paroxysmal haemoglobinsemia, and on cretinism are recorded. I have wished to draw especial attention to the tendency to an oval or sausage shape among the red corpuscles in cases of grave anaemia and to the occurrence of adventitious forms of leucocytes in leucocytosis. Some of my critics have regretted that there was so little theoretical discussion in the book. In this edition I have for the most part eliminated what little there was before. It had already become out of date. But though I cannot see my way to change the general plan of my book, there has been hardly any other suggestion of my critics which has failed to help me, I am especially indebted to those who have been good enough VI PREFACE TO THE THIRD EDITION. to send me detailed criticisms, and in almost every instance I have been glad to make the changes suggested. For such help I owe thanks above all to Dr. J, B. Herrick of Chicago, Dr. J. Ewing of New York, and Dr. Greene of the Marine Hospital Service. It is impossible to keep such a book as this up to date. While the sheets of this edition have been passing through the press, three important works upon the blood have appeared, viz.: Ehrlich and Lazarus' "Die Anaemie," Turk's "The Blood in Acute Infectious Diseases," and Coles' "The Blood." I have greatly profited by reading these books, but unfortu- nately have been unable to incorporate their observations in the present edition. 190 MARLBOROUGH STREET, BOSTON. July, 1898. PREFACE. IN order to keep the size of this book within reasonable lim- its I have omitted all historical account of the steps by which our present knowledge of the different branches of the subject has been built up. Wherever it has seemed to me that a point was definitely established, I have stated the conclusions generally accepted without special reference to the names of those who worked them out. On the other hand, where our knowledge has seemed to be insufficient I have given some of the names and findings of those who are responsible for the opinions generally current. Theoretical discussions have been omitted on account of the strictly clinical plan of the book. The absence of any account of the origin of the blood cells, the chemistry of the blood, coagulation, and many other sub- jects of great scientific interest is due to the lack of any con- siderable clinical value in them so far as at present understood. The body of data referred to from time to time as the " Mas- sachusetts Hospital Blood Counts" consists of nearly four thou- sand blood examinations, about three thousand of which were made by Drs. Moffitt, Hewes, Joslin, Denny, Franklin White, Capps, and Barney medical internes of the hospital since 1893. Permission to avail myself of these data was very kindly granted me by the visiting physicians of the hospital. To these I have added about one thousand examinations which I have made both within the hospital and outside. The technique used in all the four thousand examinations was essentially that described in the following pages. The accumulation of this body of facts and the great mass of foreign hsematological literature (untranslated) have seemed to me sufficient reasons for the existence of this book the first of its kind in English, so far as I am aware. Further, it has seemed to me a great pity that there should be no book available con- VI 11 PREFACE. taining colored illustrations of stained blood preparations which bear some resemblance to their original, and are not wholly or partly works of the imagination ("diagrammatic"). Funke, of Leipsic, has, I think, been as successful in dealing with the stained blood in the present work as he was with the fresh blood in the beautiful illustrations for W. S. Thayer's monograph on "The Malarial Fevers of Baltimore." Any one who writes on the blood must be constantly indebted to the following standard text-books: Hay em: "Du Sang," Paris, 1889. v. Limbeck : " Grundriss einer klinischen Patho- logie des Blutes," Jena, 1896 (second edition). Grawitz: "Klinische Pathologie des Blutes," Berlin, 1896. Schmaltz: "Pathologie des Blutes," Leipsic, 1896. Eieder: "Beitrage z. Kentniss der Leucocytosis," Leipsic, 1892. I have usually referred to them in the text as " Hay em, " "Eieder," etc., always meaning one of the above works. The quotations from Schreiber in the text refer to manuscript notes of his lectures in 1896, kindly loaned to me by Dr. Mark W. Bichardson. I am indebted to Dr. F. P. Henry for permission to use the cuts from his recent article on the filaria sanguinis hominis. December, 1896. PREFACE TO THE SECOND EDITION. A NEW chapter on the serum reaction in typhoid fever has been added, and the more obvious mistakes in the text have been corrected. I wish to express my thanks to those who have called my attention to mistakes in the text, especially to Dr. Joseph A. Capps, who has furnished many valuable sugges- tions. TABLE OF CONTENTS. BOOK I. Introduction. PAGE SCOPE AND VALUE OF BLOOD EXAMINATION, . . . . . .3 PAKT I. METHODS OF CLINICAL EXAMINATION OF THE BLOOD* CHAPTER 1. EXAMINATION OP THE FRESH BLOOD, . . . . . . .5 (a) Puncturing, . . 5 (6) Spreading the Blood, . . 7 (c) Prevention of Cell Death, 8 (d) Knowledge to he Gained from Fresh Blood 9 CHAPTER II. COUNTING THE CORPUSCLES, 11 (a) Sucking Up the Blood, 11 (&) Diluting the Blood, 12 (c) Adjusting the Diluted Blood in the Counting Chamber, . 14 (d) Counting the Red Corpuscles, . . . . . . .15 (e) Counting the White Corpuscles, 18 (/) Counting both Red and White Corpuscles with One Pipette, . 19 (g) Durham's Hsemocytometer, ....... 22 CHAPTER III. OLIVER'S TINTOMETER CENTRIFUGALIZING THE BLOOD HAEMOGLOBIN ESTIMATION SPECIFIC GRAVITY STAINED SPECIMENS BAC- TERIOLOGICAL EXAMINATION. 1. Oliver's Tintometer, 26 2. Daland's Haematocrit, Its Objects, ,29 Use of the Haematocrit, 30 3. Use of v. Fleischl's Haemometer, 33 Necessary Errors in Its Use, ... .37 4. Oliver's Haemoglobinoineter, . . 38 TABLE OF CONTENTS. PAGE 5. Specific Gravity Estimation, . . . . . . .40 Hammerschlag's Method, .....,. 41 6. Study of Dried and Stained Specimens, . . . . .43 (a) Preparation of Cover-Glass Specimens, . . . .43 (6) Fixing Cover-Glass Specimens, 43 (c) Staining, . . . . . .... .44 (d) Differential Counting (color analysis), . . . .46 7. Bacteriological Examination, 47 8. Other Methods of Examination, 48 PAKT H. PHYSIOLOGY or THE BLOOD. CHAPTER IV. MORPHOLOGY OP FRESH BLOOD, 50 (a) Appearances of Normal Red Corpuscles, 50 (b) Appearances of Normal White Corpuscles 52 (c) Appearances of Normal Blood Plates 53 (d) Appearances of Fibrin Network, . . . . . .53 (e) Average Diameter of Red Cells 55 (/) Normal Number of Red Cells, . . 56 Influence of Menstruation, Parturition, Lactation, . . 56 Influence of Vasomotor Changes, . . . . . .57 (g) Influence of Nutrition on the Red Cells, 57 Influence of Fatigue on the Red Cells, 58 (h) Normal Number of White Cells 58 (i) Normal Number of Blood Plates, 59 (j) Muller's Blood-dust, 59 CHAPTER V. FINER STRUCTURE OF THE BLOOD, . . . . . .61 1. Finer Structure of Red Cells, 61 2. Finer Structure of White Cells and Their Varieties, (a) Lymphocytes, 62 (6) Transitional Cells, 63 (c) Polymorphonuclear Neutrophiles, 64 (d) Eosinophiles, ...... 65 (e) Basophiles, 66 Terminology, 67 Normal Percentage of Each Variety, . . . .67 (/) Myelocytes, 69 Eosinophilic Myelocytes, ^ Cornil's "Markcellen," 71 TABLE OF CONTENTS. XI PAET III GENERAL PATHOLOGY OF THE BLOOD. CHAPTER VI. PAGE UNEQUAL DISTRIBUTION OF THE BLOOD PLETHORA DILUTION AND CONCENTRATION OF THE BLOOD, 73 1. Unequal Distribution, . . . . . . .73 2. Apparent Polycythaemia, 73 (a) General Cyanosis, . . . . . . .74 (&) Local Cyanosis, 74 (c) Feeble Circulation, 75 3. Plethora Serous or Cellular, 75 4. Concentration of the Blood, . . . . . . .76 5. Dilution of the Blood, 78 6. The Blood in High Altitudes, 79 7. Phosphorus and CO Poisoning, 80 8. Possibility of a True Plethora, 80 CHAPTER VII. ANAEMIA AND HYDR^EMIA, 82 (a) Pallor and Anaemia, 82 (6) "Tropical Anaemia," . .83 (c) Distinction Between Primary and Secondary Anaemia, . 84 Secondary Anaemia, 85 1. First Stage Loss of Color, Weight, and Size, . . . 85 2. Second Stage Poikilocytosis and Degeneration, . . 85 (a) Endoglobular Changes, 86 (6) Crenation and Deformities ; Motility, . . .86 (c) Oval Shape, 87 (d) Changes in Staining Properties, . ... 88 (e) Loss of Haemoglobin, .... ... 88 3. Third Stage Deglobularization, 89 4. Nucleated Red Corpuscles, 89 (a) Normoblasts, . 90 (b) Megaloblasts, 91 (c) Microblasts, 92 (d) Atypical Varieties, 92 Summary, 94 Hydraemia, 95 CHAPTER VIII. LEUCOCYTOSIS, LYMPHOCYTOSIS, EOSINOPHILIA, AND MYELOCYTES, . 96 Definition of Leucocytosis, 96 Physiological Leucocytosis, . . ... . . .98 Xll TABLE OF CONTENTS. PAGE (a) Effects of Nutrition and Starvation, 99 (&) Digestion Leucocytosis, . ... 99 Its Diagnostic Value, ....... 100 (c) Leucocytosis of the New-Born, 101 (d) Leucocytosis of Pregnancy, 101 (e) Leucocytosis after Parturition, 102 (/) Leucocytosis from Exercise, Massage, and Baths, . . 102 (g) Terminal Leucocytosis, 105 Pathological Leucocytosis, 106 (a) Post-hemorrhagic Leucocytosis, ..... 106 (6) Inflammatory Leucocytosis 106 (c) Toxic Leucocytosis, 109 (d) Leucocytosis due to Therapeutic and Experimental In- fluences, 110 (e) Morphology of the Leucocytes in Leucocytosis, . . 112 Absence of Leucocytosis, 113 Leucopenia, . 113 Lymphocytosis, 114 Diagnostic Value of Lymphocytosis, ... 116 Eosinophilia, ......... 116 Diminution of Eosinophiles, 119 Diagnostic and Prognostic Value of Eosinophilia, . . 119 Myelocytes, . 120 CHAPTER IX. GENERAL PATHOLOGY OP HAEMOGLOBIN AND FIBRIN, LIP^EMIA, MEL- ANAEMIA AND HEMORRHAGE, .123 1< Haemoglobin and the "Color Index," 123 2. Fibrin, 124 3. Lipaemia, ........... 125 4. Melanaemia, 126 5. Hemorrhage, 126 (a) Changes in the Blood Resulting from Hemorrhage, . . 126 (&) Blood Regeneration, . . . . . . . .127 Regeneration of Red Cells, 127 Blood Crisis, 128 Regeneration of White Cells, 128 Importance for Surgery of Blood Counting after Hemorrhage, . 129 Chronic Hemorrhage 130 TABLE OF CONTENTS. Xlll BOOK II. Special Pathology of the Blood. PART I. DISEASES OF THE BLOOD. CHAPTER I. PAGE FHE PRIMARY ANEMIAS, 1. The Blood in Pernicious Anaemia, 133-151 1. Gross Appearances, 2. Appearances of Fresh Blood, ..... 133 Red Cells and Haemoglobin, Quantitative Changes, 134-137 White Cells, 137 Quantitative Changes, 137 Blood Plates and Fibrin, ....... 138 Haemoglobin, 138 Qualitative Changes 139 1. Red Corpuscles, 139 (a) Increase in Diameter, 139 (6) Deformities in Shape, 140 (c) Staining Properties, 141 (d) Nucleated Red Corpuscles, .... 141-144 2. White Corpuscles, . .... 144-147 Characteristics of Pernicious Anaemia, Summary, . . . 146 Differential Diagnosis of Pernicious Anaemia, . 147 1. Pernicious Anaemia and Chlorosis, . .147 2. Pernicious Anaemia and Malignant Disease, , . 147 3. Pernicious Anaemia and other Secondary Anaemias, . 148 4. Pernicious Anaemia and Leukaemia, .... 148 Prognostic Value of the Blood in Pernicious Anaemia, . . 149 2 The Blood in Chlorosis 152 The Blocd in Gross, 152 1. Red Cells and Haemoglobin, 152 (a) Quantitative Changes, . . . .152 (&) Qualitative Changes 155 2. Specific Gravity, 156 3. White Cells 15 6 (a) Quantitative Changes, 156 (6) Qualitative Changes 156 4. Blood Regeneration in Chlorosis, .... 5. Chlorosis without Known Blood Changes, . . 157 6. Summary and Diagnostic Value, . . . . 157 3. Other Forms of Anaemia, ...... 158 XIV TABLE OF CONTENTS. CHAPTER II. PAGE LEUKAEMIA AND HODGKIN'S DISEASE, . . . . . . .160 1. Myelocytsemia, ...... . . . 161 1. Red Cells, . 161 (a) Quantitative Changes, 161 (6) Qualitative Changes, 162 2. White Cells, 163 (A) Quantitative Changes, . . . .'."". . 163 (B) Qualitative Changes, .165 (a) Myelocytes, 165 (6) Polymorphonuclear Cells 166 (c) Lymphocytes, 167 (d) Eosinophiles, . 167 (e) Basophiles, 168 (f) Polymorphous Condition of the Blood, . . 169 2. Lymphaemia, 169 1. Red Cells, 170 2. White cells, 170 (a) Quantitative Changes, 170 (&) Qualitative Changes, 171 Summary of Blood Changes in Leukaemia, .... 173 Differential Diagnosis of Leukaemic Blood, .... 174 (a) From Hodgkin's Disease, 174 (5) From Tumors in or near Spleen, . . . . . 174 (c) From Syphilitic or Tubercular Adenitis, . . . 175 (d) From Hydronephrosis, ...... 175 (e) From Leucocytosis, 175 (/) From Chronic Malaria or Amyloid Diser.se, . . 175 Table of Differential Diagnosis in Leukaemia, . . . 175 Effect of Intel-current Infections in Leukaemia, . . . 176 Hodgkin's Disease . 177 (a) Red Cells, 179 (6) White Cells, 181 (c) Summary and Diagnostic Value, . . . 181-182 PAET II. ACUTE INFECTIOUS DISEASES. INTRODUCTION. EFFECTS OF FEVER ON THE BLOOD, . . 188 CHAPTER III. PNEUMONIA, TYPHOID, AND DIPHTHERIA, 184 I. Pneumonia, . 184 ' 1. (a) Bacteriology of the Blood. . . .184 TABLE OF CONTENTS. ^V PAGE (&) Coagulation and Fibrin, ^ 5 (c) Concentration of the Blood, , ... 185 2. Red Cells, ... 185 3. White Cells, 185 (a) Quantitative Changes 185 (b) Qualitative Changes 187 4. Diagnosis and Prognosis, 1' II. Typhoid Fever, > 1! 1. Bacteriology and the Serum . . 1! 2. Red Cells and Haemoglobin, 191 3. White Cells, 193 4. Effect of Complications on White Cells, . . . .194 5. Qualitative Changes in the White Cells, . . . .195 6. Summary and Diagnostic Value, 196 III. Diphtheria, 197 (a) Red Cells l! (&) Haemoglobin, 1 (c) White Cells, 199 (d) Summary . . 201 CHAPTER IV. ACUTE INFECTIOUS DISEASES (Continued). I. Scarlet Fever, 203 (a) Red Cells, 203 (6) White Cells, 203 (c) Summary and Diagnostic Value 204 II. Measles, Rotheln, and Mumps, 205 III. Whooping-Cough, 306 IV. Small-pox (Variola), V. Chicken-pox, VI. Acute Articular Rheumatism, (a) Fibrin, Alkalinity, Red Cells, . 307 (6) White Cells in Acute Forms, .... (c) White Cells in Subacute Forms, (d) White Cells in Chronic and Muscular Forms, . (e) Summary and Diagnostic Value, VII. Asiatic Cholera, VIII. Erysipelas, .... IX. Tonsillitis X. Grippe XI. Septirsemia, (a) Bacteriology of the Blood, (b) Red Cells, . . (c) White Cells, (d) Summary and Diagnostic Value, XII. A. Abscess, Iodine Reaction in, .... XVI TABLE OF CONTENTS. PAGE B. Appendicitis, ... 223 (a) Leucocytosis, . . . . . . . 223-229 (6) Significance of the Absence of Leucocytosis, . . 227 (c) Differential Diagnosis, 229 C. Pus Tube, Pelvic Abscess, and Pelvic Peritonitis, . . 231 Differential Diagnosis .232 D. Otitis Media, 233 E. Osteomyelitis, 234 F. Other Abscesses, . . 234 Diagnostic Value, 235 XIII. Gonorrhoea, 236 XIV. Yellow Fever, . .236 XV. Typhus Fever, 237 XVI. Malta Fever, 238 XVII. Relapsing Fever 238 XVIII. Glanders, .238 XIX. The Bubonic Plague, . . 238 XX. Actinoraycosis, ......... 239 XXI. Trichinosis, ... ... 239 XXII. Epidemic Dropsy, 240 XXIII. Tetanus, 241 XXIV. Beri-beri, . . 241 CHAPTER V. DISEASES AFFECTING THE SEROUS MEMBRANES, ..... 242 I. Serous Pleurisy, .242 (a) Summary and Diagnostic Value, . . . . 244 (6) Purulent Pleurisy (Empyema) , . . . 244 II. Peritonitis, .245 Diagnostic Value, ... .246 III. Pericarditis (with Effusion), .... .247 Diagnostic Value, .... .248 IV. Meningitis, . . 248 Epidemic Cerebro-Spinal Meningitis, . 249 Diagnostic Value, ....... 251 PART III. CHRONIC INFECTIOUS DISEASES. CHAPTER VI. TUBERCULOSIS, SYPHILIS, AND LEPROSY, 253 I. Tuberculosis, 253 1. Red Cells and Heemoglobin, ... .253 (a) Quantitative Changes, ...... 253 TABLE OF CONTENTS. xvii PAGE (6) Qualitative Changes, . . 254 2. Leucocytes, .... . 255 (a) Changes in Phthisis, ' . . 255 (b) Changes in Bone Tuberculosis and Cold Abscess, . 259 (c) Changes in Acute Miliary Tuberculosis, . . . 261 (d) Changes in Tubercular Peritonitis, .... 264 (^ cover-glass face downward upon a / / slide so that the force of the impact """- -*^--^ will help to spread the drop of blood r thinly and evenly between slide and \L ) cover. It is recommended by Ehrlich and others to hold the cover-glass FlG - ^--^oper Method of HOW- .,, , , ,, , . ing a Cover Glass. with iorceps, but there is no harm in holding it with the fingers, provided we avoid touching either of its surfaces, i.e., hold it always as in Fig. I. 1 Slide and cover must be perfectly clean, else the blood will not spread out in a layer thin enough to avoid the corpuscles overlying each other so that not one of them is clearly seen. Further, as dirt simulates fairly closely some of the pathological appearances for which we are on the lookout, its presence on the slide leads to loss of time or to mistaken conclusions. Cover- glasses, as they come from the shops, may be coated with a substance not easily to be removed. To get them really clean 1 1 am not unmindful of Ehrlich's warning that the moisture of the fingers spoils the specimen ; but in practice I do not find it to be true ex- cept as regards the margin of the film, the good preservation of which is not essential. CLINICAL BLOOD EXAMINATION. nothing is so simple as or more effective than soap and water. After several years' use of the method of cleaning usually ad- vised (viz., strong mineral acid, followed by alcohol and then by ether), I have become converted to the use of plain soap and water as the best and simplest way of cleaning slides or cover- glasses. Kub soap over every part of the glass, wash it off thor- oughly with water, and polish it with a clean handkerchief (most towels are apt to leave a scrap of lint on the glass) . ' If slide and cover are perfectly clean, are held as in Fig. 1, and al- lowed to touch only the summit of the blood drop and not the skin, the blood will spread out properly between them, and no pressure on the cover-glass will be needed to make the layer of corpuscles thin enough. Pressure is undesirable, as it often makes all sorts of artefacts in the preparation and hastens cre- nation of the red corpuscles. Better results are obtained if slide and cover are warmed just before using. Prevention of Cell-Death. Slides so prepared are usually best examined with a one- twelfth oil-immersion lens. As a rule they keep long enough for purposes of examination without any further precautions, but if we desire to keep the blood fresh and uncoagulated for a longer period, it is best to exclude air in this way : Paint upon the slide with vaseline, cedar oil, or any gummy substance a hollow square or ring of about the size of the cover-glass, so that when the latter with its drop of blood is put down upon the slide the drop will spread out inside the ring of oil, which seals the margins of the cover-glass to the slide. Specimens so prepared will keep for many hours unchanged, and without cre- nation or coagulation, if the weather is warm or if the slide be kept in a warm place. In examining blood suspected of containing malarial para- sites it is sometimes useful to put the whole microscope into one of the warming apparatuses devised for the purpose. This is better than any of the various kinds of warm stage in use, but in clinical work there is rarely if ever any need for artificial heating 1 Further experience has convinced me that water alone is generally sufficient, provided the polishing is thorough. Tissue paper is very useful for polishing cover-glasses. After polishing, it is well to pass them through a Bunsen or alcohol flame once or twice. METHODS OF CLINICAL EXAMINATION. 9 apparatus of any kind, provided the room and the slide are warm. What Can be Learned from Fresh Blood. Examination of the fresh blood by the method just described is the best way known for ascertaining the presence or absence of 1. The Plasmodium malariae. 2. The Spirochaete of relapsing fever. 3. The Filaria sanguinis hominis. 4. Rouleaux formation among the red cells. It is also a quick and convenient method of finding out with approximate accuracy : (a) Whether the blood contains an increased amount of fibrin ; (b) "Whether any considerable anaemia or leucocy tosis ' is present ; (c) Whether or not the amount of haemoglobin in the red cells is much decreased; (d) Whether the red corpuscles are deformed ; (e) Whether the " blood plates" are increased or not. A practised observer can also make a diagnosis of leukaemia by this method in most cases, but here mistakes may easily occur. So much can sometimes be learned from a specimen pre- pared in this very quick and easy way that it should be as much a matter of routine as a urine examination. But in order to get any information from such a preparation we must previously have familiarized ourselves with the appearance of normal blood under such conditions with the size, shape, color, and refrac- tions of the red cells, white cells, and blood plates and their ratio to one another, and with the great variety of curious phenomena to be seen as a drop of blood gradually dries up between slide and cover. No book can teach this: it must be learned by actual experiment. Some of the commoner sources of error will be referred to later. Here I will mention only the Brownian movement in the protoplasm of the corpuscles, to be distinguished clearly both from the amoeboid movements of the leucocytes or of the malarial parasite and also from the irregular contractions of the dying 1 More accurately it is only the ratio of red to white corpuscles that we can determine, and when the red are very much diminished in num- ber we may be deceived into supposing that the white are increased. 10 CLINICAL BLOOD EXAMINATION. protoplasm, which give rise to pseudo-amoeboid motions in the crenated points of normal red cells or in the irregular projections of corpuscles deformed by disease (vide infra) . For a more detailed description of normal red corpuscles, white corpuscles, and blood plates the reader is referred to Part II. An account of the pathological changes to be observed in the fresh blood will be given in later chapters. CHAPTEE II. COUNTING THE CORPUSCLES. I. THE Thoma-Zeiss counter. II. Durham's modified counter. I. Out of the many instruments devised for this purpose that of Thoma-Zeiss is much the most commonly used. In the use of this instrument there are five steps or stages : 1. Puncturing the ear. 2. Diluting and mixing the blood thus obtained. 3. Adjusting a drop of di- luted blood in the counting chamber. 4. Counting the corpuscles, 5. Cleaning the pipette. To count the white corpus- cles, a different instrument is often used from that em- ployed for the red. The technique is nearly the same for both instru- ments, but for clearness' sake I shall describe them sepa- rately. To save time I shall call the small-bore pipette used for red corpuscles (Fig. 2, A} the "red counter," and FraS.-Thoma-Zeiss Pipettes. A. For red cor. the ^rge-bore pipette (Fig. puscles; B, for white corpuscles. 2, B) the " white Counter." COUNTING THE BED CORPUSCLES. (a) After puncturing the ear as above described, and as soon as the blood is flowing freely, put the point of the " red counter" 12 CLINICAL BLOOD EXAMINATION. into the drop as it emerges from the ear, and by sucking gently on the rubber tube attached to the other end, draw up blood to the mark 0.5 on the pipette. It is convenient to rest the end of the pipette on the thumb as shown in Fig. 3. It needs some practice to stop exactly at the mark, but if we happen to draw the blood up a little past the mark 0.5 no con- siderable error results, provided we draw the column down again to the mark by tapping the point of the pipette on a towel, and provided also that the instrument is perfectly clean and dry. The aim and intention, however, should always be to stop FIG. 3.-Method of Resting J Point of Pipette on tne exactly at the mark 0.5, and with a Thumb while Sucking in j^tle practice we can do it, except with nervous or delirious patients, and those who carelessly move the head just at the critical moment. With such patients we usually have to content our- selves with drawing the blood a little beyond the mark 0.5 and then drawing it down again to the mark as above de- scribed. Diluting the Blood. (V) The bottle of solution to be used for diluting the blood should be ready uncorked at the bedside. Of the many solu- tions suggested by various authors none is better than Gaivers', the formula for which is as follows : Sodii sulphat., . . . ... gr. 112 Acid, acetic. , . . . . . . 3 v. Aquse, ..' iv. Toisson's solution is also very useful and stains the white corpuscles so that they can be easily distinguished from the red. Its composition is as follows : Methyl violet, 5 B,. 025 gm. Sod. chlor., 1.000 " Sod. sulph 8.000 " Neutral glycerin, 30.000cm. Aquse destill. , 160.000cm. COUNTING THE CORPUSCLES. 13 We must wait about ten minutes after mixing before the leu- cocytes are fully stained. Except for this delay, the only diffi- culty of this solution is that it is rather difficult to clean the pipette after using it. If the white cells are counted with an- other pipette the staining fluid can be as well dispensed with. Into a bottle of one of these solutions, ready at the bedside, the point of the pipette is to be plunged as soon as the blood has been drawn up to the point 0.5 and the outside of the pipette wiped clean of blood. Suction is then exerted through the rubber tube the instant the point of the pipette is below the sur- face of the diluting solution. This suction is continued until the diluted blood has filled the bulb of the pipette and gone past it up to the point marked 101. It is not difficult to stop at this point, provided the pipette is perfectly clean and dry inside. Otherwise it is impossible. Should any mishap oc- cur at this point, the whole process must be begun over again after carefully cleaning and drying the pipette. If no acci- dent happens and the mixture is sucked up to and not past the mark 101, we have diluted the blood with two hundred times its bulk of neutral solution. If, instead of drawing the blood up to the mark 0.5 we draw it as far as the point marked 1, and then dilute as above described, the mixture will be 1 to 100. Some observers habitually use this dilution. The objections to it are : (1) That if the blood is accidentally drawn up too far (i. e. , past the mark 1) we cannot draw it down again but must painfully clean and dry out the pipette (see below, p. 16) and repeat the process. (2) If the blood contain ap- proximately the normal number of corpuscles, they will be so crowded when adjusted on the ruled surface of the disc A that it is more difficult to count them. If we use another pipette for the white corpuscles the dilution of 1 : 100 has no advantage to counterbalance these drawbacks. While sucking in the diluting solution, it is well to roll the pipette on the long axis with the fingers of the hand which holds it in the diluting fluid. This mixes the blood instantly and prevents any of it from floating on the top of the solution and thereby coming up undiluted into the narrow portion above the bulb of the pipette, where it might possibly escape thor- ough mixing. * *Care must be taken that no saliva finds its way through the rubber tube and into the pipette. Never blow through the rubber tube. 14 CLINICAL BLOOD EXAMINATION. Next we thoroughly mix the blood and diluting fluid by shaking and rolling the pipette, its ends being closed by the fingers. The little glass ball within the bulb helps this process materially. A minute's brisk rolling and shaking is as good as five minutes', as I have convinced myself by many experiments, and the distribution of the corpuscles throughout the mixture is very even, provided there is no delay in proceeding to the next step, Viz. : (c) Adjusting a Drop of Diluted Blood in the Counting CJiamber. Remove the rubber tube from the pipette and blow out the por- tion of diluting solution which last entered the pipette, and which consequently has not been thoroughly mixed with the blood in the bulb. Five or six drops should be blown out before any is used for examination. Next put upon the surface of the couDter (A, Fig. 4) a drop of such size that when the cover-glass (B) \ x i V \c FIG. 4. Thoma-Zeiss Counting Slide. A, Ruled disc ; B, cover-glass ; C, moat. is let down over it the whole of the disc A is covered with the drop without any being spilled into the "moat" (C) around it. Just how large such a drop should be can only be learned by practice. It is not literally necessary thai exactly the whole disc A should be covered, provided nine-tenths of it is covered, but any spilling over into the "moat" (C) entails serious error. After the cover-glass has been let down upon the drop, we should be able (provided the whole instrument is clean) to see concentric rainbow rings between the cover-glass and the body of the instrument. These are known as Newton's rings. A little pressure with a needle on the cover-glass will often bring them out if they do not at once appear, but they must remain visible after the pressure is taken off. Otherwise we know that fchere must be some dirt or dust under the cover-glass prevent- ing its settling exactly into position, and this will cause error in 1 If we have to pause before going on to the next step, we must take care to roll and shake the pipette again when ready to proceed. COUNTING THE COKPUSCLES. 15 the count, though not a very considerable error in most cases. (To see Newton's rings we should get our eyes near to the level of the counting chamber so that the light from window or lamp is reflected from the surface of the cover-glass.) If the above conditions are not all fulfilled the instrument should be washed and another drop tried, after shaking the pipette and blowing out a few drops as before. The cover-glass must be let down as soon as possible after the drop has been put on the disc A, and before the corpuscles have time to settle. It is best to let it down with a needle as in mounting microscopic specimens. Counting. (d) After waiting two or three minutes so that the corpuscles may settle thoroughly upon the space ruled off on the disc A, A O.lOOmtn. 4005 C.Zeiss Jena FIG. 5. Thoma-Zeiss Counting Slide. A^ Ruled disc. the counting is begun, using preferably an objective 5 of Leitz or D of Zeiss and a No. 1 or 2 eyepiece. The ruled space on the surface of the counter (A, Fig. 5). is divided into four hundred squares, every group of sixteen squares being enclosed in double lines to make it easier to know how many squares we have counted (see Fig. 6). Including the squares with double lines we have a group containing thirty- six small squares, a group convenient to count at one time as it just about fills the field of the objective Leitz Na. 5, or Zeiss D with a No. 2 eyepiece. To avoid considerable error we should count the corpuscles 16 CLINICAL BLOOD EXAMINATION. in five fields of thirty -six squares each, such as is shown in Fig. 6, taking the fields in various parts of the whole ruled space. The instrument should then be washed* and the whole process repeated with a second drop. If the count of the second drop differs widely from that of the first, a third drop should be counted and the average taken of those two which are most nearly alike. Thus at least three hundred and sixty small squares should be counted ; with such a number the error is not over three per cent for practised observers. 2 In normal blood this means counting about 2,160 corpuscles, as six or seven to a small square is about the normal average when we are using a dilution of 1 : 200 such as has been described (twelve to fourteen cells per square in a dilution of 1 : 100). Among the difficulties encountered in counting is the pres- ence of a few corpuscles on or touching one or more of the lines bounding the space to be counted. Shall we count these out or in? In counting, for in- stance, a field like that in Fig. 6, what are we to do with the cells which sit astride the lines AA, BB, etc. ? To get round this difficulty, it is best to make it a rule to count in all the corpuscles on or touching some two of the boundary lines (e.g., AA and BB) and to take no notice of any cell on or touching the lines CC and DD. In this way the exclusions just balance the inclusions. Of course all cells within these outer boundary lines are to be counted whatever their position. 1 Use only water alcohol dissolves the cement which holds the ruled disc in place. 2 See Reinert's "Zahlung der Blutkorperchen," Leipzig, 1891, p. 48 et o c o o o O r oo >o o O 6 o o o ooo n , o o o o 6 **J O O 00 o oo o o <'. o O o 00 o 00 o o o Oo o o o ) OO O o 00 o o o o 00 o < 0' 00 O O o oo o o o o O O o v o u o . Q o O o oo o o o w o o q o o lo o o oo FIG. 6. Field of Thirty-six Squares on Ruled Disc of Thoma-Zeiss Counter Covered with Normal Blood Diluted Two Hundred Times. COUNTING THE CORPUSCLES. 17 Beyond this the details must be settled by each man for him- self. My own habit is to count through the squares in the order indicated by the track of the serpentine arrow in the accompany- ing Fig. 7, and to count by twos or threes. A movable stage makes the counting easier, especially for be- ginners. Either natural or artificial light may be used, with a small aperture diaphragm, and if the instruments are clean and the diluting solution fresh and free from sediment, l there is no difficulty in deciding how many cells each square contains, and no extraneous fragments to be excluded. We must distinguish the white corpuscles from the red, not by their size but by their stain if Toisson's solution is used, otherwise by their peculiar shining look when the lens is drawn up so as to put the red cells slightly out of focus. The blood plates are not noticeable and lead to no errors. When the number of corpuscles in 360 squares has been counted the number is divided by 360 and multi- - plied by 800, 000 (i.e., by 200 to make up for dilu- tion and then by 4,000 because each square is equivalent to J^VTF of a cubic millimetre), which gives us the number of _ corpuscles per cubic mil- limetre. These figures need not - be committed to memory, ~~for we have marked on the instruments used all FIG. ?.-The Arrow Indicates the Order in which the tllQ data necessary for the squares are counted. calculation, i.e., the dilu- tion figures on the pipette and the area and depth of a single square on the counting slide. (e) The importance of cleaning the pipette as soon as the counting is done is so great that it should be reckoned as one of the regular steps on every count. First water, then alcohol, and 1 Most diluting solutions precipitate or accumulate spores, and need to- be frequently renewed or filtered. 2 18 CLINICAL BLOOD EXAMINATION. lastly ether must be sucked into the pipette and brought into contact with every part of the bulb and tube. After this air must be sucked or pumped through the tube until it is per- fectly dry and the glass ball will roll about freely in the bulb without sticking anywhere. These precautions take but two or three minutes, and if they are omitted and the blood dries in the pipette, it may take several hours' work to get it clean. Further, if it is not thor- oughly dried after cleaning, the mixing of the blood when it is used next cannot be done accurately. The first three steps of the above process (i.e., the obtaining, diluting, and mixing of the blood) must be done as swiftly as is compatible with accuracy, but when once the blood is mixed in the pipette it can be kept there indefinitely and counted at leisure. None of the corpuscles are destroyed or lost, and if the bulb is thoroughly rolled and shaken up whenever we are ready to count the blood, no error results from keeping it twenty-four hours or more in the pipette. It is not necessary, therefore, to carry a microscope to the patient's house or bedside; the pipette and the diluting solution" are all that we need to take with us, and when the blood is mixed in the pipette, the latter 's ends can be closed with a rubber band and the blood carried home and counted at leisure. The pipette should be kept approximately horizontal during the transit. COUNTING THE WHITE CORPUSCLES. To make a reasonably accurate count of white corpuscles, using the " red counter" and the dilution of 1 : 100 or 1 : 200, we need to count an immense number of squares, far more than was necessary in estimating the red cells in fact, at least ten times the whole ruled space. It is therefore far quicker and more ac- curate to use the " white counter" or large-bore pipette with a diluting solution which renders the red cells invisible and leaves only the white to be counted. Such a solution is the one-third of one-per-cent. solution of glacial acetic acid in water. With this the white corpuscles stand out very clearly and the red can barely be seen at all. The technique is the same as that already described, with the following exceptions : 1. The drop of blood needed is nearly three times as large COUNTING THE CORPUSCLES. 19 as that used in the "red counter;" it is about as big as can be made to stay on the ear without rolling off. 2. The bore of the tube being large, it fills and empties more readily. Hence our suction must be gentler, and it is rather harder to stop exactly at the mark 11. For the same reason the diluted blood will run out of the pipette if the latter is not kept nearly horizontal, and the bottle of diluting solution should accordingly be tipped up as we plunge in the point of the pipette, so that the latter is depressed as little and for as short a time as possible before suction begins. 3. Instead of counting separate fields of thirty-six squares each, we should count the whole ruled space and then repeat the process with a second drop. This takes never over fifteen minutes, often not over five, and is very accurate. The advantages of this pipette are obvious. The only draw- backs are its expense and the need of a somewhat deeper and more painful puncture to get blood enough for it. The tech- nique is not at all difficult. Counting Both Red and White Cells With the Same Pipette. We may avoid buying both large-bore and small-bore pipettes in one of the following ways : 1. We can count both red and white corpuscles with the " red counter. " 2. We can count both red and white corpuscles with the "white counter." The reason why we cannot use the " red counter" for count- ing white cells, unless modified in some way, is that in the whole ruled surface of the counting chamber not more than three or four white corpuscles are to be found in normal blood when diluted two hundred times. If we dilute less, we cannot see the cells distinctly because they are so crowded. If we find, say, three white corpuscles as the number to be used as a basis in calculating the number of white cells in a cubic millimetre, the chance of error is very great, the multiplier being so large (2,000) and the multiplicand so small (3). To get over this difficulty we may utilize the cells spread over the disc of the counting chamber outside the ruled space in one of the following ways : (a) By measuring the field of the objective used. The writer's CLINICAL BLOOD EXAMINATION. objective, No. 5 of Leitz, has a field of very nearly one-quar- ter of a square millimetre or one-quarter of the whole ruled space. Four fields of this lens, taken anywhere outside the ruled space, therefore, contain the same number of cells as will cover the whole four hundred small ruled squares, and when we have counted the white cells in a series of four fields of this lens, we have accomplished as much as if we have put a fresh drop upon the counting chamber and counted all the ruled squares over again; the latter process is tedious, the former very quick. Thus it is my practice in some cases to proceed as follows (see Fig. 8) : Supposing the large circle CCCC to represent the surface of the small disc (A, Fig. 4.) in the centre of the counting chamber, and AAAA the ruled squares in the mid- dle of this disc, four microscopic fields are taken in the di- rection away from the centre indicated by circles and ar- rows in the figure. Starting, say, to the c right of the ruled squares with the left edge of the mi- croscopic field just touching the outer boundary line of the squares, count all the white cells to be seen in the field. Then move along to the right till the corpuscles which were on the extreme right of the first field have gone out of sight to the left. Your field is then in the position of the circle marked 2 (Fig. 8). Count all the white cells in this field and so on for four fields. With my objective, four such fields are almost exactly equal to the whole ruled space AAAA. With other ob- jectives of course the number of fields is different. When we have counted four fields in each of the four direc- tions indicated by the arrows we have covered as much ground COUNTING THE CORPUSCLES. as if we have put four successive drops on the slide after the first one and counted all the ruled squares in each, and we have saved much time and labor. (&) Another and better method of attaining this same end is as follows: Cut out of black cardboard a piece of the shape shown in Fig. 9 and of such a size that it will fit into the tube of the eyepiece the square aperture allowing a space of just one-quarter of a millimetre (one hundred of the ruled squares) to be seen through it with a given objective (say Leitz No. 5). Four fields as seen through such an aper- ture can then be counted in various parts of the slide outside the ruled space as explained above. (c) For any one living where micro- scopic ruling on glass can be done at a moderate cost, by far the best way is to have the rest of the disc A (Fig. 5) ruled off as shown in Fig. 10. Leitz & Zeiss a FIG. 9. FIG. 10. Modified Ruling of Thoma-Zeiss Counting-chamber. now make to order instruments so ruled. I have not been able to hear of any one in America who could do such work at a moderate expense. 22 CLINICAL BLOOD EXAMINATION. (d) We may work out mathematically what number of squares would be contained on the whole disc were it all ruled like the central portion. This can 'be done with the aid of a micrometer eye-piece and a mechanical stage. There is some variation in individual instruments, but as a rule the disc out- side the central ruled space has an area of about two thousand of the small squares. 2. We may use the " white counter" for red corpuscles in the following way : Suck up blood only to the first mark up from the point (i.e., one-fifth of the usual distance) and then Gowers' or Toisson's solution up to the mark 11. This gives a dilution of 1 : 100, and in anaemic cases, in which the cells are not very numerous, answers well. The same pipette can then be care- fully cleaned and used for counting white cells with the acetic acid one-third per cent, and a dilution of 1 : 10 or 1 : 20. WTiatever method of counting white corpuscles is adopted, we ought to have at least one hundred corpuscles actually counted to use as the multiplicand of our computation. A single drop from the white counter with a dilution of 1 : 10 gives us normally about seventy white corpuscles in the four hundred ruled spaces, and by repeating the process with a second drop the result may be made reasonably accurate. This was the method adopted by Rieder 1 in the immense number of counts made by him. II. Durham's Modified Hcemocytometer. In the Edinburgh Medical Journal for October, 1897, Herbert E. Durham, of Cambridge, England, describes a self -filling capillary pipette which has considerable advantages over the ordinary Thoma-Zeiss instrument. The account of the device is here given in his own words. " The apparatus entails no new principle ; it is rather to be considered as an adaptation of a number of details, which to- gether seem to present some advantages. As in the Gowers' instrument, there is a separate capillary pipette for measuring the blood, one for measuring the diluting fluid, a mixing vessel, and -the counting chamber. A few words may be said about each of these. " Capillary Pipette. There is an obvious advantage in the 1 " Beitrage zur Kenntniss der Leucocytosis, " Leipzig, 1892 (Vogel). COUNTING THE CORPUSCLES. 23 use of a self-measuring pipette. It cannot go wrong by acci- dent. Durham has availed himself of the pipettes introduced by Dr. Oliver, namely, small pieces of thick- walled capillary tube 5 and 10 c.mm. in capacity. These are carefully recali- brated by the makers of Dr. Oliver's instrument The Tinto- meter Company. " There is, moreover, another important advantage attaching to Dr. Oliver's pipette; this consists in the readiness with which it may be cleansed. As he has described, all that is necessary is to pass a piece of darning cotton by means of a needle through the bore of the pipette. All the adherent serum, etc. , is completely removed thereby. Durham generally wets the end of the cotton with ether, but this is not absolutely necessary. In passing the needle, it is better to pass it into the pointed end, in case it is not withdrawn perfect^ axially, when there is a liability to chip the thinner unsupported glass. "Any one who has worked much with the Thoma-Zeiss pipette will know how troublesome it is to clean, especially when a number of observations have to be made in a limited time. Unless it is frequently cleaned out with strong acid, there is a tendency for the deposition of sticky serum remains which interfere with true readings. "For use, Dr. Oliver's pipettes are mounted by means of a FIG 11. Cross-section of Durham's Automatic Blood-Pipette. T, Glass tube (like that of medicine-dropper) ; JV, rubber nipple (like that of medicine-dropper) ; p, perforation in the nipple; c, Cork holder, perforated by capillary pipette. small cork (c) in a large glass tube ( T), which is provided with a rubber nipple (N), having a lateral perforation (p) (Fig. 11). " The mixing vessel consists of a small test tube (2f X yV i n - for 1 c.c., or 2fx| in. for -J- c.c.). Several such tubes may be kept, so that a number of observations can be made if necessary. For thoroughly mixing the blood and diluting fluid, one or more small glass globules are placed in the tube. By using different colored glass globules, different specimens can be readily differentiated. 24 CLINICAL BLOOD EXAMINATION. " For measuring the diluting fluid, pipettes containing 1 and J- c.c. are used; these are marked at 995 and 990 c.mm. and 495 and 490 c.mm. respectively. With these graduations the follow- ing dilutions may be obtained : 1 : 200, 1 : 100, and 1 : 50, with the appropriate capillary pipette. " Having measured the diluting fluid, according to the event- ual dilution desired, the blood capillary is filled by touching the exuding drop of blood and allowing it to completely fill itself. The blood may be obtained in the usual manner from the lobule of the ear, the first drops being wiped away. " The hole in the nipple allows free air-way so that there is no hindrance to the action of capillarity. When filled, any blood on the outside of the pipette is rapidly wiped off and the tube is inserted into the mixer until the point is one-half to three- fourths of an inch above the level of the contained liquid. " The nipple is then held in such a way that the hole lies under the thumb of the operator. When this is the case it is slightly squeezed, and then, while the pressure is continued, the bulb is rotated so that the hole is free again. In this way the blood is squirted out, but not sucked back again. The pro- cedure is extremely simple and really requires no practice, given an operator who is not possessed of 'five thumbs.' In order to wash out the remains of the blood the point of the capillary is dropped into the diluting fluid; the bore instantly fills itself. It is then withdrawn and the pressure and rotation of the nipple are repeated. This has to be repeated several times, and occupies a few seconds of time. It has been sug- gested that a certain amount of error is introduced by measur- ing the diluting fluid in a pipette, the inner surface of which retains some moisture ; this is extremely small in amount if the pipette is emptied slowly, and comparative readings with the Thoma-Zeiss apparatus show that the error is negligible. " To mix the blood and diluting fluid thoroughly, the mixer is placed between the opposed hands, which are rubbed backward and forward; the mixer is rotated thereby, and the glass globules cause a thorough dispersion of the corpuscles in the fluid. " A drop of sufficient size is then placed upon the counting chamber, and the cover-slip is slipped on sideways in the usual way I prefer the Thoma-Zeiss counting chamber. COUNTING THE CORPUSCLES. 25 " The advantages of this method are : "1. The ease and thoroughness with which the pipette can be cleaned. "2. The manifest advantage of the self-measurement of the blood. "3. The avoidance of the objectionable necessity of using the mouth to suck fluids into the pipette. "4. The measurement of the diluent can be done carefully and calmly beforehand, and any error corrected without taking any more blood. " 5. The greatly smaller cost of the pipette. "6. The same pipette is useful for making various dilutions in serum diagnosis, by using several mixing vessels filled before- hand with dilute fluid." CHAPTEE HI. OLIVER'S TINTOMETER CENTRIFUGALIZING THE BLOOD- HAEMOGLOBIN ESTIMATION SPECIFIC GRAVITY- STAINED SPECIMENS BACTERIOLOGICAL EXAMINATION. OLIVEE'S TINTOMETER. KECENTLY a method of estimating corpuscles by means of their optical effect, and without directly counting them, has been introduced by Dr. Oliver. For practical purposes an actual counting of the corpuscles must be considered a neces- sity ; not only since the number of leucocytes is not without im- portance (e.g., in the diagnosis of enteric fever), but also since these cells may be so abundant that they may interfere with the use of optical methods, as in the case of leukaemia. Neverthe- less the instrument is very accurate and useful in many cases. Its principle is based on the fact that if a small quantity of blood is gradually diluted with Hay em's solution 1 in a test tube whose sides are flattened so that its mouth forms a rectangle about 15 mm. by 5 mm. , and a candle flame is looked at through the mixture, there is to be seen, ivlien a certain degree of dilution is readied, a bright horizontal line on the glass (see Fig. 15). This line is made up of a large number of minute images of the flame, produced by the longitudinal striation of the glass. If the quantity and quality of blood used is in every instance the same, the degree of opacity depends wholly on the amount of Hayem's solution added. It is found that with normal blood the amount of diluting solution necessary to allow the image of the candle flame to be seen through the mixture is always the 1 Hydrargyri perchloridi, 0. 5 gm. Sodii chloridi, . . . . . . . 1.0 " " sulphatis, . .,.-' 5.0 Aquae destillatae, 200.0c.c, OLIVER'S TINTOMETER. uo FIG. same, and can be very accurately fixed, so that a variation of one per cent in the number of corpuscles can be distinguished by noting the amount of diluting solution which must be added before the image of the flame appears. To collect the blood, Oliver uses a capillary pipette containing about 10 c.mm. (one large drop), and used exactly in the same way as the v. Fleischl capillary pipette (see Fig. 12). One pipette full of normal blood is gradually diluted in the flattened tube with Hay- is. capillary em's solution until a bright Diver's horizontal li ne caused by the image of candle flame becomes visible through the mixture. The point to which the column of the mixture then reaches is marked 100, and then space between that point and the bottom of the tube is divided into 100 equal parts. The point marked 100 is then equivalent to 5,000,000 red corpuscles; 90 = 4,500,000, 80 = 4,000,000, and so on, each degree on the scale corresponding to a difference of 50,000 corpuscles (Fig. 13). Use of Oliver's Tintometer. The capillary pipette is filled in the usual way, and the outside carefully and quickly wiped if necessary. The medicine dropper (previously filled with Hayem solution) is then connected with the polished blunt end of the pipette by means of the rubber tube (Fig. 14), and blood washed into the test tube as shown in Fig. 14. If the previous haemoglobin esti- FIG. 13. Measuring mation has shown ninety to one hundred per cent of coloring matter we can safely add the diluting solution rapidly until the point marked 80 is reached. If the coloring matter is lower we must cease our rapid dilution correspondingly sooner. When we get near the point 60 S.O Tube for Oliver's Tintometer. CLINICAL BLOOD EXAMINATION. at which the flame-image is likely to appear, the diluting fluid must be added a few drops at a time. After each ad- dition put the thumb over the mouth of the tube and turn it upside down once or twice to mix the blood thoroughly, wiping FIG. 14. Use of Oliver's Tintometer. Method of washing in the blood. FIG. 15. Method of Holding the Tinto- meter while Diluting. Note the bright hori- zontal line in the fluid, indicating that enough diluting fluid has been added. the thumb each time on the edge of the tube so as to put back what fluid has adhered to it. At a certain point the image will suddenly become visible. It is seen soonest if we rotate the tube on its long axis, as the image becomes visible earliest at the sides of the tube. The whole process should be carried on in a perfectly dark room, and the diffused light of the candle must be shut off from the eye. This is best done by fitting the tube into the hand as shown in Fig. 15 with the long axis in line with candle, holding the tube close to the eye, and standing about ten feet from the candle. In the use of both of his in- struments Oliver uses only the small wax candle known as Christmas candles, whose flame is of the most convenient size. CENTRIFUGALIZING THE BLOOD. 29 THE ILEMATOCRIT. THE haematocrit of Hedin, though a comparatively new in- strument, has undergone considerable modification and improve- ment in the last few years and as remodelled and improved by Judson Daland is now coming into use in this country. Its direct and obvious object is simply to ascertain the relative volume or mass of the corpuscles and of the plasma in a drop of blood; but the hope of its advocates has usually been that it would supplant entirely or mostly the long, tedious, and eye- destroying process of counting with the Thoma-Zeiss instrument. Whereas the latter needs sometimes an hour's hard work and eye strain to make an accurate count of red cells, with Daland' s centrifugal machine one can get the result in five minutes with- out any strain on the eyes. Daland maintains the superior accuracy of his instrument in most cases as a further advantage of its use. The estimation of corpuscles depends on the length of the column of corpuscles packed down by centrifugal force at the end of a capillary tube filled with blood and whirled with great rapidity in a horizontal plane. The more corpuscles the longer the column. Wherever there is much variation in the shape or size of the cells, as in many forms of anaemia, leukaemia, etc., the hsematocrit is evidently inaccurate, inasmuch as the misshapen, under- or oversized corpuscle will pack down differently from the normal cells, three million undersized cells making a shorter column in the tube than three million healthy ones. This is recognized by the advocates of the instrument, which is accordingly recom- mended only in those cases in which we know that there are no- considerable variations in the size or shape of the red cells. These are usually cases in which no very great anaemia is pres- ent and in which consequently the labor of counting the large number of corpuscles is greatest. It seems, therefore, as if the hsematocrit might relieve us of the most irksome part of blood- counting without loss of accuracy. Against this there is to be said that we do not as yet know how far the elasticity and compressibility of otherwise 30 CLINICAL BLOOD EXAMINATION. healthy corpuscles may vary and how far such a variation may invalidate the standard of tight packing established from other cases. Further, there is known to be a certain amount of varia- tion in the size and volume of a healthy person's corpuscles, both between nations and between members of one nation, and it is yet to be shown whether this variation is sufficient to make the result of the hsematocrit liable to a greater error than those of FIG. 16. Capillary Tube of Heematocrit with Rubber Attached. the Thoma-Zeiss instrument. There is no doubt that the lat- ter is a slower, more tedious instrument ; the question is still open whether or not it is the more accurate. Daland reports wide variations between his counts and those of his colleague and between different counts by one observer at different times, using the Thoma-Zeiss instrument, while with the hsematocrit the variations are but slight. In testing these results I have made parallel counts of a pa- tient's blood with several of the house physicians at the Massa- chusetts General Hospital during the last two years and our differences have never exceeded the limit of error laid down by Eeinert namely, two per cent. I think Daland must have been unfortunate in his results. If, then, the error of the Thoma-Zeiss instrument is, as I believe, not over two per cent under ordinary circumstances and with correct technique, it does not seem likely that the haema- tocrit is a more accurate as well as a simpler and quicker in- strument. To use the Daland hsematocrit we prick the ear as usual and with the help of a bit of rubber tube attached to one end of the capillary tube (Fig.16) suck in enough blood to fill it entirely. Usually we draw in more than enough and it enters the rubber tube as well, but this is no harm. It is nearly impossible to fill the glass tube exactly and no more, inasmuch as the proximal end of it is hidden inside the rubber tube. The commonest CENTKIFUGALIZING THE BLOOD. 31 mistake at this point is incomplete filling of the capillary tube, as a very large drop is needed to do it. As soon as it is full, put the finger (greased with vaseline) FIG. 17. Daland's Hsematocrit. Two capillary tubes in place on the horizontal whirling beam. The instrument i to be fastened to the edge of some solid and bulky piece of furniture by means of the thumb- screw seen at the bottom of the cut. If not very tightly secured, it will work loose when the handle is revolved rapidly. tightly over the free end of the glass tube and then, but not till then, draw off the rubber tube and adjust the glass as quickly as possible in the place prepared for it on one of the horizontal arms of the whirling machine (Fig. 17). A similar tube (empty) 32 CLINICAL BLOOD EXAMINATION. should be put on the other arm of the crosspiece to make the balance true. "We must be quick about this, else the blood will coagulate. The handle of the instrument is then revolved at least seventy times a minute for two minutes, at the end of which time (sometimes less) the column of blood cells is packed so tight that no further whirling has any effect on its length. Great care should be taken that the horizontal beam is securely attached to the main part of the instrument, as it is capable of doing serious damage should it come off while whirling nine thousand revolutions a minute, which is the rate usually attained. It is well to put a little vaseline on the point where the blunt end of the tube rests (a, Fig. 17) to prevent any of the blood sticking there when we come to take the tube out and read it. The capillary tube is marked off into one hundred equal divisions and provided with a magnifier like that on clinical thermometers. Laid on a piece of white paper it is easy to read off the number of divisions occupied by the blood column, al- though the end of it is often frayed or bevelled in a way that precludes great accuracy. In normal blood the white corpuscles hardly show at all in the tube. They accumulate at the free end of the column of red cells, but unless a leucocytosis is present their presence is indicated, if at all, only by a slight grayish blur at the end of the red column and cannot be accu- rately measured. This blur is another difficulty in the way of deciding precisely where the end of the red-cell column is. To estimate the number of red corpuscles from the length of the column, we call each degree of the scale on the tube 100,000 cells, or a little more. Thus if the blood column in the tube ends at about the mark 50 we consider that the blood has rather more than 5,000,000 red corpuscles per cubic millimetre. So far all observers agree on the figures, but as to just how much more or less than 100,000 each degree on the scale is worth there is some variation between different observers . Daland, 1 in a long series of comparative observations of making blood counts and hsematocrit estimations on the same case, conclude that each degree of the scale on the capillary tube corresponds to 99,390 corpuscles. 1 University Med. Mag., November, 1891. HAEMOGLOBIN ESTIMATION. 33 The writer in a series of forty observations on healthy persons, in each of which a count of corpuscles with the Thoma- Zeiss instrument and a volumetric estimation with Daland's haematocrit was made, found the value of one degree on the glass scale to vary between 105,000 and 123,000 red corpuscles, the average being 112,000. It certainly seems a priori as if variations in the specific gravity of the corpuscles or in the properties of the plasma might make a considerable difference in the number of revolu- tions needed to reduce the column of corpuscles to its smallest size. So far as I can learn, the use of this instrument in Europe has been chiefly for the direct information it affords as to the volume of the red cells and the amount of respiratory surface in the blood, rather than for the indirect information it may give us as to the number of the red cells. It does not seem as yet to be supplanting the Thoma-Zeiss counter. Its bulk and the noise it makes must for the present, I think, prevent its extensive use outside of hospitals. The noise it makes is a very loud and disagreeable one, and will deter many from using it in private practice. HEMOGLOBIN ESTIMATION. /. Von FleiscliTs Hcemoglobinometer. Until recently the instrument most used both here and in Europe is that of v. Fleischl. In France Hay em rules supreme in the matter of instruments, as in everything else concerning the blood, and in England Oliver's apparatus is used to a cer- tain extent. The v. Fleischl instrument will be described first. The principle of its use is that of directly comparing the tint of the blood with various parts of a strip of colored glass (" gold- purpur") whose color shades gradually from a deep red at one end to clear glass at the other. The glass and the blood are brought before the eye side by side and a direct color judgment is attempted. 3 34 CLINICAL BLOOD EXAMINATION. Use of v, FleischVs Scemometer. (a) To use the instrument fill one side of the metallic cell (a, Fig. 19) about one-quarter full of distilled water and carry it to the bedside, together with the little capillary pipette (B, Fig. 18) and the needle for puncturing. The capillary pipette { must be scrupulously cleaned and dried before use. This is best done by drawing a needle and thread (the latter wet with alcohol and ether) through the eye of the capillary tube. When the drop of blood is flowing freely from the ear, put the end of the little pipette hori- zontally into the side of the drop, which will at once fill the tube by capillary attraction if the latter is clean and dry. Carefully but quickly wipe away any blood that may be on the outside of the pipette, and make sure that the blood in it is just flush with the surface at each end and does not present a concave or convex surface. Then put it into the water FIG. is.- "A, colored contained in one of the partitions of the metal- B, capillary ii c ce n and rattle it quickly back and forth, so that the water may be forced in first at one end and then at the other. So far in the process we must work very quick to prevent coagulation which in some cases takes place very rapidly. (6) After this the cell with the capillary tube still immersed in it may be put in place on the body of the instrument (see Fig. 19) and carried to a room or closet where daylight can be ex- cluded and artificial light used to read the instrument by. Then the expulsion of the blood from the capillary tube may be com- pleted by forcing a few drops of water from a medicine dropper through the capillary pipette and into the compartment where the mixing has been begun. Using the metal handle of the pipette as a stirrer, mix very thoroughly the blood and water in every part of the compartment, looking after the corners especially. Then using a medicine dropper, fill both com- partments of the cell to the brim with distilled water, taking .care that neither overflows into the other, and adjust the com- HEMOGLOBIN ESTIMATION. 35 partment containing the clear water so that it comes over the slip of colored glass, while through the compartment containing the blood light thrown upward by the reflector below passes directly to the eye. Turn the thumb screw (see Fig. 19, T) back and forth until the color of the glass is the same as that of the blood, and read off the number on the scale which corresponds FIG. 19. v. Fleischl's Haemometer. a, Partition into which blood is put ; a', partition into which water is put ; (?, mixing cell ; K, K, colored glass slip (see Fig. 11, A) ; P, P, metal frame on which scale is marked ; R, S, reflector ; T, screw which moves the frame, P, P. to that color. This gives the percentage of haemoglobin, 100 being the color of normal blood for men and 80-90 for women. (c) Matching the colors is not at all easy at best, but may be somewhat aided by observing the following precautions : 1. Do not stand (or sit) facing the light, but sideways (i.e., at A or B, never at C, Fig. 20) . For we wish to avoid that the image of one compartment should come on the upper half of the retina and of the other compartment on the lower half, inasmuch as the upper half of the retina is less sensitive to light than the lower and so a less accurate judge of color. By sitting as in Fig. 13, A or B, we get the compartments whose colors we are to match, on the right and left halves of the retina, which are equally sensitive in most persons. 36 CLINICAL BLOOD EXAMINATION. B 2. Use as little light as possible, and always less light for a blood having a low haemoglobin percentage than for one nearer the normal. Slight color distinctions are abolished if there is any more light than is necessary for simple illumination ; too much light dazzles us slightly and so makes us less sensitive in color dis- crimination. 3. Roll up a piece of paper (preferably black) into a tube of such size that it will fit over the metallic cell (D, Fig. 20) and rest on the platform of the instru- ment. Looking through this with one eye we can judge more accurately than with- out it. Keep the other eye closed. 4. Use first one eye and then the other, and never look more than a few seconds at a time, as the eye very quickly gets sufficiently fatigued to lose its finer sensibility. Hence the impression of a first glance is better than a long look. 5. Move the thumb screiu with short, quick turns rather tJian slowly and gradually, for sudden color changes affect the retina more than gradual ones. Suppose, for example, we have got as far as to decide that the tint of the diluted blood corresponds to that of glass somewhere between the numbers 40 and 60 on the scale. Move the screw suddenly from 40 to 55 ; the shock of the change will probably convince you that the blood color is lighter than 55. Therefore start this time at 55 and move it suddenly to, say, 45, which may show that 45 is too light. Thus by a series of quick movements of the screw getting shorter and shorter each time (with frequent rests for the eyes) we can probably get it down to a matter of doubt between, say, 42 and 45. Beyond that few persons can go and many can never learn to read without an error of five to ten per cent. FIG. 20. L, Light; A and B, right positions for observer ; C, wrong position for observer ; D, cell in place. NECESSARY ERRORS. 37 6. If the preliminary reading shows a reading of thirty per cent or less, two or three pipettes full of blood should be used and the reading divided by 2 or 3. A considerable error can thus be avoided. Necessary Errors. So far as I can see, a certain amount of error is absolutely necessary, inasmuch as the bit of colored glass to be seen at any one time through the aperture of the instrument is not (like the blood) all of one tint, but includes a variation of twenty per cent in color, i.e., if the glass appearing at one end of the aperture is opposite 50 on the scale, that seen at the other end of the aper- ture will either be at 30 or at 70. We have, therefore, to pick out as well as we can the color of the centre of the bit of glass show- ing through the cell and compare the color at that point with the color which is evenly distributed throughout the whole of the blood-and-water compartment. This is of course, strictly speaking, impossible. We can no more get hold of and sepa- rate out the color of that central point than we can seize and hold fast the present moment. It eludes our grasp. This difficulty is somewhat lessened by shutting off from view all but a small section of both compart- ments with a bit of black cardboard or metal in which a slit is cut as in Fig. 21. The slit is put at right angles to the partition which divides the cell so that the blood tint is seen at b and the glass tint at w. Many persons are not sensitive enough to colors to attain any reasonable degree of accuracy with the instrument, and there is moreover a very considerable difference be- FIG. 21. -shield tween different instruments in respect to the use with v. . Haemometer. color ot the glass slip. ' Finally the instru- ment has been shown to be entirely unreliable for percentages of haemoglobin under 20. This error, however, can be mostly eliminated by using several pipettes-full of blood and making corresponding reduction in the reading. All these difficulties render the instrument an unsatisfactory 1 Old instruments read lower than those recently manufactured. 38 CLINICAL BLOOD EXAMINATION. one in many ways. Its bulk and expense are also considerable drawbacks. II. Oliver's Hcemoglobiiwmeter. Oliver's instrument corrects two errors which are inherent in v. Fleischl's. 1. It has no sliding scale of color, but compares the blood tint successively with definite tints of glass, each of which is even. The tints are worked out to correspond to the specific dilution curve of blood, for : 2. Since every colored liquid changes color at a different rate when diluted, the dilution curve of blood does not corre- spond to that of glass (which behaves in this respect like a liquid). The glass wedge of v. Fleischl's instrument repre- sents a single color regularly diluted and does not correspond in its degrees to the colors of blood diluted at a similar rate. The scale of Dr. Oliver's instrument is measured to correspond to the actual colors of the blood's dilution curve, by means of the tintometer. In other respects the principle of the instrument is like v. Fleischl's, and the method of using the two is practically the same except that in Oliver's reflected light is used instead of transmitted light. Oliver's instrument consists of a series of twelve tinted glass discs corresponding to the haemoglobin per- centages from 10 to 120 and arranged in two rows (see Fig. 22a) . The intermediate degrees are measured by means of " riders" of colored glass, which can be laid on top of the primary color discs so as to deepen the tint seen. The capillary pipette (Fig. 22 b) is somewhat stouter than v. Fleischl's, but is used in the same way to collect the blood, which is then forced out of it with water from a medicine drop- per (which is fitted with a rubber tube to slip over the blunt end of the pipette) (Fig. 22 c) and washed into a mixing cell (Fig. 22 d) similar to v. Fleischl's, except for the absence of a central partition. Here the blood is mixed in the usual way with water and the cell filled to the brim and covered with a small glass plate. The blood thus prepared is brought close to the scale and there compared with the tint of the different standard color discs. If it matches one of them the observation is complete ; OLIVER'S H^EMOGLOBINOMETER. 39 if not we use one of the glass riders which enables us to read within two and a half degrees. A fuller set of riders can be obtained so as to make it possible to .read down to 1 per cent. The standard is usually arranged for candle-light, but another set of discs can be obtained adjusted to daylight read- FIG. 22. Oliver's Haemoglohinometer. a, Standard color disks ; b, capillary pipette; o, \vashingtube; d, mixing cell. ings. The latter are less accurate. The same precautions as to the exclusion of outer light by means of a " hydroscope" tube, resting the eye frequently, etc., must be observed with this instrument as with v. Fleischl's. [It can be obtained of J. H. 40 CLINICAL BLOOD EXAMINATION. Smith & Cie., Zurich (Wollishofen), for 115 francs plus duties and expressage, or of the Tintometer Company, 6 Farringdon Avenue, London, E. C.] The candle should be placed three or four inches from the instrument and arranged to light both the blood and the color discs alike. A word as to the use of the riders. The instrument as used for clinical work usually has two riders: the one having the deeper tint is used on the upper half of the scale, the other on the lower. Suppose we have decided that the blood color is between 60 and 70. Put the rider on the 60 disc and compare again. If the blood is darker than the 60 disc plus the rider the percentage is approximately 67| (since it is higher than 60+5 [the rider] and lower than 70). If it just matches the 60 plus its rider, the reading is 65. If the blood is paler than this, yet darker than 60, it is about 62^. An error of about 2 de- grees is obviously inevitable. ESTIMATING THE SPECIFIC GRAVITY OF THE BLOOD. The simplest and most available method for clinical use is that of Hammerschlag, l a modification of Roy's 2 method. Chloroform is heavier than blood; benzol is lighter. Mix in a urinometer glass such quantities of the two that the specific gravity taken by an ordinary urinometer is about 1059, i.e., that of normal blood. Puncture the ear, draw a drop of blood into the tube of a Thoina-Zeiss pipette, 'a small medicine dropper, or any other capillary tube, and blow it out again into the chloro- form-benzol mixture. The blood does not mix at all with these liquids but floats like a red bead. If it sinks to the bottom add chloroform, if it rises to the top add benzol, until finally the drop remains stationary in the body of the liquid, showing that its specific gravity is just that of the surrounding mixture. Then take the specific gravity of the liquid, as we do of urine, and you have the specific gravity of the drop that floated in it. The following precautions are needed : 1. Have the inside of the urinometer glass perfectly dry and clean ; otherwise the drop of blood may cling to it and flatten out against it. 1 Wien. klin. Wochenschrift, iii., 1,018, 1890. 2 Proceedings of Physiological Society, 1884. SPECIFIC GRAVITY. 41 2. It is usually well to have more than one drop of blood in the glass in case any mishap occurs with the first one. 3. Add the chloroform and benzol a few drops at a time, and after each addition stir the whole mixture thoroughly with a glass rod. 4. If we have reason to suppose the blood will be lighter than normal (i.e., if the haemoglobin is probably low, vide supra), it saves time to start with a lighter mixture of chloroform and benzol. 5. Avoid having any air within the blood drop. This can generally be seen either in the capillary tube or after the drop is in the mixture. It is safer to take the middle portion of the blood drawn into the capillary tube, as both the first and the last portions of the column are more apt to have air in them. 6. The whole process should be done as quickly as possible, else the chloroform or benzol may work into the blood drop and affect its weight. It is better to have a urinoineter with a scale running as high as 1070, but this is not essential, for the clinically important specific gravities are low, not high. The importance of the specific gravity of the blood, as hinted above, is not so much for itself, but because it runs parallel to the percentage of haemoglobin and gives a figure from which the latter can be computed. The specific gravity of the blood plasma varies very little (except in dropsy from any cause) , and in the corpuscles them- selves the variable element is the haemoglobin. 1 Consequently in most non-dropsical patients the specific gravity of the whole blood varies directly as the haemoglobin. The following excep- tions to this rule must be borne in mind. 1. In leukaemia the specific gravity is relatively higher than the haemoglobin on account of the weight of the leucocytes. 2. In pernicious anaemia with high color index (see below) the haemoglobin is about two per cent higher than we should gauge it to be judging by the specific gravity. Now, as it is far easier to take the specific gravity accurately than to use the v. Fleischl haemometer, and as the instruments needed are already in the possession of most physicians and 1 Except in dropsy in which the corpuscles themselves may get water- soaked. 42 CLINICAL BLOOD EXAMINATION. the solutions not expensive, there are evidently great advantages in taking the haemoglobin in this indirect way. The chloroform- benzol mixture can be filtered and then used over again indefi- nitely, and the bulk and weight of the urinometer with its glass and the chloroform and benzol bottles, are far less than that of the hsemoglobinometer. In dropsical cases we must still use the hsemoglobinom- eter. In other conditions I do not see why it should not be supplanted by the cheaper, easier, more accurate, and equally quick method of calculating by specific gravity. To do this one of the following tables may be used. (I. is from Harnmer- schlag, using the method above described ; II. is modified from Schmaltz, "Pathologie des Blutes," etc., Leipsic, 1896, using a direct weighing method.) Apparently a degree of specific grav- ity means much more at the top of the scale (i.e., 6.6 per cent) than at the bottom (If per cent). These tables are of course not accurate, and further research will be needed to make them so. Spec. Grav. 1033-1035 1035-1038 1038-1040 1040-1045 1045-1048 1048-1050 1050-1053 1053-1055 1055-1057 1057-1060 Haemoglobin. = 25-30 per cent. = 30-35 = 35-40 = 40-45 = 45-55 = 55-65 = 65-70 u = 70-75 = 75-85 = 85-95 II. Spec. Grav. Haemoglobin. 1030 = 20 per cent. 1035 = 30 1038 = 35 1041 = 40 1042.5= 45 1045.5= 50 1048 = 55 1049 = 60 1051 = 65 1052 = 70 1053.5= 75 1056 = 80 1057.5 = 90 1059 = 100 STUDY or THE FINER STRUCTURES OF THE BLOOD. The study of dried and stained specimens with the help of the aniline dyes gives us much of interest and importance in regard to the blood. More can be told about a given case by the study of a dried and stained cover-glass specimen than by any other single method. Preparation of Cover-Glass Specimens. (a) Covers carefully cleaned with soap and water are ar- ranged at the bedside in such position that we can quickly pick STUDY OF THE FINER STRUCTURES OF THE BLOOD. 43 them up without touching their surfaces (see Fig. I). 1 The ear is punctured in the usual way, and one of the cover-glasses touched to the summit of the drop as soon as it emerges. This cover-glass is then let fall upon another in such a way that their corners do not coincide (Fig. 23). If the covers are clean the drop spreads at once over their whole surface ; as soon as it stops spread- ing, slide off the top one witJiout lifting them apart, but exactly in the plane of their surfaces. Have a gas or alcohol flame at hand and dry instantly if you want to get the very best specimens; but this is not at all necessary for most clinical purposes. The under cover-glass is always better spread than the upper. (6) These covers have now to be fixed either by heat or by half an hour's immersion in absolute alcohol and ether (equal parts), or by the same mixture (30 c.c. each) plus five drops of a saturated alcoholic solution of corrosive sublimate (five minutes' immersion), or by exposure to the vapor of forty -five per cent formaldehyde. I 'have used all these methods, but found none of them to compare favorably with the method of heat fixation when we wish to study the leucocytes or the nuclei of any cell. When we wish to see chiefly the changes in the red cells (as in studying the malarial organism, nucleated red corpuscles, degenerative changes, etc.), the alcohol and ether method is good. But when, as in the majority of cases, it is the white cells in which our interest centres, the use of heat is very greatly to be preferred. Heat serves not simply to fix the cells on the glass and to prevent degenerative changes, but also to modify and greatly improve the staining power of the cell when Ehr- lich's triacid stain is used. The method of fixation by alcohol and ether needs little comment, the cover-glasses being simply left in the mixture half an hour or as much longer as is convenient. Half an hour 1 1 often poise them on corks so that their corners are readily accessible to the fingers. The process of making blood films is far easier if another person prepares the drop for us so that we can stand ready with a cover- glass in each hand to catch the drop as soon as it emerges. 44 CLINICAL BLOOD EXAMINATION. is enough. In most cases we use dry heat. The best way to do this is in a dry -heat sterilizer at a temperature of 140- 155 C., according to the stain used. The temperature must be watched very closely, and as soon as it reaches the desired point the heat should be removed. Gradual heating and gradual cooling are best. If we cannot easily get access to such an instrument, we can manage very well with any small iron or copper box having a door or lid and a hole for a cork which is perforated for the thermometer bulb. This supported over a gas or alcohol flame does very well. It needs about ten minutes to get the temperature to 150 C., and as soon as it gets there the specimens should be taken out. The same end can be accom- plished somewhat less accurately with a strip of copper sup- ported over a Bunsen burner or a small gas or oil stove. The copper plate should be about a foot long and two or three inches wide. Such a plate supported on an iron tripod over a flame gets, after a few minutes, to have a fixed temperature at any given distance from the flame, the heat passing off at the end of the plate as fast as it comes, and so not accumulating. On this plate find the boiling point of water by dropping small drops of water on it, and put the cover-glasses at this point face, down- ward. They may be left there for from fifteen minutes to as long .as you please; but with the stain which I have used, fifteen minutes' heating gives as good results as a longer period, and excellent specimens can often be made with five minutes' heat- ing. 1 After allowing the specimens to cool they are ready for staining. Staining. For all details of structure the Ehrlich tricolor mixture or one of the numerous modifications of it is most convenient. 1 With a little practice one can learn to make excellent specimens by simply passing the cover-glass through a Bunsen or alcohol flame about twenty times very rapidly. The rate of speed must be learned by experi- ment, i.e., such a speed and such a number of exposures to the flame as turns out to give on staining a bright yellow color to the red corpuscles (never red or brown or gray), a good definition to the blue-stained nuclei and to the violet or pink granules of the polynuclear leucocytes. These are the essentials of a well-stained specimen, and they depend (a) on the heating, (6) on the make of stain, but only slightly on the length of stain- ing (vide infra). STAINING. 45 The most useful and easily obtained of these is made by mixing : Saturated watery solution of orange G, . . . . 6 c.c. " acid fuchsin, . . . 4 c.c. To these add a few drops at a time, shaking between each ad- dition : Saturated watery solution of methyl green, . . . 6.6 c.c. Then add: Glycerin, 5 c.c. Absolute alcohol, . . . . . . . 10 " Water, . .- '. . . 15 " Shake well for one to two minutes. Let stand twenty-four hours. Do not filter. 1 G. G rubier 's colors are best. I have used only this stain for the past two years, and have never seen any other which compares with it in brilliancy and general usefulness. The staining process is remarkably simple. A drop of the stain is simply spread over the surface of the cover-glass speci- men with a glass rod and washed off again with water after two or three minutes or as much longer as is convenient. With this mixture it is impossible to overstain. If the specimen look too dark (brown or red instead of orange-yellow) it is not because of overstaining, but because of underlieating . It needs a good deal of heat to bring out the full brilliancy of the three colors, and the 100-120 C. usually recommended for heating is entirely insufficient with this stain unless continued for a long time. If overheated the specimen looks pale lemon yellow to the naked eye, and under the microscope everything is blurred and dim. I am convinced that any one who has once seen how much is brought out by a good make of triple stain will never use any other for clinical purposes. Eosin (one-per-cent alcoholic solu- tion) followed after a few minutes by Delafield's hsematoxylon for one minute, or methyl blue one-half minute, gives a very 1 An absolutely reliable triple stain from Ehrlich's latest formula can be had of Walter Dodd, apothecary to the Massachusetts General Hospital A 65-cent bottle will stain several thousand specimens. 46 CLINICAL BLOOD EXAMINATION. striking contrast stain, but does not bring out the points most essential in clinical blood work. To "control" Ehrlich's triple stain with eosin-haematoxylon or eosin-methyl blue is like con- trolling a chronometer with a fifty-cent clock. The latter stains are very valuable for the study of the finer structure of nuclei, for karyokinetic figures and basophilic granules, but not for diagnosis. After staining and washing in water, the covers are dried between layers of filter paper and mounted in Canada balsam, ready for examination with the one-twelfth oil-immersion lens, with wide-open diaphragm. ' Differential Counting. The only procedure in the microscopic examination of such specimens which needs any description is that of making the so-called "differential count" of the leucocytes (i.e., determining what percentage of the leucocytes present belongs to each of the sub-varieties as described on pp. 62-67). To do this accur- ately we should examine at least five hundred leucocytes the examination being simply the classification of them under their different sub-varieties. A movable stage is very convenient though not essential for this purpose. With such a stage the technique is simply to start with the lens in, say, the upper left- hand corner of the blood film and, by turning the screw of the mechanical stage, move the preparation slowly past the eye until the upper right-hand corner is reached. During this process as the cells appear in the field they are checked off and put down under one or another heading. Then move the stage so that the lens is just one field's diameter nearer the right-hand lower corner of the preparation, and go back again from right to left, following the serpentine track indicated above in Fig. 7. To move the lens just one field's diameter we have only to fix the eye on a cell at the extreme edge of the field, and then move the stage till that cell disappears out of sight on the oppo- 1 Of late I have used dry lenses a great deal the 7 or even the 5 of Leitz on account of their larger field. Most cells can he easily recognized with this power after we have well learned their looks by earlier study of specimens with the immersion lens. If in doubt about any cell, it is easy to pull out the tube of the microscope or put on the immersion lens. DIFFERENTIAL COUNTING. 47 site side of the field. Thus we avoid any chance of counting the same cells twice, and yet are sure not to miss seeing any. As we go back and forth in this way, we notice chiefly the white cells of course, but yet keep our eyes open for any unusual appearances in the red cells. Usually these move by in a monotonous stream, one looking much like another, but in pathological blood we must always be on the lookout for nu- cleated red cells, degenerative changes, and variations in size and shape. In malarial cases of course our scrutiny is directed chiefly upon the red cells. If we have not the help of a movable stage we try to do the same thing moving the slide with the fingers. With moderate care there is no danger of counting the same cells twice, but we cannot help missing a good many altogether, so that although accurate the process takes longer. When leucocytosis is present, at least one thousand leuco- cytes can be found in a single well-spread seven-eighth-inch cover- glass specimen. In normal blood we may need to go through two to three covers. BACTERIOLOGICAL EXAMINATION. Blood obtained by the ordinary method of puncture is not fit for bacteriological examination. 1 The following is the better way: Sterilize the skin over the flexor surface of the bend of the elbow, and wash off thoroughly the agents used for sterilization with boiled water or boiled normal salt solution. Have an as- sistant grasp the upper arm so as to prevent the venous return and distend the large veins at the elbow. Into the most promi- nent of these plunge a sterilized hollow needle connected with the bulb of a sterilized syringe. All traces of antiseptics must be carefully washed out of the needle and the syringe bulb be- fore using. When the needle penetrates the wall of the vein the blood usually begins to flow into the bulb of the syringe, and this is hastened by gently withdrawing the piston until 1-2 c.c. of blood are in the bulb. Then withdraw the needle, press a pad of sterilized gauze over the wound, and expel the blood before it 1 See Kiihnau's comparative experiments in Deut. med. Woch., 1897, No. 25. 48 CLINICAL BLOOD EXAMINATION. coagulates into a blood-serum culture tube so that it shall run down over the whole surface of the " slant" and collect a little at the bottom. The tubes are then put at once into the thermostat. In examining for the gonococcus the blood is to be mixed with equal parts of agar-agar (previously melted down so as to be mixable but not hot enough to kill the organisms) , and then plated. The farther examination of cultures falls outside the scope of this book. In the above procedure the only difficulties are: 1. Some- times it is hard to find a vein and to get the needle into it. 2. Occasionally we get the needle entirely through the vessel into the tisuses on the other side. If the blood does not flow readily into the bulb one of these two mistakes is usually the cause, but occasionally in those whose vessels are very small or whose circulation is very feeble (as in the moribund) it is very hard to get the requisite amount of blood. Only practice helps us to avoid these difficulties. The procedure causes hardly more pain than the use of an ordinary subcutaneous injection ; the process of sterilization is usually more irksome to the patient than the puncture. Bleeding is trifling, and within twenty-four hours there is usually no trace of the puncture left. A sterilized dressing with moderate pressure should be applied. OTHER METHODS or BLOOD EXAMINATION. It is perhaps worth while briefly to mention some other methods of blood examination of which no account will be given. 1. Determination of the alkalinity of the blood. No accur- ate and clinically available method has yet been devised. De- spite the interesting work of Kraus, Caro, Lowy, Biernacki, v. Limbeck, and others, *I am still unable to get hold of any clinically valuable information given by the determination of alkalinity. 2. Resistance of the red corpuscles to the influence of dis- tilled water. As is well known, water breaks up red cells, but if we add a certain amount of alkali, say NaCl, the cells re- main uninjured. The amount of NaCl which has to be added OTHER METHODS. 49 to prevent the destruction of red cells is from 0.44 to 0.48 per cent. Under certain pathological conditions it needs either more or less of the salt to keep the cells intact, i.e., they pos- sess an increased or diminished power of resistance against the destroying influences of distilled water. The degree of con- centration necessary to maintain red corpuscles intact is known as the isotonic coefficient of the blood as stated in terms of a given salt; 0.44-0.48 is thus the coefficient of normal blood cor- puscles in NaCl. Possibly this method of examining blood may in the future give us knowledge of clinical value. At present it is not clini- cally applicable. The resistance of the blood cells to the influence of elec- tricity, heat, and mechanical pressure has also been investigated in various conditions of health and disease. 3. The rapidity of coagulation varies markedly in different diseases, but no reliable way of measuring it has yet been found. 4. The amount of solids in a given quantity of blood can be determined by weighing a given amount of blood before and after six hours' drying at 65 C. Inasmuch as the haemoglobin percentage and the specific gravity run practically parallel with the amount of solids this method has no considerable clinical value. 4 PART II. PHYSIOLOGY OF THE BLOOD. CHAPTER IV. ONLY such portions of our knowledge of blood physiology will be entered upon here as are necessary for an understanding of the small group of pathological changes which can be profit- ably investigated by clinicians. This limits us for the present to the morphology of the blood, its coloring matter, and its density under physiological conditions. APPEAEANCE OF FBESH NORMAL BLOOD. A drop of normal blood spread between slide and cover-glass as directed on page 7 and examined immediately with a one- twelfth immersion lens, amazes us first of all by the entire ab- sence of any red color. All we see is a colorless liquid in which masses of very pale greenish-yellow discs are floating or lying. . I. Red Corpuscles. (a) If the blood is spread thickly the blood discs are often arranged in the form of rouleaux (Fig. 24). The entire absence of this tendency to rouleaux formation is pathological. It is to be avoided, of course, as far as possible, as it gives us only the thin edges of the corpuscles to look at and covers up much that we need to study. Thin spreading of the blood is therefore im- portant. (6) There is not much variation from the accurately round shape of each corpuscle in normal blood, except where one is indented by another. As they are moved about by the currents set in motion by the gradual drying up of the plasma and strike against each other, they bend, double up, or indent each other, PHYSIOLOGY OF THE BLOOD. 51 like bags of jelly, but yet always have a strong tendency to return elastically to their round outline when free from pres- sure. Thus a corpuscle passing through a narrow passage be- tween two leucocytes will be flattened out like a worm ; but as soon as it emerges on the other side, it. will be as round as before. (c) The central biconcavity of the cell, being thinner than the rim, is lighter colored. Just how much lighter should be learned by practice so that we may detect any abnormal pallor of the corpuscles due to lack of haemoglobin. Pallor is to be seen mostly in the centre of the cell, which in extreme cases seems almost transparent. This is not to be confounded with the highly refractile, glistening-white centres seen as a mark of necrosis as soon as the blood begins to dry up. A fuller de- scription of these appearances is given in the chapter on the malarial organisms, with some forms of which they may be confounded. (d) Slight variations in size are present among normal red discs, and here again only practice can teach us where the normal limits end and the pathological begin. Cells may be (patho- logically) all undersized or all oversized, so that a standard of comparison is not always to be looked for in the preparation itself. (e) If we focus carefully on a single red cell we can usually make out a fine, wavy, so-called molecular motion in it. This is quite different from the active amoeboid movements observed in dying cells, and from the rapid dancing of malarial pig- ment. (/) The familiar appearance of spines all over the cells usually called " crenation" need not be described here (see Fig. 27, p. 86). But it is the very earliest beginnings of crenation that lead to mistakes, as when only one projection has been developed and that points toward the eye, so that a bright spot in the cor- puscles is all we see. ( g) Unless we disinfect the skin before puncturing we must be prepared to find in fresh preparations (a) oil drops 1 (b) epi- thelium; (c) particles of "dirt;" (d) small colorless motile ! In some conditions the blood really contains fat. (Vide infra, "Li- paemia.") 52 CLINICAL BLOOD EXAMINATION. organisms about 1 n in diameter, which are not at all rare but whose nature is unknown to me. l (h) We may make a rough estimate of the number of red cells present if we take care to spread the drop of the same thickness each time. The eye gets used to the ordinary look of a well- filled field of corpuscles and notices a look of thinness if any considerable anaemia is present. (i) The degenerative changes to be seen in normal blood after long exposure to the air, which can get in between slide and cover, are described in detail later on. In pathological blood we may find these as soon as the blood is drawn. //. White Cells. (a) The white or colorless corpuscles are but little different from the red in color, the latter being so nearly colorless. We first notice them either by their amoeboid movements, or because they are not moved by the plasma currents, but stand like a rock round the sides of which the current of red cells is broken. They are slightly larger in most instances than the red cells ; but this difference shows less in the fresh specimens where the leu- cocyte keeps its spherical shape than in the dried and stained preparations, where it is usually somewhat flattened. Their shape is very irregular and their edges often look tattered. In some leucocytes the amoeboid motions are entirely absent. These are the smallest cells, and in them a single nucleus filling most of the cell can often be seen. They are much more nearly spherical and less irregular than the amoeboid cells. The large amoeboid leucocytes are more or less granular, and in certain lights these granules look quite dark and are some- times mistaken for bits of malarial pigment. This is especially true of the coarse granular cells seen occasionally; staining shows these large granules much more distinctly ( = eosino- phile see below, p. 65) ; cells of this type are the most actively amoeboid of all. (b) The most important point in connection with the leu- cocytes is their ratio to the red cells. This is estimated in fresh specimens not by any actual counting but by reference to a standard fixed in the mind by study of normal specimens, and 1 Since this was written the same appearances have been carefully studied by Mtiller and Stokes (see page 59) . PHYSIOLOGY OF THE BLOOD. 53 any considerable increase of the white cells would be noticed at once. Naturally we must not judge from any one part of the slide, as the distribution of the leucocytes may be unequal in different parts of it. ///. Blood Plates. 1 Unless the number of these elements is increased by some pathological influence, we seldom notice them at all in normal blood. This may be because we do not work quickly enough in preparing our specimen. Hayem recommends that the cover- glass be laid upon the slide before the puncture is made ; as soon as the drop emerges it is allowed to run in between slide and cover by capillary attraction, thus avoiding contact with the air. a The blood plates are irregularly shaped, very cohesive elements, about one-half the diameter of a blood disc, usually seen cling- ing together in masses like zooglcea. They are colorless and not amoeboid and look like debris. IV. Fibrin Network. After a specimen of fresh blood has stood for some time ex- posed to as much air as can creep in between slide and cover- glass, we begin to notice a network of fine straight lines in the spaces between the corpuscles. Here and there these filaments seem to radiate from a centre where irregular, colorless masses, apparently blood plates, are to be seen (Fig. 24). No stain is needed to demonstrate these fibrin threads, but a small-aperture diaphragm and very little light makes them plainer. Their only importance is that under certain pathologi- 1 It is probable that the elements included under this heading comprise several different things. It is beyond the plan of this book to discuss their origin and significance, since they possess at present no clinical value. 2 This is a very satisfactory way if we wish to see the corpuscles as fresh and unspoiled as we can. Put a cover-glass on a slide so that the edge of one corresponds with the edge of the other, and, holding them in this position with finger and thumb, put their superimposed edges into the side of the drop as it emerges. It will run in between them by capillary at- traction. The blood plates can be stained with eosin, and in the eosin- haematoxylon stain are easily seen and their number approximately estimated. 54 CLINICAL EXAMINATION OF THE BLOOD. cal conditions the fibrin network is very much increased and helps us in the diagnosis (Fig. 25) . Hence it is of importance FIG. 24. Rouleau Formation and Fibrin Network of Normal Blood. to be familiar with the ordinary closeness of the network in nor- mal blood as a standard of comparison. FIG. 35. Increased Thickness of Fibrin Network. For an account of the conditions of its increase see Chapter IX., page 124. PHYSIOLOGY OF THE BLOOD. 55 AVERAGE DIAMETER OF RED CELLS. The blood under normal conditions shows considerable varia- tions in the size of its corpuscles in the fresh state as well as in stained specimens. * The following table (v. Limbeck) shows the results of various observers. Normal Limits. Average Diameter. Welcker diameter = 4. 5-9. 5 //. 7 p, Valentin In Malinin 7.7 ft Hayem diameter = 6-S.8//.. 7.5^ Mallassez 7.6 fi Laache diameter = 6-9 v 8. 5 n Bizzozero 7. 075 n Gram . , . . diameter = 6. 7-9. 3 /* 7. 850 n Average = 1.5 n These differences depend partly on differences in the method of measuring (wet or dry), and partly on the fact that the age 1 A method of measuring, approximately accurate, and easily appli- cable in clinical work is the following : Using a camera lucida, trace on paper the divisions of a fine stage mi- crometer as seen under a one-twelfth oil immersion lens ; such micrometres are usually ruled to one one-hundreth of a millimetre. Approximate ac- curacy in our tracing can be obtained if the process is repeated till the divisions marked in successive drawings correspond accurately one with another. Care must be taken that the paper is flat upon the table beside the microscope, and. not raised on a block or otherwise ; also that the part of the paper on which we draw should be perpendicularly under the centre of the mirror and not off to one side. When a drawing has been made with these precautions, we have only to divide the space between each of the lines in our drawing into ten equal parts, and we have a scale, each division of which represents 1 // as seen under a one-twelfth oil-immersion lens, with the length of tube of the particular microscope used. To use our //-scale we have only to draw with the camera lucida any cell whose size we want to know, using always the same microscope, the same length of tube, and the same lenses, and having the drawing paper (as before) flat on the table and perpendicularly under the mirror. The drawing thus made is measured with the //-scale like any other object. With this method a cell can be measured in a few seconds and with sufficient accuracy (i.e., within 0. 5 //). 5G CLINICAL BLOOD EXAMINATION. and conditions of nutrition in the persons selected make a difference. In the new-born, and to some extent throughout childhood, the normal limits of variation are wider than in adults (3.3-10.5 />-, Hay em) . Sex appears to have no constant influence. Gram ' noted that the measurements published by observers living in southern Europe are smaller than those of northern Europe (Italians 7-7.5, Germans 7.8, Norwegians 8.5). The majority of any individual's red cells are certainly about 7.5 I* in diameter, and this may accordingly be taken as our stand- ard (Hay em counts twelve per cent under 6.6 /*, twelve per cent over 8 /*, the rest 7.5 />-) NORMAL NUMBER OF THE RED CELLS. 1. At the level of the sea and in adult life the normal number of red cells per cubic millimetre is about 5,000,000 for men and 4,500,000 for women. This is not infrequently increased in very vigorous, healthy persons ; 6,000,000 is by no means rare among healthy young men, and higher figures are seen occasionally. Thus Hewes 2 in fifty young medical students found an average of 5,809,000 per cubic millimetre; of these fifteen exceeded 6,000,000, the highest being 6,400,000, while the lowest of the whole series was 5,120,000. Altitude above the sea level raises the count invariably (see page 79). 2. The influence of menstruation, childbirth, and lactation is to diminish the red cells temporarily, the amount of the diminution depending not only on the amount of blood lost but on the capacity of the individual organism for blood regeneration. At puberty, when sexual functions are being established, we expect lower counts than after the establishment of the function. Normal pregnancy does not affect the count of red cells. 3. The count of red cells per cubic millimetre is raised by any cause inducing concentration of the blood, such as profuse sweating, and is lowered by the temporary dilution of the blood after large draughts of liquid. In these changes, which are al- ways very transient, the haemoglobin and specific gravity in a given drop are of course increased with the corpuscles. 1 Fortschritte der Medicin, 1884. 8 Transactions of the Boston Society of Medical Science, May 18, 1897. PHYSIOLOGY OF THE BLOOD. 57 Vasomotar influences affecting the calibre of the peripheral vessels (hot or cold baths, exercise, etc.) may temporarily con- centrate or dilute the blood by affecting the interchange of fluid between the vessels and the surrounding lymph spaces. By these processes the blood in the peripheral vessels may show an increase or diminution in the cellular elements, the haemoglobin and specific gravity corresponding to the greater or less concen- tration of the blood at that point (on these points see below page 76). Hayem noted that in young people especially the number of red cells varied considerably without any notable change in con- ditions. 4. Influence of Nutrition on the Number of Bed Cells. A. After a meal, especially when considerable liquid is taken, the blood is temporarily diluted and hence the count of red cells per cubic millimetre is diminished (v. Limbeck; Reinert). This is illustrated by the following case from v. Limbeck. ADULT, MALE, HEALTHY. Red Cells. White Cells. Hb 11 :15 A.M 5, 530, 000 7, 660 98 per cent. 12 M. dinner. 12:15P.M 5,320,000 6,166 1:15 " 5,480,000 8,500 2:15 " 4,733,000 12,000 3 :15 " 4, 872, 000 14, 000 89 per cent. 4:15 " 4,720,000 10,830 89 As the white cells rise (digestive leucocytosis, see below, page 83) the red fall. Fasting, by concentrating the blood, temporarily increases the number of red cells (400,000-500,000 increase after twenty- four hours' fast). B. General Nutrition. Lean, muscular people have on the average more red cells per cubic millimetre than fat people (Leichtenstern, quoted by v. Limbeck), other things being equal. * As above said, fasting (by concentrating the blood) raises the number of red blood cells, so that it is not simply hunger that 1 The influence of stasis in the obese, whose fat loads the surface of the heart, is to cause an apparent increase of red cells (see below, p. 75). 58 CLINICAL BLOOD EXAMINATION. gives us the diminution in red cells commonly found in poorly nourished people, but rather the influence of bad hygiene in the slums, etc. 5. Seasons and the time of day seem to have no influence in themselves. The same is true of race and climate. The only exception to this is reported in the work of E. Below, 1 who found in yellow fever districts an average count of only 4,700,000 red cells per cubic millimetre and the diameter of the individual cell reduced to 5.9 /* on the average (7.5 /* = normal). 6. Fatigue. Hayem noted a loss of from 500,000 to 1,000,- 000 red cells per cubic millimetre in the blood of a number of farmers after a hard summer's work, the counts made in September having been compared with those of April and al- ways found to be lower. Whether fatigue is the only cause of this diminution may be doubted. 7. Age. In the new-born the number of red cells is very high for a few days (7,000,000 to 8,800,000), but falls at the end of seven to ten days (see page 86) . In the very old a certain degree of anaemia is, so to speak, physiological; but this, which like the plethora of the new- born is to be referred not to the fact of age, but to concomitant influences, is by no means invariable. Schmaltz reports 6,766,000 red cells in a man of eighty-one and 4,816,000 in a woman of seventy -four. NOBMAL NUMBER OF WHITE CELLS. The figure usually given for adults is 7,500 per cubic milli- metre. This varies a good deal, according to the nutrition of the individual (see page 81) and also at different times of the day, owing to influences not explained. The influence of digestion will be mentioned later. In animals a slight shock 2 is sufficient materially to affect the count of leucocytes; 5,000 to 10,000 may be called the normal limits. Eomberg finds 9,058 as the average count in fifty -five healthy young women. There is, I believe, no evidence to show whether or not mental disturbances (fear, rage, emotion of various kinds) affect their 1 "Deut. Tropenhygiene," Berlin, 1895. O. Coblanz. 2 L6witt: "Studien z. Physiol. und Pathol. d. Blutes," etc., Jena, 1892. Fischer. PHYSIOLOGY OF THE BLOOD. 59 number, but my impression is that they do. Other causes of variation will be discussed under Leucocytosis. BLOOD PLATES. The number of blood plates is from 400,000 to 700,000 under normal conditions. They are the chief constituents of white thrombi, and wherever they are diminished (e.g. t , in haemo- philia, purpura) clotting is apt to be slow. They are increased in leukaemia and in many cases of grave anaemia. In the severer types of many infectious diseases (typhus, erysipelas, malaria) they are diminished, and in malaria they are sometimes wholly absent during the fever. In pneumonia and tuberculosis they are normal or increased. In purpura and haemophilia they are sometimes much diminished or absent. The physiological limits of the amount of hcemoglobin and of the specific gravity have already been mentioned. Under phys- iological conditions their variations follow those of the count of red cells. MULLER'S "BLOOD DUST." Miiller 1 has recently described under the title of " Haemo- conien," or blood dust, a constituent of normal and pathological blood not hitherto noticed. This consists of small round color- less granules about the size of the finest fat drops or about i-1/* in diameter, their size being very variable. They are highly refractile and have rapid dancing (molecular) motion, but no power of locomotion. They are insoluble in alcohol and ether, not stained by osmic acid, and take no part in the forma- tion of fibrin. Stokes and Wegefarth, 2 who have confirmed Miiller's observations, note that the " blood dust" can be seen much more clearly by the light of a Welsbach gas burner than by daylight. The latter observers present a body of evidence tending strongly to show that these bodies are the extruded granules of neutrophilic and eosinophilic leucocytes. Granules apparently identical with them can be stained in fresh speci- mens with eosin or Ehrlich's triacid stain in a way apparently 'Centralbl. fur allg. Path., etc., viii., 1896. 2 Johns Hopkins Hospital Bulletin, December, 1897. 60 CLINICAL BLOOD EXAMINATION. like that of the intracellular granules. ! They are also to be seen in pus and in hydrocele fluid. I have frequently noticed these granules in studying fresh blood, but hitherto supposed them to come from the patient's skin (see page 52, footnote). No special diagnostic or prog- nostic significance has yet been attached to them, though the work of Kanthack and Stokes renders them of great interest with reference to the problem of immunity. They remind one somewhat of the description given by Kahane of the supposed organism of cancer. 1 Nicholls: Phil. Med. Jour., Feb. 26, 1898. Examination of the Blood. PLATE I. Fig. i. Varieties of Leucocytes. Polymorphonuclear ^ &2 H\ L-Myelocytes Small Lymphocytes --Large Lymphocytes Eosinophile M|*.. Eosinophilic "Myelocyte The Malaria! Organism. N J -;; v /5 6 C'$: ijf v :.v- M^ sr. ,* Fig. 2. Fig. 3. B. C. < 'ahf.t fee. I.itli. AIIHI v. !;. A. Fank4, L.ipzig. PLATE I. FIG. 1. (a) Polymorphonuclear Neutrophiles. Note the varieties in size and shape of granules, the irregular staining of the nuclei, the light space around them, their relatively central position in the cell. (&) Myelocytes. Note identity of granules with those just described ; the even, pale stain of nuclei ; their position near the surface (edge) of the cell. The two cells figured indicate the usual variations in the size of the whole cell. (c) Small Lymphocytes. In the cell at the left note transparent proto- plasm ; in the cell next to it note very pale pink ring of protoplasm around nucleus which is deeply stained, especially at the periphery. The next cell has an indented nucleus ; its protoplasm relatively distinct. The cell on the extreme right shows no protoplasm and is probably necrotic. In all note absence of granules with this stain. With basic stains blue granules appear in the protoplasm. (d) Large Lymphocytes. Note pale-stained nuclei and protoplasm, ir- regularity of outline ; indented nucleus in one. Every intermediate stage between these and the "small" lymphocytes occurs, and the distinction between them is arbitrary. (e) Eosinophile. Note irregular shape, loose connection of granules, their copper color, their uniform and relatively large size, and spherical shape. (/) Eosinophilic Myelocyte. Note similarity to (6) ordinary my elocytes except as regards granules. Color of granules may be as in (e) ordinary eosinophile. All the above were stained with the Ehrlich triacid stain, and drawn" with camera lucida. Oil-immersion objective one-twelfth and ocular No. iii. (Leitz). F IG< 2. Malarial Parasites in Fresh (Unstained) Blood (Tertian Forms) . N, JV, normal red corpuscles ; 1, red cell containing hyaline body ; 2, 3, 4, 5, successive stages in the development of the parasite, showing acquisition of pigment; 6, 7, full-grown parasites, the corpuscle no longer visible ; 8, beginning of segmentation ; 9, segmentation. In 6 and 7 note brownish blur behind the pigment dots. Drawn as in Fig. 1. FIG. 3. Tertian Parasite Stained with Eosin and Methyl Blue. The remains of the corpuscle containing the parasite stain pink, the parasite blue, and its pigment black. The stages of growth correspond with the numbers attached. Note in Figs. 1, 2, 3, and 4 the shape of the parasite, shown better than in fresh specimen. [Owing to a mistake the cells in Fig. 3 are not drawn according to a single scale and their relative sizes must be disregarded.] CHAPTEE V. FINER STRUCTURE OF THE BLOOD. I. APPEARANCES or DRIED AND STAINED SPECIMENS. COVER-GLASS specimens prepared and stained as above di- rected give us more information of interest and importance than can be obtained from any other one method of blood examination. Approximate ideas of the quantity of red cells, of white cells, and of haemoglobin can be formed, parasites and bacteria can be seen, and the whole mass of evidence based on the finer structure of the leucocytes can only be obtained in this way. The ap- pearances of a specimen of normal blood prepared in this way are as follows : RED CELLS. 1. The haemoglobin stains with the orange G of the tricolor mixture, and in a properly heated specimen the red cells are of a brilliant yellow or pale orange tint. If overheated they have a feebly stained, washed-out look, while if underheated they are more or less broivn or gray. 1 The degree of pallor of the centres corresponding to the amount of haemoglobin in the corpuscle can be gauged much more accurately with this stain than in the fresh preparations. The color of the edges is not much affected by pathological changes, the centres being the test. But in cases with extreme poverty of haemoglobin the colored rim may be reduced to a mere shell and the rest may be almost completely colorless. The power to estimate the amount of coloring matter in this way can be easily acquired. An approximate idea of the number of red cells may be formed by any observer who has learned to use a uniform technique in each case and to spread the blood of a standard thickness. 1 This is a fruitful source of error. Many suppose that because their specimens come out too dark they must be "burnt," and so heat less. In. fact the dark tint means that the specimen is not heated enough. 62 CLINICAL BLOOD EXAMINATION. 2. Nothing is seen of the fibrin or blood plates as a rule. In normal cases the plasma does not stain at all. A certain amount of debris is often present, usually pink stained. WHITE COKPUSCLES. 3. The chief purpose and use of the " triple stain" is for dis- tinguishing the varieties of white corpuscles, and the pathologi- cal states of the red. About the normal red cells it gives us no information that cannot be obtained as well by various other stains, but our knowledge of normal leucocytes has been im- mensely enlarged by its use. In normal blood stained as above directed, we recognize the following varieties of white cells : (1) Small lymphocytes or " small hyaline forms" (Kanthack) (see Plate I.). These consist mostly of a round blue nucleus about the size of a red cell, and surrounded by a thin coating of protoplasm, faintly stained or invisible (with Ehrlich's triple stain). The nucleus may be considerably smaller than a red cell, and may or may not be deeply stained. In my experience it is usually pale-stained, but slight differences in technique will greatly affect its staining power. The larger it is the more apt it is to be pale (see Plate I. ) . (2) There is no line to be drawn between this form and that now to be described, namely, the " large lymphocyte" or " large mononuclear cell," which is simply larger and paler. The small lymphocyte is the form most frequently seen in the lymph channels and in chyle and at the periphery of the follicles of adenoid tissue. Whether it grows in the circulating blood into any other form of leucocyte is uncertain. In the so-called "large lymphocyte" the nucleus occupies relatively less of the cell than in the small lymphocyte. In many cases we do not see in the blood any intermediate forms. Lymphocytes are either "small" (5-10 /* in diameter) or "large" (13-15 ;>- in diameter) (see Plate I.). In other cases we find every intermediate size, both of nucleus and of the cells as a whole, and in such cases it is absurd to attempt a division into "large" or "small," though we may be able to say in a general way which size predominates. FINER STRUCTURE OF THE BLOOD. 63 The theory that the " large" mononuclear cells or " large hya- line forms" (Kanthack) come from the spleen and the small mononuclear from the lymph glands has been abandoned on all sides of late years. The protoplasm of all lymphocytes, as has been said, is always hard to stain with Ehrlich's triple stain. Sometimes it has a faint pinkish tinge, more frequently it is grayish or very light blue, and in some cases it .stands out brilliantly trans- parent and colorless against the faint purplish tinge of the sur- rounding plasma (see Plate I.). Although non-granular with the triple stain many of the lymphocytes show basophilic granules at the periphery of their protoplasm when stained with eosin and methylene blue. When thus stained a ring of un- stained protoplasm appears around the nucleus. Sometimes the protoplasm stains diffusely blue with basic stains and no granules can be made out. I have described the lymphocytes so far as "mononuclear," but it is not rare to find even very small ones (6 /* in diameter) whose nucleus has a deep cut in one side or has divided into two parts. I believe it is commoner to find a divided nucleus in the small forms than in the " large lymphocytes." The inap- plicability of the term "small mononuclear cells" or "large mononuclear cells" to this variety of corpuscle is evident. The distinguishing mark is not the single nucleus but the absence of granules, with Ehrlich's stain. (See also below under "Mast cells.") In the smaller forms of lymphocytes the nucleus, even when dividing, is compact and fills most of the cell. But in the large forms, instead of simply being larger and paler, the nucleus may begin to bend and branch in the cell, and then we get the so-called (3) " Transitional forms" (Ehrlich), which are no bigger than the larger size of -lymphocytes, from which they differ only in that they have an indentation in their nucleus either a narrow cut or a bay so wide that a " horseshoe" nucleus results. This is the transitional form according to this nomenclature. There is no reason for calling it so, as all the forms of leucocytes are transitional, but there is some convenience in the name. Like most large lymphocytes it is pale all through pale in both nucleus and protoplasm and often escapes notice in hasty ex- 64 CLINICAL BLOOD EXAMINATION. animations. Sometimes its protoplasm is sparsely covered with faint neutrophilic granules. (4) The cells usually known as " polynuclear" are more properly called polymorplionuclear neutrophiles. These cells constitute the vast majority of those found in ordinary pus. The main difference between them and those last described is in the possession of granules, best seen when stained by Ehr- lich's methods. The nucleus stains usually quite deep blue or greenish-blue, and irregularly, i.e., more intensely in some parts than in others. It is very irregular in shape, being twisted about in the body of the cell. Here and there it may dive down so deeply beneath the surface of the cell that it is hidden under a thick layer of granules, reappearing in another part of the cell so that it seems to be broken in two. Occasionally, no doubt, this is actually the case, but generally there are " underground connections" between the apparently separate pieces of nucleus. Now and then we see a cell (degenerating) where the granules have fallen away, leaving the nucleus like a short, thick snake, very rarely two, or like several sausages joined by strings. One never sees any two of these cells whose nuclei are of the same shape. Hence the term "polyrnorphonuclear." The windings and twistings of the nucleus have suggested compari- sons to the letters Z, S, E, etc. The granules which fill the body of the cell and in which the nucleus is embedded stain well only with triple stains like Ehr- lich's. Acid stains like eosin, and basic stains like methylene blue, do not bring them out clearly. Hence the term "neutro- philic," which is not strictly accurate; more properly they are faintly oxyphilic J and can be faintly stained with eosin. [Hence Kanthack and other English observers have called them " fine granular oxyphiles," while the term "coarse granular oxy- philes" is applied to the cells generally known as eosinophiles. These terms are in some respects more accurats than Ehrlich's, but are even more cumbrous than his.] With Ehrlich's triacid mixture the granules stain violet or purple, sometimes pink. They are very small and irregular in shape and size, contrast- ing with, the large, round, " eosinophile" granules (see below). The cells being spherical the granules lie over and around the nucleus, not simply at the side of it. In their interstices we 1 Ehrlich's stain is really a differential acid stain and not neutral. FINER STRUCTURE OF THE BLOOD. 65 sometimes seem to see a pinkish background of cell substance, but this is probably composed of granules somewhat out of focus. In normal blood these " neutrophilic" granules, which are so small that except with very high powers they look like a diffuse stain, ^--. rarely if ever occur except in cells whose nucleus has reached the polymorphous stage. Occasionally we seem to see mono- nuclear neutrophiles, having a round nucleus with neutrophilic granules, but careful focussing usually shows that the appear- ance of a round or rod-shaped nucleus is given by the tight coiling of the ribbon- like nucleus round one of its ends, or else FlG - 26 - that a horseshoe nucleus is seen from the point of view indicated in Fig. 26. Thus if the eye be at the point A the nucleus will appear of the shape indicated in B (Fig. 26). (5) The eosinophile or " coarse granular oxyphile" cell has, like its predecessor, a polymorphous nucleus and granules ; but the nucleus is paler and more loosely connected to the granules, and the latter are spherical or oval, of uniform size, and much larger than any seen in the neutrophilic cell. They have strong affinity for acid coloring matters (eosin, acid-fuchsin, etc.), hence their name. In specimens stained with eosin or eosin and methyl blue they are very brightly colored pink. With the Ehrlich triacid mixture they are more of a copper or burnt-sienna color. Some individual granules stain much darker than others in the same cell, and may be of a different nature. The eosinophiles are the most actively amosboid of all the corpuscles, and it may be for this reason that the different parts of the cell seem so loosely strung together. Or, if the hy- pothesis of Kanthack and Hardy, recently supported by Stokes, be true their loose arrangement may serve to make them easily detached for bactericidal purposes (see above, page 59) . The granules may be all at one side of the cell and the nucleus on the other, and in cover-glass specimens we very frequently find actual separation of the two. Whether or not the actual sepa- ration is brought about by the technique of spreading the blood is unimportant, as we find such broken cells much more often 5 66 CLINICAL BLOOD EXAMINATION. among the eosinophiles than among any other variety which argues a looser structure. Sometimes there seem to be two or more distinct and separate nuclei in the cell, no " underground connection" being traceable. The granules are seldom over the nucleus as we see it in cover-glass preparations, but cluster round it loosely. The cell as a whole is usually a little smaller than the " neu- trophile" and more irregular in shape. In stained specimens the neutrophile is seldom seen with a pseudopod extended, whereas the eosinophile often shows it. The staining of the nucleus is more even as well as paler than that of the neutrophile, and with the Ehrlich stain often has a robin' s-egg tint. To sum up: The four varieties which we usually find among leucocytes in the blood are these : 1. Small lymphocytes, or "hyaline cells." 2. Large lymphocytes and transitional forms. 3. Polymorphonuclear neutrophiles, or fine granular oxy- philes. 4. Eosinophiles, or coarse granular oxyphiles. 5. A fifth variety of leucocyte the basophilic " mast cell" has lately been described as a constituent of normal blood, though in very small numbers. In leukaemia it is very common, but no special significance is attached to it. With Ehrlich's stain the basophilic granules of this cell are not seen or appear only as clear white spots. Stained with the following solution they are easily seen : Dahlia (saturated alcoholic solution filtered) , . . . 50 Glacial acetic acid, . . ... . . . 10-15 Distilled water, . . , .... . v . 100 Covers should be left twenty-four hours in this mixture, then washed and mounted in the ordinary way. The nucleus is usually trilobed. Though very common in all connective tissue as well as in the wall of the intestine and the serous cavities, these cells rarely stray into the circulating blood. TEEMS. No one can feel more unsatisfied with the terminology used in this book than the writer. It rests partly on a theory of the origin of the cells ("lymphocytes"), partly on the properties of FINER STRUCTURE OF THE BLOOD. 67 the nucleus ("polymorphonuclear"), and partly on affinities for aniline dyes (" neutrophile" " eosinophile") . All that can be said for it is that it discards certain very misleading names like " splenocy te" (a term applied by some to the large lymphocytes according to the now exploded theory that they come from the spleen) , or like " small mononuclear" to designate cells not rarely polynuclear. The cumbrous word " polymorphonuclear" is a shade better than "polynuclear," and that is all to be said in its favor. It is greatly to be hoped that we may ere long have a new and improved terminology by some competent student. The English terms above referred to are so cumbrous that I have not as yet felt compelled by their slightly greater accuracy to adopt them unconditionally, but they certainly show a tendency in the right direction. For some unknown reason we do usually find the leucocytes of the blood only in these four forms (the frequent presence of transitional stages between "small" and "large" lymphocytes has been mentioned as an exception). There are far too few transitional forms between the four varieties to be seen in the circulating blood for us to suppose that they grow from one type to another there. It may be that they all have a common leucocyte ancestor and are "specialized" in various extra- vascular tissues into the forms which we meet with in the blood. There seem to be well-marked sets of leucocyte forms adapted respectively to the blood, the serous spaces, the intes- tinal wall, the marrow, the connective tissues, and the adenoid tissues, as has been well shown by recent English observations. NORMAL PERCENTAGE or EACH VARIETY. In the blood of healthy adults the proportions of the differ- ent varieties above described are the following : , , j Small lymphocytes, 20-30 per cent. ( Large " 4-8 (6) Polymorphonuclear neutrophiles, . . 62-70 " (c) Eosinophiles, ....... -4 " (d) "Mast cells," ...... tiri (a) In infancy the percentage of lymphocytes is much larger 68 CLINICAL BLOOD EXAMINATION. (forty to sixty * per cent) and the polymorphomiclear neutro- philes are only eighteen or forty per cent. In a variety of debilitated conditions not usually thought of as definite diseases, the number of lymphocytes is comparatively large and that of the polymorphonuclear cells small. The general vigor and health of the individual can sometimes be estimated simply from the leucocytes. Persons calling them- selves well, but never vigorous or active, may show no more than fifty per cent of polymorphonuclear cells, the lymphocytes running up to forty or even fifty per cent. Not all cases of debility show this change, and we are not yet in a position to say under just what conditions it occurs. It certainly is not peculiar to tuberculosis as Holmes has sup- posed. Presumably the conditions are such as decrease the nutritive value of the plasma. (b) Changes in the percentage of neutrophiles will be dis- cussed later. (c) The percentage of eosinophiles often changes in a way hard to explain. We know that eosinophiles are present in large numbers in various parts of the body outside the blood-vessels (bone marrow, gastro-intestinal tract, ccelomic spaces, thymus gland), and in many ways they seem to live their life in com- parative independence of the' other members of the leucocyte group. In the free interchange of fluid and cells that is constantly going on between blood-vessels and lymphatic tissues and spaces, it is evident that a part of the life history of the leu- cocytes goes on outside the vessels, and there is reason to sup- pose that it is chiefly outside the vessels that cells divide and produce others like themselves. At any rate, we rarely find evidence of mitosis or amitosis in the circulating leucocytes, while in the lymph glands and the marrow, and elsewhere, such dividing forms are common. The bone marrow seems to be such a dividing-place for eosinophiles. They are always numerous there and mitoses are often seen in them. Indeed their number is so small in normal circulating blood that they might almost be said to be there "by mistake," belonging normally elsewhere. Whether or not this has any connection with their active amoeboid properties, I do not know. The " mast cells" are even more " an accident" in FINER STRUCTURE OF THE BLOOD. 69 the blood, and Ehrlich denies that they are a constituent of normal blood. The increase or decrease of eosinophiles in the circulating blood does not follow that of the polymorphonuclear neutro- philes, in fact is often inversely proportional to it, and the eosinophiles are often markedly increased in a blood otherwise normal, for reasons wholly unknown to us. An increase in the lymphocytes or neutrophiles does not occur without other blood changes, and points, not to disease of one place or function, but to general conditions like inflam- mation or malnutrition. The diagnostic indications of an in- crease of the eosinophiles are more specific (vide infra, articles "Trichinosis," "Asthma," "Dermatitis Herpetif ormis, " etc.). I have spoken of the eosinophiles and " mast cells" as com- parative strangers, though not intruders in the circulating bood. They are thus intermediate between the regular inhabitants (lymphocytes and neutrophiles) and the variety next to be men- tioned, which are real intruders i.e., never found in normal blood. These are the MYELOCYTES (EHRLICH). The normal l abiding-place of these cells appears to be the bone marrow, hence their name of myelocytes or marrow cells. They are perhaps the most numerous leucocytes to be found in the marrow, although lymphocytes and polymorphonuclear cells are also to be found there, and eosinophiles and basophiles are numerous. I describe them here because they are peculiar to no one dis- ease and are occasional visitors of the blood in various diseased conditions and in conditions on the borderland between the pathological and the physiological (starvation various intoxi- cations) . The myelocyte (see Plate L), like the polymorphonuclear 'Frankel (15th Congresse fur innere Med., 1897) has reported the find- ing of myelocytes in the swollen lymph glands of a case of scarlet fever. In leukaemia they are found in the metastases and infiltrations in various organs as well as in the blood. Otherwise they are confined to the marrow so far as I know, except that very small numbers may enter the blood in conditions involving leucocytosis or grave anaemia. 70 CLINICAL BLOOD EXAMINATION. neutrophile, is recognizable only by Ehrlich's staining methods. With Ehrlich's triple stain it appears as a spherical cell nearly filled by a large, pale-stained nucleus immersed in neutrophilic granules. One sees at once how little it differs from the large lymphocytes (simply in having granules) and from the poly- morphonuclear neutrophile (only in the shape of its nucleus). Were it present in normal blood we should undoubtedly con- sider it an intermediate stage between the large lymphocyte and the polymorphonuclear neutrophile. I see no sufficient reason for thinking otherwise merely because it does not appear in normal blood. The leucocytes are so cosmopolitan in their habits that we can hardly call them blood cells at all. It is better to think of "blood leucocytes," "gland leucocytes," and "marrow leucocytes" (perhaps "skin and mucous membrane leucocytes" too, see page 117) and to consider that " missing links" among blood leucocytes are to be looked for among those that live and grow up elsewhere. Perhaps the condi- tions in the marrow (rest, nutrition?) are such as bring out sides of the leucocyte nature suppressed in the blood. Staining with eosin and methylene blue shows that myelocytes also con- tain fine basophilic (blue) granules. The suggestion of their transitional nature is thereby increased, since they have baso- phile granules in common with the lymphocytes, and neutrophile granules in common with the polymorphonuclear cells. I am indebted to Dr. H. F. Hewes for calling my attention to this point. With Ehrlich's stain the granules of the myelocyte are precisely those of the polymorphonuclear leucocyte and need no second description (see Plate I.). The nucleus, by which alone we distinguish the myelocyte, shows none of the twists and turns characteristic of the polymorphonuclear neutrophile, but is usually spherical or egg-shaped, and is in close contact with the cell wall for a comparatively large portion of its extent i.e., if egg-shaped it is placed eccentrically. Not infrequently the nucleus shows signs of old age (vacuoles) or of mitosis, for not infrequently we find two nuclei at the poles of the cell. It is then to be distinguished from the polymorphonuclear neutrophile by the fact of its having the nuclei in close contact with the surface of the whole cell for a comparatively large portion of their extent, while in the poly- morphonuclear leucocyte the nucleus abruptly leaves the sur- FINER STRUCTURE OF THE BLOOD. 71 faces again if it chances to approach it. The dividing myelo- cyte is also to be distinguished from the polymorphonuclear neutrophile by the even staining of the nucleus in the former. Size of Myelocytes. Almost every account of the myelocyte which has come to my notice speaks of it as a very large cell, the largest variety of leucocyte ever seen in the blood. This is true of many of them ; diameters of 18-21 P. are not uncommon, but we also find them of every other size down to 10-11 fj. diameter, that is, down to the size of a lymphocyte. This is true both of the myelocytes in the circulating (leukgemic) blood and of those in the marrow. No distinction from other varieties of leucocyte can be based on size alone, unless we say their average size is greater than the average size of the leu- cocyte. Perhaps the following table may be of interest: Average diameter of 100 myelocytes = 15.75 n. " " " 100 poly morphonuclearneutroph lies = 13.50 p. " " 100 "large" lymphocytes = 13 u. a " " 100 eosinophiles = 12 ju. tt *''* 100 "small" lymphocytes = 10 ft. tt u u 100 red corpuscles (normal) =7.5 ;w. EOSINOPHILIC MYELOCYTES. Under the same conditions where we expect to find the or- dinary (neutrophilic) myelocyte, we often find a small number of cells identical with them in all respects, except in possessing, eosinophilic in place of neutrophilic granules. Such cells are found in abundance in the marrow, and this fact together with the resemblance to the ordinary myelocyte both in morphology and in the conditions of their occurrence, seems to me to justify the term eosinophilic myelocyte. COBNIL'S "MARK CELLS." By most observers these are supposed to be the same as Ehrlich's "mark cells" or myelocytes. Cornil worked before the days of Ehrlich's staining methods and therefore before the presence of neutrophilic granules could be used to distinguish a 72 CLINICAL BLOOD EXAMINATION. myelocyte from a large lymphocyte. Corral's description of them would answer for either. Schreiber considers Cornil to Lave discovered a different variety of non-granular cell, but the description of it given by Schreiber seems to me to leave it in- distinguishable from a large lymphocyte. Mononuclear Neutrophiles. Capps 1 observed in general paralysis of the insane a variety of leucocyte possessing a deep-staining centrally placed nucleus like that of a lymphocyte, but containing also neutrophilic granules. He considers it either a variation from the ordinary type of marrow-bred cell visiting the blood temporarily, or more likely an ordi- nary lymphocyte in which the granules have developed before the nucleus has become polymorphous. Thayer has observed similar cells, but has given no explanation of them. Klein 2 mentions them under the name above given and figures them in his plates, but does not comment on them. I am inclined to think them ordinary myelocytes. So far I have described the type cell of each variety. As we should expect, atypical forms are numerous. Some of the com- moner ones are as follows : 1. Small lymphocytes whose nucleus is pale blue instead of dark blue. 2. Large lymphocytes, whose protoplasm has evidently for- saken them (degeneration forms, often more or less deformed or tattered) . 3. Cells on the borderland between the "marrow cell" and the " poly morphonuclear leucocyte, " the nucleus having some of the characters of each variety. 4. Cells tiie nature of whose granules we cannot settle (eosi- nophilic or neutrophilic). Other rare varieties will be mentioned under leukaemia. 1 American Journal of Medical Sciences, June, 1896. 2 Volkmanu's Sammlung klin. Vortrage, December, 1893. PART III. GENERAL PATHOLOGY OF THE BLOOD. CHAPTER VI. UNEQUAL DISTRIBUTION OF BLOOD PLETHORA DILUTION AND CONCENTRATION OF THE BLOOD. I. Unequal Distribution. How far is the single drop used for blood examination typi- cal of the whole? It has been experimentally proved that specimens of the blood of the smaller venous and arterial twigs do not differ from each other materially in corpuscular richness. Capillary blood is slightly richer in corpuscles than that either of veins or of arteries. But as capillary blood is everywhere of the same corpuscular richness, we may consider one capillary net- work or set of venules as typical as another, provided our tech- nique is good, that is, provided lymph is not squeezed into the drop by strong pressure. It is indifferent, therefore, so far as accuracy is concerned, whether the drop of blood be obtained from one or another part of the body. All standard estimates of the number of corpuscles per cubic millimetre of normal blood refer to capillary blood. 2. Apparent Polycyihcemia. So far we are speaking of normal conditions. It is a famil- iar fact, however, that the vessels of a given part of the body can be overcrowded with blood, e.g., by the use of an Esmarch bandage. A drop taken from such a part would certainly not be typical. Now as the same effect can be produced by a variety of diseases, under these conditions we must modify considerably any inferences made from examination of a single drop. 74 CLINICAL BLOOD EXAMINATION. Such conditions, entailing a false polycythaemia or apparent increase in the number of corpuscles are : I. Any disease involving either (a) general cyanosis or (6) cyanosis of the part from which the drop of blood is drawn. (a) General cyanosis results either from cardiac insufficiency (valvular or parietal disease of the heart itself, blocking of the lung circulation by emphysema or thrombosis), from insufficient aeration of the blood (pneumonia, congenital malformation of the heart), interference with the heart's action by pressure of tumors, effusions (pericardial, pleural, peritoneal), or enlarged organs (liver, spleen), or from vasomotor disturbances. It is evident that some of these conditions (e.g., congenital heart disease) may not involve any peripheral stasis at all, and in the absence of this it is not easy to account for the increased number of corpuscles in the drop. Some observers have supposed that there is a real overproduction of blood cells under these condi- tions ; others suppose that the life of the individual corpuscle being lengthened, reproduction of cells at the normal rate soon leads to the "glut." There seems to be no reason to suppose that there is in these cases any unequal distribution of cells in favor of the peripherj^, such as is obviously the condition in ordinary cyanosis with stasis. Whatever the explanation may be, there is no doubt of the fact that general cyanosis from any cause whatever produces an increase of cells in a drop such as we usually examine. The cases of cyanosis which I have classed under " vaso- motor" (for want of a better explanation), cases in which, in the absence of disease in any organ, the skin and mucous mem- branes are persistently and markedly bluish, are not very uncommon. I have seen three such, all in stout, elderly women. In one the cells in a drop of blood from the ear, finger, or toe were more than double the normal number (see below, page 81). (b) Local Cyanosis. The pressure of a tumor, or any other hindrance to the circulation of any part, may give a similar in- crease in the number of corpuscles in a measured amount of blood from that part. Here again vasomotor conditions may cause cyanosis and apparent polycythsemia. In markedly cyanotic patients the count of red cells is notably above normal, we should naturally guess the reason, and make GENERAL PATHOLOGY OF THE BLOOD. 75 allowances. Error is more likely to arise where we have cyanosis in a person whose blood is poor in red corpuscles. The combi- nation of these two factors may give us a normal blood count and lead us to overlook the anaemia. Thus a person might have really a severe anaemia and yet the count of red cells be actually above the normal.' This element of stasis should never be lost sight of. Many high counts reported in pneumonia or hysteria are to be explained by abnormalities not of production or destruc- tion but of distribution of the blood cells. II. Certain patients, whoso circulations are feeble without being feeble enough to produce actual cyanosis, first give us- evidence of the fact by an increase in the count of blood corpus- cles in a given amount of peripheral blood. Following up the hint thus given, one may sometimes be brought to note and in- vestigate an element in the case which might otherwise have- been lost sight of. "With these exceptions the drop of blood taken at the periph- ery is typical. We have next to consider some general condi- tions under which a person's whole blood may be inferred to be abnormal from the findings in a drop taken from the periphery. Consideration of special diseases will follow later. FULL-BLOODEDNESS (PLETHORA) AND ITS OPPOSITE. There is no direct evidence for the existence of any long- standing over-filling or under-filling of the blood-vessels ; there is a good deal of experimental evidence to show that if by arti- ficial means we succeed in forcing into the vessels an abnormal amount of fluid (transfusion of blood or normal salt solution large draughts of water), it does not stay there many hours, but comes out by the kidneys. The red-faced persons popularly known as "full-blooded" show no abnormalities in their blood discoverable by any means of investigation known to us. The condition is probably de- pendent on the presence of a rich capillary network near the surface of the skin, or a dilatation of individual venules and arterioles at the periphery. Such a person may be markedly anaemic without any considerable changes in the color of the face. The fact that people of such complexion often end their lives with a ruptured cerebral artery is due presumably to the- 76 CLINICAL BLOOD EXAMINATION. circumstance that " high living" produces in the same individual dilated peripheral capillaries and weakened arterial walls. Temporary increase or diminution in the amount of fluid with- in the vessels can be brought about not only by a change in the mechanical conditions of pressure and osmosis, but by any influ- ence affecting the tone of the peripheral vessels. We have then: (a) Temporary serous plethora or dilution of the blood from transfusion of fluid in large amounts or its ingestion by mouth or rectum. (b) From decreased blood pressure, as in acute failures of compensation in cardiac disease. (c) From vasomotor dilatation. As an example of this last Grawitz reduced the specific grav- ity of the blood from 1041 to 1038.7 within eight minutes by the inhalation of nitrite of amyl. This decrease of specific grav- ity can only mean an increased amount of watery constituents in the blood, as there was no evidence of any destruction of the heavier elements of the blood, and only water (and chlorides) pass through the vessel walls easily. In the above case the specific gravity was again at 1041 within a few minutes. (d) In cases of severe anaemia which recover, the blood regen- eration may attain such vigor that the number of red cells shoots up above normal, even as high as 7,700,000. This is temporary cellular plethora or polycythsemia. (e) The same condition can be temporarily produced by transfusion of actual blood from one individual to another. It lasts but a few days as a rule. The polycythsemia of the new-born will be discussed later. Concentration of the Blood. It is obvious that influences opposite to those producing temporary full-bloodedness will produce temporary lack of fluid within the vessels. So acute diarrhoea, purgation, deprivation of liquids (as in starvation), rapidly accumulated serous effu- sions, profuse vomiting or sweating (by skin and lungs) produce a temporary concentration of the blood by draining out its diffus- ible elements (water chiefly) . All these influences are transitory. More permanent drains on the system, like chronic diarrhoea, diabetes insipidus or nielli tus, or long-standing suppurations, show no evidence of lessening the volume of blood in the vessels. GENERAL PATHOLOGY OF THE BLOOD. 77 They drain albumin out of the serum and corpuscles and so decrease the weight of the blood (see below, page 85) , but the blood volume is not changed. Indeed, any influence has to work very quickly in order to concentrate the blood, for in an aston- ishingly short time the other tissues repay the vessels their loss of fluid and the normal blood volume is restored. The same temporary effects can be produced by influences con- stricting the vessels (cold, pain, suprarenal extract) , and a concen- tration of the blood results which lasts a few minutes or hours. 1 In all these interchanges of contents between the blood- vessels and the other tissues it is, as above said, the watery ele- ments chiefly that change. The red cells are not affected by the give-and-take of the vessels and tissues, and although cold pro- duces in the peripheral circulation an increase in the number of white cells greater than can be accounted for by simple Concen- tration, the weight of evidence seems to be against any new pro- duction of cells and in favor of a change only in distribution, the white cells accumulating at the periphery. Now as the number of cells is not affected by these tempo- rary variations in the volume of liquid within the vessels, it fol- lows that the number to be counted in a cubic millimetre, though typical of the whole blood at that time, is not to be reckoned from in the ordinary way. For example, after a severe diarrhoea or in phthisis after a night-sweat the blood may be temporarily so concentrated that we find 6,000,000 or more red corpuscles per cubic millimetre. Under normal conditions of the blood mass we should infer from such a count that the body contained one- sixth more red corpuscles than usual. Here obviously it only means (if anaemia is absent) that the blood mass is reduced by one-sixth by concentration. It is only in such sudden reduc- tions of blood volume that we can measure the amount lost by 1 Oliver has shown recently (Lancet, June 27th, 1896) that any influence causing rise of blood pressure will slightly concentrate the blood. Thus raising the arm over the head and holding it there by muscular effort slightly concentrates the blood in that arm. Electrical stimulation or massage of the arm has the same effect. Lowering blood pressure, as when the arm is supported passively over the head, dilutes the blood. This confirms the results of Mitchell (Med. News, May, 1893) and of Cheron (Comptes Rend, de 1'Acad. d. Sciences, 1896, No. vi.). Oliver uses a new method for estimating the number of red cells, the accuracy of which has not yet been tested by others (see above, page 27) . 78 CLINICAL BLOOD EXAMINATION. this method. Long-standing causes of drain on the plasma might at any time act as destroyers of red corpuscles as well, through the changes in the nutritive fluids in which they live. Further, it is only where we know the number of corpuscles just before the sudden drain on the plasma comes, that we can measure the amount of plasma lost by the amount of apparent increase in the red cells. Stasis and any other cause that heaps up corpuscles at the periphery must also be excluded before we can judge of the loss of plasma in this way. The conditions of an abnormal concentration of the blood are those already alluded to as temporarily sucking away its watery constituents, namely: (a) Watery diarrhoea, especially in cholera and other acute diseases accompanied by diarrhoea ; (6) Large and rapidly accumulating serous effusions (slow accumulations would give time for the blood to take up water from the tissues and make up for its loss) ; (c) Profuse sweats; (d) Persistent vomiting or starvation of liquids ; (e) Increased blood pressure (exercise, massage, electricity). Blood already lacking in red cells, if suddenly concentrated by such a loss of fluid, might deceive us into supposing it normal, because the number of cells in a cubic millimetre might be normal. In the presence, therefore, of any such reason for concentration of the blood, we, should always modify our ordi- nary methods of inference from the Hood count. For example, v. Limbeck records a case of hepatic cirrhosis with ascites, where before tapping the ascites the count of red cells was 3,280,000 per cubic millimetre. Within twenty-four hours after tapping there were 5,160,000 cells per cubic millimetre, the reaccumulation of the ascitic fluid going on so fast that the blood was unable to adjust itself and became overconcentrated. A careless observation might have inferred a great gain in the corpuscular richness of the whole blood, when in fact not a cor- puscle has been gained and those present have probably grown poorer in albumin. Dilution of the Blood. Causes of temporary dilution of the blood are less common than those of temporary concentration. GENERAL PATHOLOGY OF THE BLOOD. 79 Immediately after the inhalation of nitrite of amyl or the in- gestion of a large amount of fluid by mouth or rectum, the blood would be diluted so that a blood count would show a diminu- tion in the number of cells per cubic millimetre, which yet would be due to no changes in the number of red cells in the body, and might be wrongly taken for an anaemia. The dilu- tion in cases of heart disease will be discussed later (see p. 296). Any condition involving loivered blood pressure has the effect of diluting the blood by allowing the entrance of perivascular lymph. Summing up the discussion so far : There is no evidence for a chronic plethora nor for a chronic diminution in the volume of the blood. Where such takes place temporarily, it is by the ad- dition or subtraction of water and salts only, and not of the cor- puscles or organic materials, so that we must guard against false inferences from the resulting apparent increase or decrease of corpuscles per cubic millimetre. But although there is no positive evidence of a true increase in the whole amount of blood in the vessels (except temporarily) , there are some conditions which lead to an increased richness of the peripheral blood in red corpuscles even after excluding the influence of stasis or loss of fluid. Such a condition of what appears to be true polycythaemia is found: 1. In persons living at high altitudes; 2. In persons suffering from phosphorus or CO poisoning. 1. The Slood in High Altitudes. The polycythsemia of those living at high altitudes increases the higher one goes. Koppe ' gives the following tables : Place. Height above sea level. Red cells. Author. Christiania. o 4 974 000 Laache. Gottingen 148 metres 5,225 000 Schafer. Tubingen .... 314 " 5,322 000 Reinert. Zurich 414 " 5, 752, 000 Stierlin. Auerbach 425 " 5 748 000 Koppe Reiboldsgriin . 700 " 5 900 000 Arosa 1 800 " 7 000 000 Eerier. The Cordilleras 4 392 " 8 000 000 Viault 1 Munch, raed. Woch., 1890, No. 41. 80 CLINICAL BLOOD EXAMINATION. This extraordinary change takes place within two weeks of the time of taking up residence in a high place, and independent of any change in diet or manner of living. The sick and the well are equally affected and animals show similar changes. The haemoglobin is also considerably increased, although it lags somewhat behind the corpuscles. Koppe states that the individual corpuscles under these con- ditions are so much smaller that their volume in a given amount of blood (as determined by the hsematocrit) is not in- creased at all. On returning to low land, the blood returns within a short time to its normal condition. Many explanations have been offered for this interesting phenomenon. If it were a true new production of corpuscles we should expect some signs of blood destruction (icterus, hsemo- globinuria) on returning to the sea level. But there are no such signs. On the other hand, if the polycythaemia were a simple result of concentration due to the dryness of the high air, one would expect that the blood would quickly adapt itself, as in other (temporary) concentrations, by taking up water from the tissues. But in fact it does not do so. The cause of the in- crease is still a mystery. 2. PhospJiorus and CO Poisoning. The polycythsemia of acute phosphorus poisoning may reach as high as 8,650,000. This may be partly explained by concentration due to the occurrence of vomiting, but in some cases the increase seems out of proportion to the amount of vomiting. With illuminating-gas poisoning there is usually no vomiting to speak of, and the cause of the marked increase in the red cells is unknown. Von Limbeck in two cases of CO poisoning showed respectively 6,630,000 and 5,700,000 red cells. Miinzer and Palma 1 record 5,700,000. The white cells are also in- creased (see page 330). Possibility of a True Plethora. Although there is no direct evidence that the whole blood mass in its relation to the weight of the body ever varies more 'Zeit. f. Heilk., vol. 15, p. 1. GENERAL PATHOLOGY OF THE BLOOD. 81 than temporarily from the traditional 1 to 13, one cannot help getting the impression in some cases that the blood mass is increased or diminished, though we cannot prove it. Thus in certain cases of phthisis in which in spite of all the signs and symptoms of anaemia the blood count is normal, and simple con- centration of a really anaemic blood seems to be excluded by the absence of a cause for such concentration, we cannot help thinking that the whole amount of blood is too small. Again, it is possible that in individuals who eat and drink much and exercise little the blood-vessels may gradually accommodate themselves so as to hold a large bulk of liquid, and thus a true plethora or " full-blooded" condition might be brought about. Experiments show, however, that fat, "sedentary" animals (pigs) have less blood in proportion to their weight than lean, active animals (horses, dogs). Pigs' blood is only one twenty- second of their total weight, while horses' is one-tenth. I have repeatedly examined the blood in two patients with chronic cyanosis without known cause and found from 8,000,000 to 9,000,000 corpuscles in each case. One of these patients died of cerebral hemorrhage, and the autopsy revealed no lesions whatever except a stuffing of all the internal organs with blood and the results of chronic passive congestion everywhere. This certainly seems like a true plethora. Young, fast-growing animals have relatively more blood than adults, and males more than females. Our impression that old people are more or less " dried up" gets some support from the analogy of these animal experiments. Until some method is devised for estimating the total amount of blood during life, we shall never be sure upon this point. 6 CHAPTEK VII. ANJEMIA AND HYDR^EMIA. 1. ANEMIA. IN the absence of any proof that the total volume of blood can be more than temporality diminished, our definition of anaemia must be this: A deficiency in corpuscle substance) i.e., a deficiency in red corpuscles, in haemoglobin, or in both. l It is important to bear in mind that the color of the skin is not a safe guide in judging whether a person is anaemic. Thus out of 100 cases shown to be anaemic by actual blood examina- tion, Townsend 8 found a good color in the cheeks of 4 and a fair color in 7 others. Eighty-nine were pale. The color of the lips is but little better as a guide, as the following table from Townsend' s article shows : Table of Color in One Hundred Cases of Ancemia. Pale. Fair. Good. Nails 95 5 Cheeks 89 7 4 Tongue 84 15 1.2 Lips 76 21 2.4 Conjunctivas 64 25.5 10.5 My own impression would be that the lips and conjunctivae were better guides than they are shown to be in this table. In examining the color of the nails, the fingers should be flexed, as full extension may partly cut off the circulation under the nails. 1 This is a clinical definition and makes no attempt to go to the root of the matter. I have little doubt that chemical or other changes in the serum are the cause of the corpuscular changes, which only mirror the deeper disease. But these chemical changes are as yet so little understood that we have to judge of their presence chiefly by their effect on the cor- puscles. 2 Townsend : Boston Medical and Surgical Journal, May 28th, 1896. ANAEMIA AND HYDR^MIA. 83 A. K. Stone l and his assistants estimated the haemoglobin of 189 female patients who looked anaemic, and found over 75 per cent of haemoglobin in 89, or nearly one-half of them. For a woman a haemoglobin percentage of 75 per cent or more means practically normal blood. 8 The most striking example of the fallacy of judging of anaemia by the color of the skin and mucous membranes is in the so-called " tropical ancemia" Practically all persons belong- ing to white races who take up their residence in the tropics acquire after a time an extreme pallor of the skin and mucous membranes, and this appearance has usually received the title of "tropical anaemia." It turns out, however, from the careful studies of several different investigators, that the blood of such persons shows absolutely no anaemia or other variation from the normal. 3 The appearance of the skin is probably due to the action of the extreme climate on the peripheral nerves and vessels. Tropical anaemia is a condition not of the blood, but of the skin and subcutaneous tissues. Every one's experience includes a few persons who are per- fectly well despite an almost bloodless condition of the skin. On the other hand, anaemia may exist where there is a good color in the face. We are to judge of anaemia, then, solely by the blood exam- ination, and this judgment can be accurately made on the basis of the small fraction of a drop used for examination, provided always that our technique is good, and provided we make allow- ances for a considerable error wherever there is reason to sup- pose that any venous and capillary stasis is present, or that the blood is temporarily concentrated or diluted. Distinction between Primary and Secondary Ancemia, In one sense all anaemia is secondary. It is due to some cause, a symptom in a chain of events. But in some cases we know the cause and in some we do not. (a) Primary anaemia is that in which the causal factors are 1 Boston Medical and Surgical Journal, August 23d, 1894. 2 Where the haemoglobin is high the number of corpuscles is never con- siderably diminished. 3 So far as present methods of examination go. 84 CLINICAL BLOOD EXAMINATION. either entirely unknown or are apparently insufficient to cause so severe a disease. This division, like most of our statements about the blood, is a rough-and-ready one, held to provisionally until a better classification is discovered. It has a certain utility if not used with any less simple meaning than that given above. In view of our ignorance of the blood-making functions, there is little difference between saying that a primary anaemia is a disease of the blood-making organs and saying that it is one whose cause is unknown, especially as the pathological appear- ances in the bone marrow recorded in cases of so-called primary anaemia do not differ from those which can be brought about experimentally by bleeding. There is no good evidence that there are any primary diseases of the blood-making functions. A case of secondary anaemia is one in which we have an obvious cause such as hemorrhage or malaria for the loss of corpuscle substance. Eemove the cause and the anaemia ceases. Sometimes, however, after removal of the cause, e.g., after cure of a case of syphilis, the anaemia set agoing by the syph- ilis persists. On the other hand, there are few patients with "primary" anaemia who cannot recall some event in their past lives sufficient to account for a certain grade of ancemia (e.g., a nervous shock, a hemorrhage, an attack of tertian malaria). Yet if the anaemia that occurs after so slight a cause is of the pernicious or fatal type, we may fairly call it "primary." By this we mean that though the " cause" assigned might produce some anaemia, it was not sufficient to produce tliis fatal anaemia and has presumably little or no connection with it. "Primary" means not the absence of any cause of anaemia in the history, but the absence of any sufficient cause so far as is known. An attack of tertian malaria or a history of bleeding piles does- not cause fatal anaemia in 999 out of 1,000 people who have such a history. In the 1000th it is a case of post lioc and not propter hoc. Given the unknown cause that does lead to " primary" an- aemia, and it might be that a pregnancy, a nervous shock, or the presence of intestinal parasites would act as the straw that breaks the camel's back; but the important causal factor is the unknown factor. It is, then, by their etiology and not by their symptoms or by the blood examination alone that we distinguish primary from secondary anaemia. It is true that in the majority of cases we can tell from the blood examination alone whether a case is without known cause. ANAEMIA AND HYDRJEMIA. 85 ( = "primary") or symptomatic (= "secondary"). But there appear to be enough exceptions to this rule to make us cautious about stating it as a law. To be discussed under Secondary. VII. Secondary Ancemia. I. First Stage. I denned anaemia above as a diminution in corpuscle substance. In the milder types of this condition the number of red corpuscles is not diminished at all, but the indi- vidual cell is small, pale, and of light weight, through loss of nitrogenous matter. This is appreciated : (a) As a lack of coloring matter ; (6) As a lowering of the specific gravity. In the mildest grades of secondary anaemia there are no further changes. Such cases are those due to errors in hygiene bad air, poor food, lack of light or exercise to small hem- orrhages, and to the earlier stages of the diseases next to be mentioned. The lack of coloring matter is usually not present in every cell, as is seen in the stained specimens. Some are very pale at the centre, while others are well stained. II. Second Stage. Usually the next changes to appear are, like those already mentioned, qualitative, the number of red cells still remaining normal or approximately so. The individual cell as seen in fresh preparations is more or less deformed and varies from its normal diameter, dwarfed forms usually being commoner than the giant forms. These variations in size and shape are sometimes termed " poikilocy- tosis," and the dwarf and gianL forms are called respectively microcytes and macrocytes. Maragliano 1 has included the above changes, together with others about to be described, under the heading of Necrobiosis in the red corpuscles, attributing them to a patho- logical condition of the serum. The changes, united under this heading may be divided for convenience' sake into: (a) Endoglobular changes. 1 XI. Cong, f . Inn. Med. , Leipzig, 1892. 86 CLINICAL BLOOD EXAMINATION. (b) Poikylocytosis and crenation. (c) Changes in staining properties. (d) Changes involving motility in the corpuscle as a whole, or in parts of it. (e) Decrease in the average diameter of corpuscles with loss of the power to form rouleaux. All these changes may be watched in normal blood outside the vessels, as necrosis gradually comes on from contact with the air. Under pathological conditions the same changes may occur outside the body, but more quickly than usual (as other diseased tissues decompose more quickly after death than those of a sound man suddenly killed), or inside the body. (a) Endoglobular Changes (see Fig. 27, a). These consist in the appearance of clear hyaline spaces of various shapes within the corpuscle, round, triangu- lar, rod-shaped, etc. In the fresh specimen they change their shape rapidly and continually; in dried and stained specimens they appear as sharply outlined light spaces in the corpuscle. In normal blood these changes occur after thirty to seventy minutes outside the vessels. In some patho- logical conditions specimens show them the instant the blood is collected, and presumably they were present before it left the vessels. (b) Crenation and Poikilocytosi s (Fig. 27, b). What we ordinarily know as crenation in the corpuscles is the same sort of process which, occuring within the vessels, we call poikilocy- tosis. A lump rises at one or more points in the corpuscle, becomes more pointed, and gradually the whole cell acquires amoeboid motions, assuming in succession the various shapes with which we are familiar in poikylocytes. (c) The pointed projections may break off and move about actively in the plasma. These motions, as well as the preceding FIG. 27. Degenerative Changes in Bed Cells. AND HYDRJ3MIA. 87 amoeboid movement of the whole corpuscle, are to be explained as irregular contractions of the necrobiotic protoplasm, similar in a general way to the actions of a hen after its head is cut off. These motions are not to be confounded with the finer Brownian molecular" movement to be seen in any healthy cell. The or small bits broken off (Fig. 27, c) are doubtless the dwarf cells seen in dried and stained preparations. Curiously enough, these fragments tend again to assume the biconcavity character- istic of normal cells, as a drop of fat breaks into smaller but eimilar drops. FIG. 28. Elongated or Oval Corpuscles in a Case of Pernicious Anaemia. (d) Oval Shape. A great many cases of pernicious anaemia and some other diseases (see Epidemic Dropsy, page 240) show a marked tendency to oval shapes in corpuscles not otherwise considerably deformed. Even in normal blood I think there are a small number of oval forms, and in anaemia this number may be greatly increased till, as in Fig. 28, we get ,all the cells elongated. The same appearance can be produced by roughness in spreading the blood, but in such case the deformed corpuscles all point one way. 88 CLINICAL BLOOD EXAMINATION. (e) Changes in Staining Properties. Normal red corpuscles have affinity only for acid stains (eosin). The same degenera- tive changes that lead to the alterations in shape and size above described alter the staining properties of the cell as well, so that it takes up two or three colors (according to the number present in the stain) , either as a diffuse mixture or irregularly, some parts of the cell taking color differently from others. This has been termed a " polychromatophilic" or degenerative change. Some observers have supposed it to be rather of the nature of regeneration, believing that the cells take color in this unortho- dox way because they are half -developed, but the weight of evi- dence is that they are degenerative changes. (/) In many secondary anaemias, especially in those asso- ciated with inflammations, the average diameter of the cells is lessened, and the rouleaux are not formed. ' ( g ) Cells may lose their haemoglobin altogether, leaving only the shell of the corpuscle behind (see Fig. 27, d). Now all these necrobiotic changes are characteristic of the severer grades of secondary anaemia such as occur in cancerous cachexia, phthisis, nephritis, etc. The changes of staining affinity are less common than the others, and usually represent the severest grades of anaemia, but they have also been noted occasionally in smallpox, measles, scarlet fever, typhus, and purpura. In pernicious anaemia they are, as a rule, much more common than in any other disease. Maragliano considers these degener- ative changes to be due to toxic plasma. A lessened resistance to the ordinary plasma-environment on the part of the red cells would also explain them, and in such affections as paroxysmal haemoglobinuria it seems the most probable cause. In syphilis the abnormal sensitiveness of the red cells to the influence of mercury seems another instance where the red cells are imma- ture, decrepit, or weak. In syphilitic children, for instance, mercury easily gives anaemia, while in healthy children it does not. This will be discussed more fully under syphilis. The necrobiotic phenomena above described have been ob- served by Maragliano in carcinoma, lead poisoning, leukaemia, pernicious anaemia, purpura, cirrhotic liver, nephritis, pneu- 1 But in the severest forms of anaemia the diameters are apt to be in- creased (see below, Pernicious Anaemia). ANAEMIA AND HYDRJEMIA. 89 monia, malaria, typhoid, erysipelas, and tuberculosis. Celli and Guarnieri (Fortschritte der Medicin, 1889, No. 14) found them in measles and scarlet fever. Weintraub (Virchow's Ar- chiv, Vol. 131) noted them in epilepsy, pyaemia, and catarrhal jaundice. A decreased resistance to pressure of electric currents and other influences has also been noted by v. Limbeck in some cases. Such weakening of the red cells experimentally produced in animals by poisons has been found (Mya and Sanarelli, Arch, ital. di Biolog., XVII., 1892) to increase the susceptibility to infectious diseases. III. Third Stage. Here the number as well as the quality of the red cells begins to suffer. So far I have mentioned only the qualitative changes in secondary anaemia and have purposely made these changes more prominent than the actual diminution in the count of red cells, because it is only comparatively rarely and in very marked cases that the diminution in red corpuscles is considerable. The blood characteristic of most cases of sec- ondary anaemia is one in which the number of red cells is ap- proximately normal. The important exceptions to this rule are: 1. The anaemias of infancy and early childhood. 2. Large hemorrhages (soon after their occurrence). 3. Malaria. 4. Acute septicaemia. The direct and rapid destruction of the corpuscles by the malarial organism or hemorrhage account for this. Of sepsis and the anaemias of infancy we shall speak later. IV. Fourth Stage. The blood of secondary anaemia shows often evidence not only of degeneration and destruction of the cells but also of regenerative changes, namely : Nucleated Red Cells. These are usually divided into three groups (a) Normoblasts. (&) Megaloblasts. (c) Microblasts. Normoblasts. (a) The first are normally present in moderate number in the bone marrow of healthy persons, and in great numbers in the 90 CLINICAL BLOOD EXAMINATION. marrow after hemorrhage. They are generally considered to be a younger stage in the life of the corpuscle than the non- nucleated forms seen in the circulating blood. Hence the ap- pearance in the peripheral circulation of this form of nucleated cell is considered to mean that, in the comparatively plentiful reproduction of red cells called forth in the marrow by the anaemia, a certain number of red cells leave the nursery (the marrow) before they are grown up and circulate for a time in their immature state. A normoblast, then, represents an im- mature red corpuscle (see Plate IV.). In size and color it is like an ordinary red cell except that we find, usually somewhat to one side of it, a round nucleus about one-half the diameter of the whole cell. With Ehrlich's tricolor mixture, this nucleus stains very deep blue, nearly black, and is sharply outlined against the pale yellow of the cell body around it. The cell often looks as if it were pushing its nucleus out, i.e., in many instances we see the nucleus projecting over the edge of the corpuscle, or half out of it, and occasionally we find it lying beside the corpuscle from which it has just emerged; but this appearance is probably an artefact and not, as Ehiiich thought, the regular way of disposing of the normoblast nucleus. Very frequently the nucleus has toward the centre a light spot, sometimes so brilliant that it looks like the reflection of light from the surface of a drop of ink or any dark liquid, what artists call the "high light." Occasionally there are several of these light spots in a nucleus, or it may be all light blue-gray ex- cept a dark blue rim. This is the commonest type of normo- blast. But now and then we meet with one when the nucleus is more or less separated into two or more pieces. These pieces are usually connected by pale-staining "bridges," perhaps ra- diating from a centre so that the nucleus is "rosette-shaped," or it may take any one of a large number of different shapes. The parts of the nucleus which are nearest the periphery of the cell usually stain deeper than the " bridges" which join them. Sometimes the nucleus breaks apart completely and we find two or more separate unconnected nuclei within the single cell. 1 Or one of the pieces may be outside the cell and the others inside. 1 Apparent^ the nucleus is absorbed or degenerates (see Israel and Pap- penheim : Virchow's Archiv, vol. 143). ANEMIA AND HYDR^MIA. 91 Barest of all is the appearance of true mitosis in the nucleus of a normoblast. Megaloblasts. (b) The typical megaloblast as usually described is so unlike the normoblast that we should not naturally think of them as near relations. It does not occur anywhere in the healthy adult body, not even in the bone marrow. In the early foetal marrow and in the marrow and circulating blood of grave forms of anaemia it is to be found, usually in company with a certain number of nor- moblasts. Ehrlich described megaloblasts as the sign or product of a different type of blood formation, namely, the foetal type, and considered those anaemias in which it occurs as tend- ing to a return of the blood to the foetal state. [This has been recently confirmed by the researches of Pappenheim, though Ehrlich 's detailed theory of two methods of blood formation is now generally discredited. All nucleated red cells lose their nuclei by "absorption"; none by extrusion.] Ehrlich regarded the presence of megaloblasts as a bad prog- nostic sign, and believed that a pernicious or fatal anaemia was characterized by an excess of these cells over the normoblasts. He recognized that they might be found in various milder forms of anaemia; but here the prevailing type is the normo- blast, and regeneration may be more active than degenera- tion. In his general conception of the prognostic import of megaloblasts Ehrlich has been supported by the weight of later clinical observation, although it has been shown that the anaemia due to intestinal parasites can be cured, despite the presence of the megaloblasts as the prevailing type of nucleated red cells. So far as I know this is the only exception to Ehrlich 's rule. The typical megaloblast is an abnormally large cell (11 to 20 ,a in diameter, frequently showing marks of degeneration (poly- chromatophilia) in its protoplasm, which is therefore brownish or purplish with the Ehrlich-Biondi stain. Its nucleus is very large, filling most of the cell, and contrasts with the normoblast nucleus not only by its greater size but by the pale, even stain which it takes up. The commonest color of the nucleus with 92 CLINICAL BLOOD EXAMINATION. the Ehrlich-Biondi stain is pale green or robin 's-egg color. It is not stained evenly but dotted over with purplish granules arranged in a fine mesh like the knots in a fish-net (see Plate IV.). Outside the nucleus there is usually a narrow band of clear white, apparently an empty space, separating the nucleus from the encircling protoplasm. The protoplasm close round this colorless ring is usually stained more deeply than the rest of the cell. Cracks and " flaws" are sometimes to be seen in the proto- plasm, giving evidence, as its purplish stain does, of the necro- biotic changes described by Maragliano. The outline of the whole may be quite circular : oftener it is oval or somewhat irregular, but rarely much deformed. Microblasts. (c) Microblasts, which are rarer than either of the varieties just described, consist of a nucleus like that of a normoblast or smaller, and contained in a cell body smaller than the normal red corpuscle. In the writer's experience the cell body is usu- ally reduced to a few shreds of discolored protoplasm hanging about the nucleus (see Plate IV.). Their clinical significance is usually supposed to be that of megaloblasts. "Atypical Forms." As a rule we find in a given specimen of blood only typical normoblasts, microblasts, or megaloblasts, and accordingly can easily reckon up the number of each kind and see which type of blood formation predominates. Sometimes there are a few cells present, about the classification of which we cannot come to a decision, and I have occasionally seen a specimen of blood con- taining a large number of nucleated red cells no one of ivhicli could strictly be classed either as a "normoblast," a "megalo- blast," or a " microblast, " as these are defined above. The researches of Pappenheim have thrown much light on this difficulty. While insisting with Ehrlich that the megalo- blast and the normoblast represent respectively the early fostal and the post-uterine types of blood formation, and that there are no real "transitions" from the one to the other, he yet recognizes that the two varieties are not absolutely to be differ- entiated by any of the ordinarily accepted criteria such as size, Examination of the Blood. PLATE IV. Normal red cells m ' ifii Normoblasts Microblast Cells in karyo- - -> kinesis M f i mo ;. Varieties of Nucleated Red Cells. m. m. m. m. = Typical megaloblasts. D. D. D. D. = Cells with dividing nuclei, o. o. o. o. o. o. o. = Other (unnamed) varieties of nucleated red corpuscles. Scale of u .' Polychromatophilis cells CeUs deformed in size or shape. K. C. Gabot fee. Lith. Ant. v. K. A.. Funka, Leipzig. PLATE IV. (1 ) m,m,m, m Young megaloblasts. (2) D,D,D,D; the upper two are probably old normoblasts with de- generating nuclei, and the lower two old megaloblasts with nuclei is a similar condition. (3) 0,0,0,0, etc. The two cells in the lower right-hand corner are probably old megaloblasts whose nuclei are nearly absorbed. The three cells immediately to the left of these are probably young normoblasts the lowest one being the youngest. The other four cells marked "0,0,0,0" (those to the extreme left) are probably middle-aged megaloblasts. The two labelled " Normoblasts" are really old normoblasts. The appearance of extrusion of the nucleus on one of them is probably an artefact. The large cell on the extreme upper right-hand corner is probably a megalo- blast with a " pyknotic" or oedematous (degenerating) nucleus. (4) In the young or "typical" megaloblasts (m,m,m,m) note the white line around the nucleus, the variations in its tint, and, in two of them, the discolorations of the protoplasm (polychromatophilia), especially near the nucleus. The lower of the two cells in karyokinesis shows this best. (5) In the microblast note the ragged edge of the protoplasm. (6) In the lower portion of the plate ("cells deformed in size or shape") an actual field from a case of pernicious anaemia was copied. Macrocytes (or large cells) , microcytes (or small cells) , and m isshapen cells or poiki- locytes are shown. (7) The " polychromatophilic cells" in the lower right-hand corner were stained with the same mixture as those to the left of them, but have taken up other colors besides the orange G, which alone is taken up by normal red cells. ANAEMIA AND HYDRJEMIA. 93 color of nucleus, etc. Most " megaloblasts, " he admits, are larger than most normoblasts, but there are occasional giant nor- moblasts and dwarf megaloblasts which by size alone are indis- tinguishable. The large, pale, delicately netted nucleus of the "megaloblasts" is simply a young nucleus. All young nuclei are relatively large and pale, while the small dark nucleus of the normoblast is simply an old or degenerating nucleus. The real criteria of the two varieties, according to Pappenheim, is not the size or color of nucleus nor of the whole cell, but the structure of the nuclear network. This is a point difficult to make out by ordinary staining methods and not easily appre- ciated. Luckily for us, most "megaloblasts" are larger than most " normoblasts ;" and further, most of them as seen in the blood are young (i.e., have large pale nuclei with delicate chro- matin network), while most "normoblasts" are old, as shown by their small, dark, coarse-skeined nucleus, so that in the majority of cases Ehrlich's criteria for the two varieties are sufficiently correct for diagnostic purposes. Pappenheim of course wishes to abandon the terms " megaloblast" and "normoblast" alto- gether, but since size still remains the most easily recognized criterion of " megaloblasts" and " normoblasts" I shall continue to use the terms. On the chances, then, any nucleated red cell over 10 ;j. diameter should be classed as a megaloblast ivhatever the appearance of its nucleus, and any nucleated red cell under 10 v- diameter is probably a normoblast whatever the appearance of its nucleus. Microblasts simply represent degenerating forms (usually normoblasts) whose protoplasm is falling away. The clinical significance of the two varieties is just such as Ehrlich supposed. [These points will be made clearer by reference to Plate IV. and the remarks intended to explain it.] In most cases of severe secondary anaemia we find a few normoblasts. In very severe forms, whatever the cause, we may or may not find an occasional megaloblast. But these are much rarer than the normoblasts, even in the severest types of secondary anaemia. The only exceptions to this rule are the anaemias due to intestinal parasites, in which, though secondary and curable, the megaloblasts in some cases predominate over the normoblasts. Summing up the changes characteristic of secondary 94 CLINICAL BLOOD EXAMINATION. anaemia, which includes almost all the important pathological appearances occurring in red cells, we have : I. ( (a) Lack of haemoglobin. } specifics Characteristic of mild cases. TIT ' (b) Lowered ( gravity. ) II. The above and necrobiotic ) Characteristic of moderate changes of Maragliano. j cases. (a) Lack of red cells. ~1 (b) Presence of normo- [Characteristic of severe blasts and the above [ cases. (I. and II.). IV. Megaloblasts and the above ) Characteristic 1 * of very se- (I., II., and III.). ) vere cases. The changes in the white cells will be discussed in the next chapter. Among the commonest causes of secondary anaemia are : I. Infective and febrile diseases, acute or chronic. II. Malignant disease. III. Chronic suppurations, nephritis, chronic dysen- tery, cirrhosis of the liver. IV. Bad hygiene, pregnancy, and lactation. V. Intestinal parasites. VI. Poisons (lead, arsenic, etc.). To discuss the way in which each of these influences acts in producing anaemia is tempting, but falls outside the plan of this book. The following are good examples of the condition of the blood : SECONDARY ANJEMIA. o 1 1 Red cells. White cells. Per cent, haemo- globin. Remarks. 1 23 F. 1,656,000 2,048,000 1,808,000 1,568,000 4,248,000 2,300 2,600 3,200 1,300 2,300 18 24 30 58 60 72 Post-malarial. After 7 days' treatment. " 14 " 24 " " 34 " " " 43 " " 9, 45 M 2,208,000 50 Post-malarial 3 4 5 6 9 41 32 40 F. F. F. M. 2,186,000 3,240,000 2,920,000 2,040,000 8,500 5,400 12, 000 9,600 30 59 47 10 Post-hemorrhagic (8 hours) . Post-hemorrhagic (bleeding fibroid). After abortion with hemorrhage. Chronic dysentery. Differential count in case 6 showed polynuclear cells, 66. 3 ; myelo- cytes, 1.4; 8 normoblasts, 5 megaloblasts seen in counting 400 leuco- cytes. ANAEMIA AND HYDR^MIA. 95 2. HYDB^MIA. (a) Seen from the opposite point of view almost all cases of anaemia are hydraemic. That is, if the total volume of blood is to remain approximately constant (as it appears to do) , any loss of solids (corpuscle substance) must be made up by water taken in from the tissues. Hence any anaemic person's blood is thin, watery, or hydraemic. Women's blood is somewhat more hydraemic than men's, because less rich in cells. Ordinary chlorosis and secondary anaemia show no more water than normal in the serum, but the cells are probably somewhat water- logged. (6) In many conditions of dropsy, whether from heart or kidney, we may have more water than normal, both in the plasma and in the corpuscles themselves, which are capable of taking up considerably more than their normal amount of water. (c) Any temporary dilution of the blood under the conditions mentioned above (ingestion of liquid, lowered blood pressure, etc.) is from one point of view a hydj-aemic condition. No special clinical significance attaches to it other than that of anaemia, whose correlative it is. CHAPTEK YIII. LEUCOCYTOSIS LYMPHOCYTOSISEOSINOPHILIA MYELOCYTES. MUCH confusion Las been caused in the past by the failure to see in leuksemic blood anything more than an extreme and permanent form of leucocytosis, while leucocytosis was thought of as a mild and temporary leukaemia. We know now that they are totally different phenomena, differing not in the number, but in the kind of cells present in the increased numbers. Definition. There are many difficulties in denning leucocytosis. To my mind the term is best used to mean : An increase in the number of leucocytes in the peripheral blood over the number normal in the individual case, this increase never involving a diminution in the polymorphoniiclear varieties, but generally a marked absolute and relative gain over the number previously present. (a) I say "in the peripheral blood" because certain ob- servers hold that leucocytosis is not a real increase in the total number of leucocytes in the blood, but only an affair of distribution, the cells being drawn or attracted to the pe- riphery and out of the internal organs. Whether this theory be true or not, it is accurate to say that in the drop which we draw (whether also in the internal organs or not), the leuco- cytes are present in increased numbers per cubic millimetre. (b) In persons not usually to be considered sick, but simply somewhat wizened or ill-nourished, the normal count of white cells may be as low as three thousand per cubic millimetre. For such an individual ten thousand cells per cubic millimetre would be a decidedly pathological condition. On the other hand, there are persons, usually those of notable vigor and good nutrition, whose white cells rarely fall below ten thousand. Obviously we must take account of these differences both in our definition and in our practice if we are to reason correctly from the data of blood examination. LEUCOCYTOSIS. 97 (c) Further we must lay stress upon the varieties of leuco- cytes whose increase constitutes leucocytosis in distinction from either variety of leukaemia (splenic-myelogenous, or lymphatic). For instance, given a count of eighty thousand leucocytes per cubic millimetre, we cannot tell without knowing the va- rieties of cells present whether the case is a genuine leukaemia or a merely leucocytosis symptomatic of pneumonia, suppura- tion, malignant disease, or other conditions. (d) Thus defined leucocytosis is of two kinds. 1. That in which the relative proportions of the different varieties to each other is unchanged. 2. That in which the increase is made up solely or largely by a gain in the polymorphonuclear leucocytes. The latter includes nearly all pathological leucocytoses, the former being confined chiefly to the physiological leucocytoses next to be described. (e) Lastly, in order to be sure that the polymorphonuclear cells are not decreased, we must know what the normal percent- age/or that individual is. The normal percentage of these cells in infancy is from twenty-eight to forty per cent. In adults it is much higher, but varies like the total count, according to conditions of nutrition, etc. Thus the normal for adults is usually set at from sixty to seventy per cent, but no one indi- vidual's blood shows such variations in health, and if we include the obviously ill-nourished, but not actually sick, and also those in blooming health, we shall have to widen our nor- mal limits considerably. From fifty to seventy-five per cent are within normal limits according to the above conception. But obviously we can make no absolute judgment by a standard so vague. It is much better, I think, to consider each indi- vidual as his own standard within these limits, his count of polymorphonuclear cells being a fair measure of the soundness and vigor of his metabolism. Thus, in an obviously debilitated individual, we should consider seventy-two per cent of these cells very high, while in a vigorous athlete it might not be so. It is the endeavor to include all these limiting conditions that has made my definition so long and involved. It gives us, if it turns out to be true, some better way of classing individuals than as "sick" or "well" as regards their blood state. We find out how well or how sick their blood is (to a certain extent) , (a) 7 98 CLINICAL BLOOD EXAMINATION. by the total number of leucocytes present, and (b) by the pro- portion of polymorphonuclear neutrophiles in a given one thou- sand of these leucocytes. These data tell us approximately hoiv normal or hoia abnormal a given individual's blood is. When a given disease like pneumonia occurs, we need to know, if pos- sible, what is the ordinary leucocyte count and differential count of that case, on top of which a leucocy tosis may (or may not) be built up. Condition of stasis, temporary blood concentration, dilution, and vasomotor disturbances must, of course, be excluded or allowed for, since these may increase not only the total leu- cocyte count, but often the percentage of polymorphonuclear cells. Whether or not differences of race make any difference in the normal count of white cells, I cannot say, but certainly the average of a group of college athletes would be higher than that of some country towns in New England where everybody is more or less under-nourished ; and if one is to practise among all sorts and conditions of men, I think he cannot but expect to find people's leucocytes vary all the way from 3,000 to 10,500 per cubic millimetre, without there being more than malnutrition to account for the lower figures. Into the theories of how leucocytosis is brought about I shall not enter ; no one of them as yet commands general assent. "We may divide leucocytoses for convenience' sake into : 1. Physiological leucocytoses. 2. Pathological leucocytoses. PHYSIOLOGICAL LEUCOCYTOSES. 1. Leucocytosis of the new-born. 2. Leucocytosis of digestion. 3. Leucocytosis of pregnancy. 4. Leucocytosis post-partum. 5. Leucocytosis after violent exercise, massage, and cold baths. 6. Leucocytosis of the moribund state. The Leucocytosis as Affected by Digestion. (a) Total abstinence from food lowers the leucocyte count. In the blood of the professional faster Succi, the number sank within his first week's fast to 861 per cubic millimetre. After LEUCOCYTOSIS. 99 the first week it rose to 1,530, and remained there throughout his thirty days' abstinence (Luciani '). The polymorphonu- clear cells and eosinophiles are said by Tauszk to be increased in chronic starvation. Yon Limbeck counted the blood of a melancholic patient who had fasted a week, and found 2,800 white cells per cubic millimetre. These facts support the idea that the number of leucocytes depends (within certain limits) on the individual's assimilation of food. In cancer of the gullet we find similar low figures. (b) After a meal rich in proteids the leucocyte count rises about thirty-three per cent in most sound persons. Ten thou- sand cells may perhaps be considered the average, three to four hours after a proteid meal, but if the count before a meal is only 4,000 or 5,000, digestion will perhaps not raise it above 7,000, while vigorous adults may show 13,000. Digestion leu- cocytosis is always relative to the count of the individual's blood when fasting. This is to be obtained preferably before breakfast, as during the day the leucocy tosis caused by one meal may not be gone before the influence of the next meal begins. Occasionally we see sound persons with little or no digestive leucocytosis. Some of these cases are to be explained by habit- ual constipation (v. Limbeck) ; in others the reason is more obscure. But there is no doubt of its being the rule after meals of mixed or proteid diet. In herbivorous animals, and presumably in vegetarians, it is not found. Any disease of the gastro-intestinal tract, whether functional or organic, may prevent the appearance of the digestion leuco- cytosis (see later under Diseases of the stomach, page 280). In anaemic and debilitated conditions it is frequently absent. In children it is especially marked. Schiff 2 records a case of a healthy infant whose blood an hour after birth showed 19,500 (see next section), after its first meal 27,625, and after its fourth meal 36,000 white cells per cubic millimetre. After the second day this gradually diminished. Food seems to call forth a greater leucocytosis in proportion as it is a novelty in the stomach. Cases of gastric ulcer who had been fed exclusively by rectum for some weeks show a 1 "Des Hungern," German translation by O. Frankel. Hamburg, 1890. 2 Zeit. f. Heilk., xi., 1890. 100 CLINICAL BLOOD EXAMINATION. greater leucocytosis after their first meal than later. Perhaps the size of the digestion leucocytosis in the new-born is to be similarly explained. In diabetics the digestion leucocytosis is sometimes very large. The leucocytosis can usually be observed one hour after a meal, increases for two, three, or even five hours according to the slowness of digestion, then falls again. Burian and Schur ' found an increase of the polymorpho- nuclear varieties in those cases in which an increase of the total count took place at all. Eosinophiles show no regular changes. Diagnostic Value. 1. When we wish to know whether a person is accurate in such statements as that they have "eaten nothing for a week," we can get evidence from the leucocyte count, which should be very low if the assertion be true. Whenever we cannot communicate with a patient and wish to know how much food he has taken of late, we can form some idea from the blood examination. In the case of a patient who spoke only Russian, I was led to look for a stenosis of the gullet by the lowness of the leucocyte count (2,700), and the probang confirmed the suspicion. 2. As suggested above, we can form some idea of a person's general vigor, nutrition, and capacity to assimilate food by the number of leucocytes and the proportion of mononuclear cells, as compared with the average figures for that age and locality. Persons debilitated from any reason are apt to show it in their blood by the changes above mentioned, the element of hysteria being sometimes recognizable by other signs (see below: " Eosinophilia, " page 116). 3. Slowness of digestion is indicated by a late appearance of the digestion leucocytosis. The inferences to be drawn from the blood in diseases of the gastro-intestinal tract will be discussed later (page 274). 4. Perhaps the chief importance of digestion leucocytosis is as a possible cause of false inferences, through being taken for a pathological increase. Bearing this in mind, we must always examine the blood as near a meal as possible, or better still before breakfast. 'Wien. klin. Woch., February llth, 1897. LEUCOCYTOSIS. 101 Leucocytosis of the New-Born. The following table is compiled from the best authorities on the subject (Schiff, Gundobin, Bayer, Hayem, and others) : Age. Bed cells. Leucocytes. At birth 5 900 000 17 000 to 21,000(26,- End of first day 7 000 000 to 8 800 000 000 to 36, 000 after first feeding) . 24 000 " second day Generally increased. 30 000 " fourth day . 6, 000, 000 20, 000 " seventh day 5,000,000 15,000 Tenth day 10 000 to 14 000 Twelfth to eighteenth day . . . 12,000 Sixth month 12,000 Sixth year and upward 7,500 The increase is explained by Lepine, v. Limbeck, and others as a combination of blood concentration with large digestion leucocytosis. Gundobin and others are opposed to this theory. Certainly the influence of digestion on infant's blood is much greater than in adults. After a meal 30,000 leucocytes is never a very high count in infants under two years. A fuller discussion of the subject will be found in the chapter on the blood in infancy. The Leucocytosis of Pregnancy. Most primiparce show during the latter montJis of pregnancy a moderate increase of all varieties of leucocytes. Thirteen thou- sand cells per cubic centimetre is about the average count. In multipart it occurs in only about fifty per cent of the cases. Digestion leucocytosis "on top of" the constant preg- nancy leucocytosis, so to speak, does not occur. As mentioned above, the relative percentage of the different types of leucocyte remains unchanged, so that all varieties must be equally increased (eosinophiles excepted). The fact that digestion does not increase the pregnancy leucocytosis, leads to the suggestion that the whole thing may be only a prolonged digestion leucocytosis the mother having to eat for two. The swelling of thejbreasts may also account for part of the leuco- 102 CLINICAL BLOOD EXAMINATION. cytosis. In the last weeks of pregnancy the leucocytosis increases till at the beginning of labor it is often 16,000 to 18,000. It has no diagnostic value, as it is not present during the earlier months of pregnancy when diagnosis is difficult, and in the later months such conditions as hydatiform mole and fibroid tumors might raise the count of white cells as much as pregnancy. Leucocytosis After Parturition. The following charts illustrate the course of the leucocyte curve from the time of parturition till the end of the second week after it. All were primiparae excepting Nos. 5, 8, and 9. There was no sepsis in any case, and the temperature charts were practi- cally normal after the second day. No reasons are known for the variations between the different cases. All were counted at the same hour of the day, and under the same conditions of nutrition. All nursed their children. The only importance of this leucocytosis is that it might be confounded with a pathological leucocytosis in a case suspected of being septic. Just how long the leucocytosis is prolonged during lactation has not been studied so far as I am aware, but it certainly may go on several weeks. Violent exercise, massage, and short cold baths have been shown to cause a temporary increase in the number of leuco- cytes in the peripheral blood, all varieties of the cell being equally increased. The explanation usually given is that the blood is concentrated by vasomotor contraction and rise of blood pressure. Schultz (Deut. Arch. f. Idin. Med., 1893, page 234) found the leucocytosis of exercise amount to about the same as that of digestion, 11,000 to 13,000. He also noted that in dogs merely opening the peritoneum aseptically or breaking a leg caused leucocytosis. Thayer studied twenty cases of typhoid and found an aver- age of 7,724 white cells before and 13,170 after a Brand bath (Johns Hopkins Medical Bidletin, April, 1893). The increase took place equally in all varieties. Winternitz (Imperio-Koyal Medical Society, Vienna, February, 1893) came to a similar conclusion and found also that prolonged cold bathing decreased LEUCOCYTOSI8. 103 the number of white cells (dry cold does the same) . A patient was recently brought to the Massachusetts Hospital who had fallen through a hole in the ice and been some minutes in the s a II ' ^ s- a 1-1 g w g I! 5 ^ c8 O .2 2 icy water. His temperature was 91.8 by the rectum. Blood count showed 17,500 leucocytes per cubic millimetre. Next day he was perfectly well. On the contrary, short hot baths decrease and prolonged ones increase the number of leucocytes. 104 CLINICAL BLOOD EXAMINATION. Local arm baths have a similar effect, raising the count of leucocytes in the blood of the immersed arm if cold and short, and lowering it if hot and short, while prolonged immersion has an opposite effect. In the other arm the counts go up when M aj ffl 7T \ - J* 11 s B ^i K 00 O > V A V 05 ,3 0> il 3 I 1 a 1 8 8 those of the immersed arm go down, and vice versa (Eovighi). 1 JMitchell 2 found that the leucocytes showed distinct increase (as ! Arch. Ital. d. Clin. Med., xxxii., 3, 1893. 2 American Journal of the Medical Sciences, May, 1894. LEUCOCYTOSIS. 105 well as the red cells and haemoglobin) after one hour's gen- eral massage. All these forms of leucocytosis are usually explained b^ changes in blood pressure, and vasomotor changes affecting the calibre of the peripheral vessels and consequently their con- tents. Terminal Leucocytosis. The leucocytosis of the moribund state, though by no means invariable, occurs in many cases, whether from the influence of a terminal infection or from stasis. Where death is sudden or rapid it does not occur. It seems to be analogous to the ter- minal rise of temperature seen at the close of many chronic non-febrile affections. The longer the patient is moribund the higher the count reaches. In pernicious anaemia the increase may be so great as to simulate lymphatic leukaemia. Such a case occurred in the writer's own experience. The patient had presented the signs and symptoms of pernicious anaemia, and the blood was typical of the disease in all respects except for the lack of nucleated red cells. Slides taken on the day of death showed a ratio of one white to fifteen red cells, the small lymphocytes greatly predominat- ing, but the autopsy revealed simply the lesions of pernicious anaemia. The differential count of one thousand leucocytes on the day of death showed: Lymphocytes, 91.7 per cent; poly- morphonuclear cells, 7.7 per cent; eosinophiles, 0.5 per cent. Four megaloblasts were seen while counting these. The total leucocyte count was unfortunately not made. In ordinary cases the differential count shows an increase in the polymorphonuclear leucocytes. Thus in a case reported by Eieder, in which the leucocyte count rose during the last two days of life from 7,800 to 59,300, the polymorphonuclear cells constituted 87.5 per cent of the whole 59,300. PATHOLOGICAL LEUCOCYTOSES. For convenience' sake these may be divided as follows : 1. Post-hem orrhagic leucocytosis. 2. Inflammatory leucocytosis. 3. Toxic leucocytosis. 106 CLINICAL BLOOD EXAMINATION. 4. Leucocytosis in malignant disease. 5. Leucocyjosis due to therapeutic and experimental in- fluences. 1. Post-hemorrhagic Leucocytosis. Within an hour after a large hemorrhage we find commonly a considerable increase (16,000-18,000). In hemorrhage from the stomach this disappears again usually within a day or two, while in ordinary traumatic hemorrhage it persists longer. This last fact may perhaps be explained, as v. Limbeck sug- gests, by the local conditions in the wound rather than by the loss of blood in itself. The polymorphonuclear leucocytes are usually increased relatively and absolutely as in other forms of pathological leu- cocytosis. Sometimes we have lymphocytosis (see page 114). The degree of increase in the white cells is parallel in a general way to the anaemia produced in the individual, i.e., it depends on his powers of recuperation rather than on the amount of blood lost. Its duration follows the same rule/ 2. Inflammatory Leucocytosis. I use the term "inflammatory leucocytosis" rather than " leucocytosis of infectious diseases" because there is a consider- able number of infectious diseases in which no leucocytosis oc- curs, while it accompanies almost all forms and cases of inflam- mation. Nevertheless I shall class under this heading some diseases in which inflammation plays but a very subordinate role. I. Although purulent and gangrenous processes usually cause a higher count of white cells than serous processes, the amount of the exudation is not a measure of the amount of leucocytosis. It seems to depend rather on the resultant of two forces, viz., the severity of the infection and the resisting power of the indi- vidual.. These factors may interact in various ways : 1. Infection mild : resistance good = small leucocytosis. 2. " less mild : " less good = moderate leucocytoeis. 3. " severe: " good = very marked leucocytosis. 4. " " " poor == no leucocytosis. 1 Further account of the blood after hemorrhage will be found on page 126 et seq. PATHOLOGICAL LEUCOCYTOSIS. 107 This will be illustrated later under "Pneumonia " and under "Sepsis." Experiments on animals show that whereas moder- ate sized doses of septic cultures, not sufficient to kill the animal, are followed by leucocytosis, larger doses after which death follows speedily, do not raise the leucocyte count at all. Animals weakened by any cause show less leucocytosis to a moderate dose than strong animals. If the individual reacts from the shock his leucocytes are in- creased again and rise above normal. If reaction fails the leu- cocytes do not rise. II. Inflammatory leucocytoses differ from physiological leu- cocytoses (a) In being usually of larger extent. (6) In being almost aways accompanied by a relative and ab- solute increase in the percentage of polymorphonuclear cells. III. The course of the leucocytosis as regards both amount and duration shows, like the temperature chart, certain more or less characteristic differences in different diseases. IV. In some cases in which the absolute number of leuco- cytes is not increased, we see a relative increase in the polymor- phonuclear cells, pointing to the fact that influences are at work similar to those which produce an absolute increase. Y. That the amount of exudation is not of itself a measure of the amount of leucocytosis is shown by the fact that erysipelas or scarlet fever may be accompanied by as high a count as the average count in pneumonia or empyema. That purulent exudations usually have more effect on the white cells than do serous ones is due, I suppose, to the fact that a purulent inflammation usually means a severer infection. YI. No direct connection exists between leucocytosis and fever, many febrile affections running their course with a normal leucocyte count. When both leucocytosis and fever are due to the same causes they rise and fall together, but the correspond- ence is rarely accurate, and marked leucocytosis may exist with- out fever. VII. Acute, rapidly spreading inflammations seem to pro- duce a greater leucocytosis (other things being equal) than those in which the process is relatively chronic and stationary. For instance, an appendicitis, when well walled off and station- ary, shows less increase in white blood cells than while its le- 108 CLINICAL BLOOD EXAMINATION. sions are progressing. But peracute, overwhelming general sep- sis may have no effect on the leucocytes, the reactive power of the organism being crushed. YIII. Most inflammatory leucocytoses are preceded by a temporary diminution in the number of leucocytes. This oc- curs in animals from shock of any kind (blows on the head, tying to the etherizing board) , and it seems not unlikely that the cause is the same in all cases. The following is a list of the more important inflammatory or infectious conditions in which leucocytosis appears: 1. Infectious diseases ivith comparatively slight local inflamma- tory processes : (a) Asiatic cholera. (b) Relapsing fever. (c) Typhus fever (according to the majority of observers). (d) Scarlet fever. (e) Diphtheria and follicular tonsillitis. (/) Syphilis (secondary stage). (a) Erysipelas. (h) The bubonic plague. (i) Yellow fever (some cases) . 2. Infectious diseases with more extensive, local lesions : (a) Pneumonia. (6) Small-pox (suppurative stage). (c) Malignant endocarditis, puerperal septicaemia, and all pyaemic and septicsemic conditions. (d) Actinomycosis. (e) Trichinosis. (/) Glanders. (g) Acute multiple neuritis (febrile stages). (h) Acute articular rheumatism. (i) Septic meningitis and cerebro-spinal meningitis. (J) Cholangitis, cholecystitis, and ernpyema of the gall blad- der. (k) Acute pancreatitis. (I) Endometritis, cystitis (some acute cases). (m) Gonorrho3a. 3. Local inflammatory processes : (a) Abscesses of all kinds and situations, such as Felon. PATHOLOGICAL LEUCOCYTOSIS. 109' Carbuncle, furunculosis. Tonsillar and retropharyngeal abscess. Appendicitis, phlebitis (some cases). Pyonephrosis, perinephritic abscess and pyelonephritis. Osteomyelitis, empyema. Psoas and hip abscess when not simply tubercular. Abscess of lung, liver, spleen, ovary, prostate. Salpingitis and pelvic peritonitis, epididymitis. (b) Inflammations of the serous membranes including : Pericarditis, peritonitis, arthritis (serous or purulent, non- tubercular), conjunctivitis. (c) Gangrenous inflammations, as of the Appendix, lung, bowel, mouth (noma). (d) Many inflammatory skin diseases, such as dermatitis, pemphigus, pellagra, herpes zoster, prurigo, some cases of uni- versal eczema, etc. 3. Toxic Leucocytosis. Under this heading I have grouped most of the conditions not obviously to be explained as infectious or inflammatory (though some may turn out to be such) and not due to malig- nant disease or therapeutic agencies. This classification is chiefly for convenience' sake and represents only a guess at the real explanation of the leucocy tosis : (a) Leucocytosis of illuminating-gas poisoning. (b) " quinine poisoning. (c) " rickets (many cases). (d) " the uric-acid diathesis, gout. (e) " acute yellow atrophy of the liver. . (/) " advanced cirrhosis of the liver (some cases) especially with jaundice. (g) " acute gastro-intestinal disorders (pto- mains ?) . (li) " chronic nephritis, usually in ursemic cases. (i) after injections of tuberculin and thyroid extract. (/) after injection of normal salt solution (in- travenous) . (k) after ingestion of salicylates. (?) during and after etherization. HO CLINICAL BLOOD EXAMINATION. Possibly the leucocytosis of acute delirium belongs also in this group. 4. Leucocytosis of Malignant Disease. Very likely this belongs more properly under one or another of the classes just mentioned. Some observers think that it oc- curs only from the inflammation excited in the periphery of some malignant tumors ; others that it is due to absorption of morbid products from the tumor itself ; others again that it is to be accounted for by the cachectic state associated with the growth of the tumors. The details and conditions of its occur- rence will be discussed later (page 332). 5. Leucocytosis Due to Therapeutic and Experimental Influences. Pohl ' found that most of the so-called tonics and stomachics produce a slight increase in the white cells in animals, particu- larly the vegetable tonics like tincture of gentian, and oil of anise seed, while bismuth, bicarbonate of soda, and iron had no such effect. Quinine, caffeine, and ethyl alcohol gave likewise negative results. Yon Limbeck found leucocytosis in men after oil of peppermint and oil of anise seed. Binz 2 got the same results with camphor. In all these ex- periments the substances were given by the mouth. Using subcutaneous or intravenous injections, Lowit experi- mented on animals with hemialbumose, peptone, pepsin, nucle- inic acid, nuclein, extract of blood-leech, pyocyanin, tuberculin, curare, uric acid, urate of sodium, and urea. All but the last of these produce temporary decrease followed by increase of leucocytes. Goldschneider and Jacob 3 used extracts of various organs. Extract of spleen, marrow, and thymus produced leucocytosis preceded, as in Lowit' s experiments, by a brief diminution in the number of leucocytes, while extract of pancreas, thyroid, kidney, and liver had no effect. Winternitz 4 injected a large variety of substances subcutane- 'Arch. f. exp. Path. u. Pharm., 1889, vol. xv. 8 Arch. f. exp. Path. u. Pharm., vol. v., p. 122. 3 Arch. f. Anat. u. Physiol., 1893, p. 567. 4 Arch. f. exp. Path. u. Pharm., vol. xxxv., p. 77. PATHOLOGICAL LEUCOCYTOSIS. Ill ously and found that the degree of leucocy tosis was parallel to the degree of local reaction excited. For example, neutral salts and weak acids or alkalies pro- duced slight local inflammation and a leucocy tosis of from forty to seventy -five per cent of the original count. But irritants like turpentine, croton oil, nitrate of silver, sulphate of copper, mercury, antimony, digitoxin, etc., produced local suppuration (aseptic) and much greater leucocytosis (two hundred to three hundred per cent). Pilocarpine and antipyrin have been found by v. Jaksch and others to produce marked increase in the number of leucocytes when given subcutaneously . During the use of thyroid ex- tract Kichter (Centralblatt f. inn. Med., 1896, p. 3) noted leucocytosis. A large number of observations on the effects of injections of bacteria or their toxins agree in the following results. 1. Where the dose is very large the leucocytes are reduced, and the animal dies. 2. Where the dose is not sufficient to kill the animal the temporary diminution in the leucocytes is soon followed by leu- cocytosis. 3. Where the dose is slowly fatal the count of leucocytes oscillates up and down within wide limits. 4. Animals previously rendered immune to the poison in- jected show little or no leucocytosis. 5. Leucocytosis is more easily called forth and of greater ex- tent in young animals. 6. Most pathogenic organisms act similarly, but bacilli and toxins of tuberculosis as a rule cause no leucocytosis. 7. There is no evidence that any one variety of leucocyte is attracted by any particular bacillus or toxin. In the above sketch of therapeutic and experimental forms of leucocytosis no attempt has been made to give anything but the more interesting and important outlines of the immense amount of work done. Cell Structure of the Leucocytes in Leucocytosis. Hitherto we have spoken as if leucocytosis meant only an increased number of the normal cells, but one cannot study the cell forms in extensive pathological leucocytosis without noting 112 CLINICAL BLOOD EXAMINATION. in many cases qualitative changes in the individual cells. These are chiefly : 1. A greater or less approximation of the nuclei of polymor- phonuclear neutrophiles to the appearances of the myelocyte nucleus. As will be mentioned later under leukaemia, we find in every blood containing many myelocytes numerous cells whose nucleus is on the border-line between the myelocyte and the polymorphonuclear stage, so far as appearances go. Now in leucocytosis we find the same "border-line" cells in smaller FIG. 29. Atypical Leucocytes seen in Leucocytosis. 1, Leucocytes with polar arrange- ment of nuclei (mitosis?); 2 and 3, leucocytes with nuclei resembling those of myelo- cytes; 4, leucocyte containing two kinds of granules. numbers, the likeness to the myelocyte sometimes passing into identity in one to three per cent of the cells. 2. A greater or less approximation of the appearance of the large lymphocytes' protoplasm to that of myelocyte protoplasm, i.e., a diffuse violet or purple color exactly as in the myelocyte but non-granular. Engel makes a separate variety of this cell, giving it the useless name of " mononuclear cell. " 3. Other finer changes, such as the number, size, and stain- ing power of the neutrophilic granulations, polar position of the nuclei, etc. (see Fig. 29), require further study. PATHOLOGICAL LEUCOCYTOSIS. 113 Changes like the above militate against the idea that leu- cocytosis is simply a matter of distribution in the peripheral or internal vessels. Absence of Leucocytosis. It is of fully as great a practical assistance to us to know that in certain infective diseases leucocytosis is regularly absent as to know those conditions in which it is to be expected. Among the most important diseases in which leucocytosis is conspicuously absent are : (a) Typhoid fever. (6) Malaria. (c) Grippe (most cases). (d) Measles. (e ) Rotheln and mumps. (/) Cystitis. ( g) Tuberculosis, including Incipient phthisis. Miliary tuberculosis. Tubercular peritonitis. ostitis and periostitis, pleurisy, pericarditis. 1 In some of these affections, notably in miliary tubercle and the later weeks of typhoid, the leucocytes are diminished. Further details will be given under the special diseases. LEUCOPENIA. Definition. A diminution in the number of white cells in the peripheral circulation as compared with the number normal for the given individual. 1. The effects of starvation and malnutrition in producing leucopenia have already been described. Such leucopenia is usually associated with lymphocytosis (see below). Cancer of the gullet is an example of this class. 2. Short hot baths or prolonged cold baths produce tem- porarily the same result (Winternitz, loc. cit.). 1 Tubercular meningitis often does show leucocytosis (vide infra, page 266). 114 CLINICAL BLOOD EXAMINATION. 3. Most of the infective diseases in which there is no leuco- cytosis are sometimes characterized by leucopenia, e.g., grippe, measles, miliary tuberculosis, and other forms of pure tuber- cular infection, malaria, and especially typhoid, in the later weeks of which it is almost invariable, and is accompanied by lymphocytosis. Where a case of leukaemia is complicated by an infective disease (pneumonia, septicaemia) the number of leucocytes may fall below the normal. In a case recently occurring at the Mass- achusetts General Hospital in which a lymphatic leukaemia was terminated by septicaemia from glandular suppuration, the white cells fell gradually from 40,000 three weeks before death to 419 per cubic millimetre on the day of death. I have never heard of a lower count than this. The differential count was unchanged (lymphocytes ninety-eight per cent). 4. In pernicious anaemia the count is usually very low and may fall below 1,000 cells per cubic millimetre. Other forms of anaemia (rachitic, syphilitic) occasionally produce the same re- sult. LYMPHOCYTOSIS. Lymphocytosis is a relative increase in the lymphocytes in the blood, with or tvithout an increase of the total leucocytes count The increase is relative to the percentage of lymphocytes normal for the individual. When lymphocytosis and an increase of the total leucocyte count are present we cannot distinguish the blood from that of lymphatic leukaemia, and the distinction must depend upon the course and symptoms of the case. ' 1. Such a condition (relative to the adult) occurs in healthy infant's blood and in many diseases of infancy, the blood seem- ing to have a tendency to return to the infantile type. This is especially true of cholera infantum and any gastro-intestinal trouble. Anything that retards the infant's normal gain in weight or general development retards its blood development as well. Thus a child of three, convalescent from a summer diar- rhoea, may have fifty to sixty per cent of lymphocytes, which 1 The lymphocytosis of chlorosis has been mistaken for lymphatic leukaemia (Schreiber) owing to too exclusive reliance on the results of the blood examination. The patient recovered. Such cases are very rare, and the difficulty hardly ever arises. LYMPHOCYTOSIS. 115 would be normal for an infant of a few weeks, but for three years old is very high. 2. Hereditary syphilis is perhaps the best-known cause of relative lymphocytosis in children. Scurvy may produce the same result. Dividing the anaemias of children into two groups, those that do and those that do not produce leucocytosis, it appears that the great majority of those whose total leucocyte count is normal show a relative lymphocytosis. This is the case irrespective of whether there is enlargement of the spleen or not. Sometimes the smaller, sometimes the larger lymphocytes are in the majority. Often no division between the two kinds is possible. 3. In adults some forms of debility may be associated with relative lymphocytosis as above noted (page 97). It is most marked, however, in chlorosis, pernicious anaemia, and the anae- mia secondary to syphilis, in the later weeks of typhoid fever and in lactation. 4. Certain cases of Graves' disease show marked lymphocy- tosis. How such cases differ from those that do not show it I have not been able to determine. 5. It occurs also in haemophilia, goitre, in some cases of cervical adenitis, whether tubercular or lymphomatous, and in tumors of the spleen. 6. During the administration of thyroid extract a lymphocy- tosis has been recently noted by Perry (New York Medical Rec- Becord, August 29th, 1896). 7. The larger forms of lymphocytes are increased in some splenic tumors (chronic " ague cake"), at the end of scarlet fever, iu imeumonia with delayed resolution (some cases), in measles, certain forms of phthisis and in the non-suppurative stages of small-pox ; also in many of the same diseases in which the small lymphocytes are increased. 8. So far I have referred chiefly to relative lymphocytosis. Absolute lymphocytosis is very rare outside of lymphatic leu- kaemia. One case occurred at the Massachusetts General Hos- pital in 1894 a child of six, who passed through an attack of bronchopneuinonia with uneventful recovery, the only pecu- liarity of the case being the marked increase of white cells run- ning up to 94,600, sixty-nine per cent of which were lymphocytes. During convalescence the blood became normal and the child 116 CLINICAL BLOOD EXAMINATION. left the hospital well in all respects. The case will be referred to later in the account of the blood of pneumonia. Diagnostic Value of Lymphocytosis. 1. I have already suggested that the degree of health in persons not organically diseased might perhaps prove to vary directly with the percentage of polymorphonuclear cells in the blood. 2. In children the same percentage is to a certain extent a measure of the child's degree of development causes of leu- cocytosis being excluded, and the percentage normal for a child of the patient's age being taken as the standard. 3. The diagnosis of obscure syphilitic disease may be supported by the coincidence of lymphocytosis with eosino- philia. 4. Absolute lymphocytosis in the presence of glandular tumors is our mainstay in the diagnosis of lymphatic leukaemia. EOSINOPHILIA. Definition. An increase in the percentage of eosinophiles in the circulatory blood, with or without an increase in the total leucocyte count. The researches of Neusser, Zappert, Weiss, Klein, and others have brought the eosinophilic cells once more into the promi- nence which they lost when it became apparent that they were in no way peculiar to leukaemia. 1. Leukaemia is occasionally associated with eosinophilia (see below, page 167), but in the majority of cases this is not so. As in normal blood, from one to three per cent of them are to be found. 2. In infancy the percentage of eosinophiles is very often higher than in adults, so that in them eosinophilia may be con- sidered physiological. In adults its presence is often unex- plained. The eosinophiles are the most seemingly capricious of all blood cells. A certain amount of light has been thrown on them by the observations of Neusser and his pupils (Weiss, Schreiber, Klein, ai)d others). 3. Neusser noticed that eosinophilia occurs EOSINOPHILIA. 117 (A) In many affections of the bones (sarcoma, leukaemia, osteomalacia). (B) In many affections of the skin (pemphigus, pellagra, and others). (C) In troubles involving the female genitals, especially the ovaries. (D) In disturbances of the sympathetic nervous system. That there is some relation between these seemingly uncon- nected sets of phenomena is shown by various other facts besides the presence of eosinophilia in them all. (a) Bone and genitals. Osteomalacia is most apt to occur in pregnancy and is cured in some cases by castration. (6) Genitals and sympathetic nervous system. The presence of all sorts of psychoses and vasomotor troubles associated with menstruation, pregnancy, and the climacteric, and the so-called " reflex" disturbances in connection with uter- ine or ovarian disease, are well known. (c) The connection of the skin with both of the last-men- tioned systems is seen in the trophic disorders and sympathetic dermatoses of hysteria and ovarian disease. Working out the suggestions of this theory Neusser and his pupils have found relative eosinophilia in the following affec- tions : 1. Bones. Osteomalacia, malignant bone-tumors, pernicious anaemia (some cases), splenic-nay elogenous leukaemia (occasionally). [The writer has seen slight eosinophilia in osteomyelitis.] Possibly the relative eosinophilia of normal infants' blood may be connected with the great activity of their bone growth. 2. Diseases affecting the skin. Urticaria, pellagra, dermatitis herpetiformis, and pemphi- gus (constantly) ; some varieties of herpes, prurigo, eczema, lymph odermia perniciosa ; after vaccination, in scarlet fever and syphilis (not measles or small-pox), ichthyosis, lupus, myx- oedema. 1 1 1 do not vouch for these or for any of Neusser's statements, which are frequently incorrect. In a single case each of pemphigus and prurigo I have found only four per cent and three per cent of eosinophiles. On the 118 CLINICAL BLOOD EXAMINATION. 3. Genitals. Gonorrhoea, prostatitis, many ovarian tumors, before and during the early days of menstruation, puerperal mania, and the psychoses 'of menstruation, of the puerperium, and of the climacteric; in sexual neurasthenia, after coitus, and in lacta- tion. 4. Sympathetic Nervous System. The psychoses last men- tioned, hysteria, Basedow's disease, and some of those given under the next heading. 5. Besides these general groups, Neusser has noticed another class of cases characterized by eosinophilia, namely, those in which some member of the group of xanihin bases is supposed to be in the system. In the so-called uric-acid diathesis the nuclein derivatives are transformed in the intestine into one of the xanthin bases, and their presence in the system appears to give rise to eosinophilia. At any rate we regularly find eosinophilia (according to Neusser) in diseases thought to be characterized by an excess of these substances in the system. Examples of this are found in gout, bronchial asthma, emphysema, certain forms of migraine and epilepsy, oxaluria, uraemia, tetanus, some gastro-intestinal troubles, ankylostomiasis, after injections of nuclein, pilocar- pine, tuberculin, iron preparations, and in most non-malignant liver diseases. All these Neusser believes stimulate the sym- pathetic nervous system and hence the bone marrow, through the production of xanthin bases. In asthmatic patients he suc- ceeded in producing a paroxysm by injecting nuclein subcuta- neously. Possibly under this heading comes the eosinophilia after antipyrin, and that sometimes found in chlorosis, scurvy, nephritis, chronic malaria, and phthisical patients with cavities. In the latter cases it has been suggested that the patients may other hand, in a case of dermatitis herpetiformis in which I lately ex- amined, the differential count of five hundred cells showed : Poymorphonuclear neutrophiles, . . . 47 per cent. Small lymphocytes, ....... 25 " Large " 8 " Eosinophiles, 19 " Myelocytes, . 1 " EOSINOPHILES. 119 inoculate themselves with tuberculin absorbed from their lung cavities. 6. Tumors of the spleen are also accompanied by eosinophilia in some cases. Neusser does not explain this under the theory above sketched. Many acute mental troubles show eosinophilia, while chronic cases do not. Other causes of eosinophilia are phosphorus poisoning and injections of campherin. In Osier's clinic there have recently been observed three cases of trichinosis in which the eosinophilic cells were from the first increased, and continued to increase till in one case at the time of death there was 68 per cent of eosinophiles in a leu- cocytosis of 17,000. I have had one similar case. We also find eosinophilia in some cases of syphilis and syphilitic disease of the spinal cord (tabes dorsalis). DIMINUTION IN EOSINOPHILES. 1. During digestion. 2. After castration. 3. In febrile stages of pneumonia, grippe, typhoid, diph- theria, sepsis, and most infectious diseases accompanied by leu- cocytosis. That this is not due simply to the presence of fever is shown by the fact that in malaria and scarlet fever, despite high fever, eosinophiles may be increased. 4. In the moribund state eosinophiles are diminished or absent. In the post-critical stages of pneumonia and other infectious diseases the eosinophiles swing up above the normal. 4. Malignant disease, hemorrhage, and most of other causes, of leucocytosis also diminish the eosinophiles. DIAGNOSTIC AND PROGNOSTIC VALUE OF EOSINOPHILIA. Neusser has suggested the following points : 1 1. In the diagnosis between puerperal mania and puerperal sepsis, eosinophilia points to the former. 2. Between a tumor connected with the genital system and one not so connected, eosinophilia points to the former. 1 For none of which I can vouch. 120 CLINICAL BLOOD EXAMINATION. 3. In determining whether a given case of hysteria, neurosis, or psychosis is likely to be benefited by castration, the presence of eosinophilia favors the operation. 4. In malignant disease an eosinophilia points to a metas- tasis in the osseous system (tumors of the spleen are not in- cluded in this rule) . 5. In cases of doubtful syphilis eosinophilia combined with lymphocytosis (see above) speaks in favor of syphilis. 6. The diagnosis of any obscure form of " uric-acid diathesis" is helped by finding an increase of eosinophiles. 7. In distinguishing malignant liver disease from other liver disease eosinophilia points to the latter. 1. In the prognosis of chlorosis, eosinophilia is favorable. 2. In the prognosis of scarlet fever and scarlatinal nephritis the greater the eosinophilia the better the prognosis. 3. After hemorrhage increased eosinophiles show active re- generation of blood and good prognosis. 4. In pernicious anaemia eosinophilia is favorable for the same reason. MYELOCYTES. The occurrence of the myelocyte of Ehrlich in the circulat- ing blood is always to be looked upon as pathological, that is, as the intrusion of a variety of leucocyte naturally a stranger to the circulating blood and a permanent inhabitant of the marrow. Although it is so close morphologically to other varieties of leucocytes that we should certainly suppose it to be an inter- mediate stage between the large lymphocytes and the polymor- phonuclear neutrophiles, the fact that it does not occur outside the marrow in health speaks against the supposition. Of the occurrence of the myelocyte in leukaemia and perni- cious anaemia mention will be made under those diseases. The object of this section is to give a list of the other conditions under which it appears. Neusser ' has found small percentages of myelocytes in urae- mia, carbonic-acid poisoning, diabetes, syphilis, puerperal mania, osteomalacia, Basedow's disease, and sarcoma, also dur- ing menstruation. 1 Cited in Klein: Volkmann's "Saraml. klin. Vortrage," December, 1893. MYELOCYTES. 121 Capps found considerable percentages near death in general paralysis (see Book II., page 313). J. J. Thomas found them in myxcedema. The majority of other references to them in literature relate to different forms of grave anaemia. For example : (1) Hay em 1 speaks of cells apparently myelocytes (he did not use Ehrlich's methods) in cases of extreme anaemia. (2) E. Krebs 2 found them in severe anaemia. (3) Loos s describes them in the anaemia of hereditary syphi- lis, and Eille 4 finds them in the anaemia of acquired syphilis. (4) Neusser 5 mentions their presence both in pernicious anaemia and in chlorosis. (5) Hammerschlag 6 made a similar observation. (6) Engel 7 noted their presence in a case of what he cau- tiously calls "pseudo-pernicious anaemia ," and in diphtheria. (7) Arnold 5 mentions them. (8) Klein 9 gives a list of various diseases (besides leukaemia) , in which they have been found, many of which are essentially anaemic conditions. (9) Holmes 10 has found them in phthisis. I can confirm this observation. (10) The writer " found them especially in the anaemia secon- dary to malignant disease (see page 351). Besides these conditions the writer has found them occa- sionally in almost all the conditions in which leucocytosis or grave anaemia is present for example, in pneumonia, malaria, sepsis, peritonitis, granulating wounds, osteomyelitis, phleb- itis, rickets, Hodgkin's and Addison's disease, tuberculosis, and other diseases. 1 "Du Sang," Paris, 1889, p. 382. 2 Inaug. Dissert., Berlin, 1892. 3 Wien. klin. Woch., 1892, p. 291. *Loc. cit., 1893, No. 9. 6 Zoc. cit., 1892, No. 42. 6 Berlin klin. Woch., August 20th, 1894. 7 Virchow's Archiv, vol. cxxxv. 8 LOG. cit., vol. cxl. 9 Volkmann's "Sammlung klin. Vortrage," December, 1893. 10 New York Medical Record, September 5th, 1896. 11 Boston Medical and Surgical Journal, loc. cit. 122 CLINICAL BLOOD EXAMINATION. The most curious example of their occurrence known to me is the following : Mrs. W - had been starving herself more or less for six months from motives of economy. Two weeks before I first saw her she began to suffer with cystitis. From both these troubles she made a rapid recovery, which has persisted now eighteen months. There was at the first count a leucocytosis of 15,100; partly due to cyanosis, as she had just been having a chill. The red cells were 7,300,000. Haemoglobin, eighty -seven per cent. Differential counts were as follows : Date. Number of cells counted. May2d 800. Per cent. May 6th 1,000. Per cent. May 7th 400. Per cent. May 8th 400. Per cent. May 13th 1,000. Per cent. " Polynuclear neutrophiles". . . L y m phocy tcs. 82.7 8.6 82.2 12.5 83.6 9.4 80.2 11 3 68.5 25 2 Large mononuclear 8.2 1.5 2.0 6 5 2 Myelocytes .5 3.5 4.0 2 5 6 Eosinophiles . ... .0 .3 1.0 .0 5 What caused the presence of myelocytes I do not know. At that time I had never seen them in any curable disease and was alarmed by their appearing, but this case proves that they are not always of any importance. In a general way their presence seems to have about the same* significance as that of normoblasts, but they occur much more frequently. As a rule I think they indicate an acceleration of the function of those organs (marrow ?), by which red corpus- cles and granular leucocytes are furnished to the blood. Such an acceleration may be supposed to take place in leucocytosis, leukaemia, and pernicious anaemia, which are the chief condi- tions in which myelocytes appear in the blood. CHAPTEE IX. GENERAL PATHOLOGY OF THE BLOOD AS REGARDS HAEMO- GLOBIN, FIBRIN, LIP^EMIA, MELAN^EMIA AND HEMORRHAGE. HEMOGLOBIN. As stated above, the haemoglobin may increase and diminish in lines parallel to those of the red cells. In that case we sup- pose the amount of haemoglobin per corpuscle to be normal and the color index or valeur globulaire is said to = 1. Where the haemoglobin is diminished more than the count of corpuscles, we say that the color index is less than 1. For example, if a man has 5,000,000 red cells per cubic millimetre and only 50 per cent of haemoglobin, we estimate the color index by simply reducing the count of cells to a stated percentage (5,000,000 cells = 100 per cent of cells) and dividing this percentage into the haemo- globin percentage i.e., T VV = 0.5 = the color index. There- fore 4,000,000 red cells (= 80 per cent) with 60 per cent of haemoglobin give a color index of f^ = 0.75. The color index rarely goes above 1, except in pernicious anaemia (see below) . As a rule when the red cells are above the normal the haemoglobin rises equally, sometimes it lags behind a little, but rarely if ever does it rise higher than the cells. In most anaemias, as has been pointed out, the haemoglobin suffers markedly before any considerable loss of red cells takes place. In other words, the corpuscles seem to get thin before they die, and except in malaria, hemorrhage, and a few other cases they are not destroyed while in the full vigor of health. 1 The loss of haemoglobin is loss of albumin, the chief constit- uent of the cells, and hence is usually loss of weight. In general the changes in the haemoglobin are best studied in connection with changes in the count of red cells, and so far as they have not already been mentioned will come in under the various special diseases. 1 This is of course not literal. There is no reason to suppose that good- sized corpuscles get smaller. It is more likely that a smaller generation is sent out by the blood-making organs. 124 CLINICAL BLOOD EXAMINATION. FIBRIN. The fibrin network to be seen in normal blood during coagu- lation (see page 54) is increased in a considerable number of conditions. Hayem has studied these minutely, and described several varieties of arrangement of fibrin fibres as characteristic of special diseases, that is, he studied fibrin qualitatively as well as quantitatively, and also as regards the rapidity of its formation. The rate of fibrin formation is often not the same as the rate of coagulation. It is not parallel to the number of leucocytes or blood plates, at least not in all cases (malignant diseases, scurvy). In a general way we expect increased fibrin in infectious and inflammatory diseases, but there are notable exceptions to this. The greater the exudation and the freer it is (in a cavity or on the surface) the thicker the fibrin network, while so-called interstitial inflammations or such conditions as parenchyma- tous nephritis show little increase in fibrin. The seat of the lesions has no considerable influence, except as it modifies the nature of the lesion. An abscess in one place has the same effect as an abscess elsewhere, provided it is equally free or equally confined, and of the same contents. Tuberculosis does not increase fibrin if uncomplicated. Leucocytosis and fibrin behave alike in many respects, espe- cially in relation to the vigor of resistance which the individual opposes to a given infection. When the individual is so weak- ened that he does not react well against the infection, the leu- cocytes and fibrin are but slightly increased, whereas in a vigor- ous individual the same infection would have markedly increased both fibrin and leucocytes. But neoplasms raise the count of leucocytes without changing the amount of fibrin. In a general way fibrin increases and decreases as fever does, but often persists after fever is gone. The most marked fibrin networks are seen in pneumonia, acute articular rheumatism, suppurative diseases, and in scurvy. In erysipelas it follows the leucocytes (increased in severe, not in mild cases). In the early days of grippe it is increased. The fever of hysteria or chlorosis shows no increase of fibrin LIP^EMIA. 125 and post-hemorrhagic anaemia with or without fever shows none. Fibrin is diminished in pernicious anaemia, not increased in leukaemia, typhoid, malaria, malignant disease, non-suppurative diseases of liver, nephritis (except interstitial nephritis, where it may be increased), heart disease, purpura, haemoglobinuria (sometimes decreased). The most valuable point about the fibrin appears to be the absence of any increase in malignant disease, whereby a diag- nosis between the affection and a suppuration may be helped. Otherwise the information given by it is chiefly confirmatory of impressions given by other features in blood examination. LIP^MIA. The blood invariably contains small quantities of fat, espe- cially during digestion (v. Jaksch '). In the blood of persons suffering from a variety of diseases such as phthisis, diabetes mellitus, obesity, alcoholism., ne- phritis, and in some dyspnoeic conditions, suppressed menses, pregnancy, icterus, typhus, malaria, mental disease, diseases of the heart and pancreas, as well as in health, fat is occasionally to be seen in considerable quantities. Grawitz 2 finds that if the blood is collected in a fine capillary tube and this is kept in a horizontal position for some time, fat rises to the surface like cream, and can be seen with an oil-immersion lens in the form of fine drops. Gumprecht 3 demonstrated it with osmic acid, which stains the fat drops black, and proved them to be fat by dissolving them in ether, xylol, etc. Lipaemia has no special significance so far as is known, and is not characteristic of the diseases above mentioned. Its cause is unknown. [In almost any preparation of the fresh blood fat drops are to be seen unless the patient's skin is washed with alcohol be- fore puncturing. Even with these precautions a few drops, may often be seen in healthy people's blood.] 1 "Klin. Diagnostik," p. 75 (English translation). 2 Loc. ctt., p. 180. 3 Dent. md. Woch., 1894, No. 39. 126 CLINICAL BLOOD EXAMINATION. MELAN^MIA. In malaria the occurrence of a black pigment in the leu- cocytes which have taken plasmodia into themselves, is gener- ally to be seen during and shortly after a paroxysm. Pigment free in the blood is only to be seen at the moment of seg- mentation among the new generation of parasites. The same condition has been observed in relapsing fever and in persons ; suffering from melanotic malignant tumors, the pigment being always in the white corpuscles. Presumably it must at some time be free in the plasma, but it is rarely if ever seen outside the cells. In Addison's disease Tschirkoff ' observed pigment in the leucocytes. HEMORRHAGE. Women can stand a greater hemorrhage and yet live than men can. Children, on the other hand, succumb to compara- tively slight hemorrhages (cf. Blood in Infancy, page 385). In- dividual differences make a great difference in the ability to survive hemorrhage, and no exact amount of blood can be ^stated as the maximum that any one can lose and yet survive. Changes in the Blood Eesulling from Hemorrhage. The red cells and haemoglobin of course suffer proportionally .at first; later the haemoglobin in the newly formed cells is always deficient (see below). The striking point in the blood after hemorrhage is the evi- dence it gives us that ey^en before the hemorrhage has ceased the other tissues begin to contribute fluid to make up the volume upon which life depends. The serum is markedly diluted by this fluid, but still serves to give the heart something to contract on and so prevents blood pressure from falling as fast as it otherwise would do. Were it not for such contributions from neighboring tissues the organism could sustain but slight hemorrhage without succumbing at once. We have then after .hemorrhage a diluted or hydrsemic blood, even though we do it. f. klin. Med., vol. xix., 1891. BLOOD REGENERATION. 127 not assist the efforts of nature by contributing fluid by intra- venous or rectal injection. Behier reports a case due to trauma in which the count was only 688,000 per cubic millimetre. Coagulation increases in rapidity the more blood is lost, so that after severe hemorrhage it takes place almost instantly. BLOOD [REGENERATION. The regeneration of the blood after hemorrhage may be taken as typical of the same process in anaemia from other causes. The length of time needed for full restoration to normal depends not merely on the (a) amount of blood lost, but also on the (b) age and nutrition of the patient as well as upon (c) the methods of treatment carried out and the existence of (d) other disease (typhoid, malignant disease, phthisis, etc.). Allowing for these other conditions we may say that, other things being favorable, the loss of I. Less than 1 per cent of the blood mass is made up in 2 to 5 days. II. From 1 to 3 " " " 5 rt 14 " III. 3 " 4 " " " " " 14 30 " The last amount means a very severe hemorrhage. Few surgical operations involve the loss of over three per cent, and after such accordingly we expect the blood to be normal again in two weeks, provided the individual is otherwise sound (see Malignant Disease, page 335) . Young, well-nourished persons are of course quicker in mak- ing up losses than the old and weak. Blood Condition During Regeneration. 1. Red Cells. (A) As previously mentioned, the haemoglo- bin becomes relatively low as soon as the regenerative process is well established, and as recoverj^ progresses the red cells are almost always normal in numbers for some time before the stat- ure, weight, and color of the individual cells is what it should be. A color index of 0.50-0.60 is not unusual in short, what some call a " chlorotic" condition of the blood. (B) Qualitative changes are those already described on page 72, namely : (a) Deformities in size and shape with an average diminution in size; (b) polychromatophilic cells; and (c) nucle- 128 CLINICAL BLOOD EXAMINATION. ated corpuscles. These latter are almost exclusively of the nor- moblast type, but an occasional megaloblast has been observed. Blood Crises. Yon Noorden was the first to notice that in some cases nucleated corpuscles are to be found in the circu- lation in great numbers for a few hours only, the blood examina- tion both before and after showing few or none at all. The name of " blood crisis" has been given to these sudden outpour- ings of nucleated red cells ; they are to be observed during re- covery from various forms of anaemia. 2. Willie Cells. Immediately after a loss of blood we can usually find a decided leucocytosis despite the dilution of the blood (see above, Post-hemorrhagic Leucocytosis). This leucocytosis is in no way different from those occurring from other causes. The percentage of polymorphonuclear cells is usually increased, and the eosinophiles often disappear. As pointed out by Stengel we may have a lymphocytosis after hemorrhage. A case of anaemia from bleeding piles, in which the red cells were 2,723,000 and the haemoglobin 35 per cent, showed in a total leucocyte count of 4,200, 69 per cent of small lymphocytes and only 28 per cent of polymorphonuclear cells. Leucocytosis if present is rarely very high, seldom reaching over 30,000. It is not invariably present, or if present sometimes is of very short duration. Thus in a patieut whose red cells were reduced to 3,200,000 by a profuse uterine hemorrhage the white cells counted next day were only 8,000; while in the next bed of the hospital was a woman crushed in a railroad accident whose red cells were 1,280,000, and the white cells 28,000, the usual state of things. The leucocytes may be increased even by a cerebral hemor- rhage which is not large enough considerably to affect the red cells in most cases. Ten apoplectic cases (with autopsy) ob- served at the Massachusetts Hospital showed such counts as the following : 1. Eed cells 5,512,000, white cells 25,000, Hb. 85 per cent. 2. Bed cells 5,560,000, white cells 15,600, Hb. 90 per cent. Whether the leucocytes are here affected by any influence other than that of hemorrhage I do not know. The effect of transfusion (intravenous saline solution) is ap- parently at first to increase the leucocytosis. D , a patient with traumatic rupture of the urethra, had BLOOD REGENERATION. 129 had severe hemorrhage for forty-eight hours before it was checked at 1 P.M., November 1st, 1895. At 4 P.M., his pulse being 165, the count showed: red cells, 3,304,000; white cells, 10,400. He was at once given a pint of sterilized normal salt solution by intravenous injection under the strictest asepsis. Ten minutes after the transfusion the leucocytes numbered 32,400. One hour later they were 24,700, and the red cells 3,632,000. Four hours later leucocytes, 31,900; red cells, 3,046,000. The later counts were as follows : Red cells. White cells. November 2d : good pulse 3, 608, 000 34, 600 3d (5 P. M. ) : good pulse 2, 944, 000 30, 200 3d (4 p. M. ) : good pulse 2, 928, 000 15, 800 13th 3,360,000 16,600 A good recovery was made. IMPORTANCE FOB SURGERY OF BLOOD COUNTING AFTER HEMORRHAGE. Mikulicz, who as a surgeon should speak with authority and who always takes account of the condition of the blood in his cases, lays down (following Laker) the following rule: Never operate on any case when the Immoglobin is below thirty per cent. The question of operating at once or waiting for recovery from "shock," is a very common one in the accident rooms of any hospital and is generally settled on general impressions of the patient's vigor. We know, say, that he has lost blood, but we have no way of ascertaining how much. If his " shock" is due to hemorrhage he may need transfusion ; if it is due to cerebral concussion or compression, the transfusion will do more harm than good. The blood count can settle these questions, and could reveal much which is now obscure, if it were more fre- quently employed in surgical cases and a standard like that of Mikulicz worked out. In cases of suspected ruptured tube in extra-uterine preg- nancy, the question of whether the patient is suffering from in- ternal concealed hemorrhage can be settled in many cases by the blood count, which will show a decided loss of red cells if 9 130 CLINICAL BLOOD EXAMINATION. the hemorrhage is large, and thereby distinguish the condition from peritonitis, obstruction, or strangulated hernia, none of which affects the red cells. Any other concealed hemorrhage, as for instance from ruptured kidney or spleen or liver, may be indicated by the blood count when by other physical signs the diagnosis might be very difficult. Summary. The blood count is of importance after cases of supposed hemorrhage. 1. To ascertain whether such has taken place. 2. Its extent. 3. Whether operation is to be immediate or not. 4. Whether transfusion is indicated. 5. How soon the patient has got back enough blood to make operation worth while. CHRONIC HEMORRHAGE. Piles, uterine disease, haemophilia, purpura, and other causes may produce a long-standing drain on the blood. Some patients apparently can lose a little blood almost daily for years without acquiring any severe anaemia, and if the indi- vidual is otherwise sound and does not suffer from an underly- ing disease like phthisis, cancer, or nephritis, he can probably go on for a long time without showing any bad effects from the repeated small hemorrhages. How much he can stand we have no way of judging, for we cannot measure the amount of blood lost. When, however, such small repeated losses do produce an anaemia, regeneration is apt to be much slower than after a single large hemorrhage. The longer the drain has been going on the poorer the chance for recovery, and the slower the latter will be if it does take place. Gain in body weight does not always mean gain in corpuscle substance as well (see Malignant Disease, page 335) . BOOK II. SPECIAL PATHOLOGY OF THE BLOOD. PART I. DISEASES OF THE BLOOD. CHAPTEE I. THE PRIMARY ANAEMIAS. 1. THE BLOOD IN PERNICIOUS ANEMIA. THE definition of the disease has been sufficiently explained before (see page 84) and we can proceed at once to the descrip- tion of the blood. 1. Gross appearances. (a) The drop as it emerges from the puncture is often exces- sively pale and watery, but not more so than may occasionally be seen in secondary anaemia or chlorosis. Sometimes it is not nearly so pale as in other cases with equally low counts, a fact which may be due to the increased color index sometimes present (see below). In one case (color index 1.2) I have seen the blood as red as normal. Another appearance, which I have frequently observed in this and other anaemias, is an uneven, streaked color in tho drop, as if the cells were unequally divided in the plasma. (&) As striking as the color of the drop is its great fluidity ; the rapidity with which it slips off the ear or finger often makes it difficult to suck it up in time. It is usually very slow in coagulating. 2. The fresh specimen in most cases shows no rouleaux forma- tion, a diminution in blood plates and fibrin, and usually great variations in the size and shape of the corpuscles with a ten- dency to an oval shape and an increase in the average diameter. Not infrequently the deformed corpuscle shows active pseudo- amoeboid motions of its projecting points or of the cell as a whole. The great lack both of red and white cells is noticeable even in the fresh specimen. 134 SPECIAL PATHOLOGY OF THE BLOOD. Red Cells and Haemoglobin. (a) Quantitative changes (see Table I.). The average count of red cells in the sixty-eight cases of my table is about 1,200,000, which may be taken as the average count in patients seen at the stage of the disease at which they feel sick enough to seek medi- cal advice. 1 We very rarely get an opportunity to examine the blood in the early stages of the disease, so that we have to judge of them chiefly from the evidence given during the remission so commonly observed. In the relapse following such a remission the blood count may fall from 5,000,000 to 1,000,000 in a period of from six weeks to six months. In the later stages of the dis- ease 500,000 red cells per cubic millimetre is not rare, and if the diminution has been gradual, the patient may be up and about and able to do light work with a count no greater than this. I had an opportunity to observe such a case in the wards of Dr. "F. C. Shattuck at the Massachusetts General Hospital five years ago, where for several weeks the blood count remained at or near 500,000, yet the patient was outdoors daily, read the papers, and seemed perfectly comfortable. Evidently it is not the anaemia itself which kills the patient. The lowest count on record is that reported by Quincke 143,000 per cubic millimetre. TABLE I. (a) FIRST COUNT. (6) HlGHES T COUNT. (c) LOWES! c 1 COUNT. Total 1 Red cells. Per cent haemo- globin. Red cells. Per cent haemo- globin. Red cells. Per cent haemo- globin. number of exam- inations. 1 2 3 4 5 450,000 490,000 503,000 510,000 600,000 10 ? 10 20 24 658, 000 490,000 522,000 680,000 18 ? 18 ? 363,000 410,000 368, 000 510, 000 10 9 10 20 6 2 6 3 1 6 7 8 9 10 11 12 13 630,000 670,000 680,000 694,000 735,000 784, 000 842,000 896,000 V 20 20 ? 14 ? 18 658,000 670,000 680,000 2,654,000 1,500,000 784,000 842, 000 896,000 ? '26' v 14 ? 18 450,000 670, 000 680,000 694,000 730,000 784,000 842,000 430, 000 ? '26 20 ? 14 ? 6 13 1 1 8 3 1 1 3 l Cf. Schaumann : Out of his 38 cases, 1 was over 2,000,000; 26 be- tween 1,000,000 and 2,000,000 ; 11 below 1,000,000 ; average 1,290,000. THE BLOOD IN PERNICIOUS ANJEMIA. 135 TABLE I . ( Continued) . 6 fc (a) FIRST COUNT. (6) HIGHEST COUNT. (c) LOWEST COUNT. Total number of exam- inations. Red cells. Per cent haemo- globin. Red cells. Per cent haemo- globin. Red cells. Per cent haemo- globin. 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 896, 000 962,000 963, 000 984, 000 988, 000 992,593 1,018,000 1,096,400 1,064,000 1,060,524 1,092,000 1,111,000 1,113,000 1,126,000 1,137,000 1,140,000 1,150,000 1,150,000 1,176,000 1,200,000 1,226,284 1,270,000 1,280,000 1,288,000 1,289,000 1,296,000 1,300,000 1,336,000 1,344,000 f, 364, 000 1,493,000 1,498,000 1,500,000 1,516,000 1,582,000 1,583,000 1,600,000 1,627,000 1,632,000 1,755,000 1,768,000 1,800,000 1,800,000 1,800,000 1,819,000 1,872,000 1,884,000 1,920,900 1,929,000 17 15 38 28 ? 34 20 12 23 35 25 ? 18 ? * 20 20 ? 30 29 15 25 16 24 ? 32 25 28 28 23 35 32 20-30 35 35 20 20 25 ? 28 20 ? 28 25 30 34 25 '33' 38 3,800,000 1,028,080 70 15 608,000 962, 000 15 15 20 3 1 3 1 2 1 4 5 1 5 3 8 2 4 3 16 1 1 1 12 2 16 3 5 3 2 3 3 1 2 1 8 3 12 4 10 1 7 2 6 5 1 3 3 4 4 1 1,080,000 ? 992,593 34 1,096,400 1,420,000 1,150,000 1,111,000 2,820,000 1,126,000 1,137,000 1,140,000 2,802,000 13 25 '39' ? ? ? 20 20 624,000 1,064,000 13 23 ? ? ? 15 25 893,000 756, 000 1,038,000 1,100,000 550,000 622,160 1,150,000 4,450,000 2,700,000 1,328,000 1,910,000 1,628,000 1,296,000 1,300,000 1,336,000 1,344,000 65 30 29 ? 34 25 28 28 23 762,000 "664,000 1,288,000 1,121,000 1,248,000 970,000 956,000 758,000 ? '22" '34' 25 30 20 17 1,500,000 ? 3,700,000 1,916,000 4,760,000 1,768,000 4,032,000 53 35 52 40 80 1,460,000 1,516,000 1,624,000 1,500,000 1,288,000 30 35 30 20 23 1,632,000 1,755,000 2,458,000 2,868,000 28 20 ? 41 '36' 25 '35' 1,180,000 1,117,000 1,768,000 1,508,000 30 20 ? 31 1,800,000 1,872,000 1,889,314 1,930,000 1,768,000 1,144,000 1,330,000 1,600,000 22 30 '28* 136 SPECIAL PATHOLOGY OF THE BLOOD. TABLE I. (Continued). 6 & (a) FIRST COUNT. (6) HIGHEST COUNT. (c) LOWEST COUNT. Total number of exam- inations. Red cells. Per cent haemo- globin. Red cells. Per cent, haemo- globin. Red cells. Per cent haemo- globin. 63 64 65 66 67 68 1,946,000 1,984.000 2,000,000 2,080,000 2,076,000 2, 524, 000 Average= 1,200,000 40 39 20 25 15 26 26 39 20 70 45 26 3 3 2 9 10 5 293 ?e = 4 + 1,984,000 2,000,000 5,056,000 4,500,000 2,524,000 598,000 1,200,000 1,632,000 1,384,000 1,280,000 15 20 50(?) 10 13 Averai The great but temporary improvements above alluded to, followed by relapse, occur either with or without treatment. In the course of a few months the count of red cells may rise w '.4000000 -"Haemoglobin vn/\rvi j. . t, 70 \ N / 9 ."X .. 60 3000000 V ) " " V \ 50' i r / \ -V' J 40; zoooooo _5J \ j \ , 3$ I 2 \ Vf f r ! * 20/ 1000000 N - v* 10; V Red Cells Haem 4500000 CHART ]i ^ 3000000 50' /u ,2,000000 40% 30% .1000000 ao% 10% 1* / _7 y r - ^ > * / \ . 2 3 ... ** . .. ^TT^edCeUs 3000000 60% Z5 ooooo sol ZQOOOOO 40} 1500000 30f 1000000 Ml 500000 10/ v CHART m j \ ^ ... ^^ ^; - ., ".. "^ ^ \ ^ s \ \ tHE BLOOD IN PERNICIOUS ANEMIA. 137 to normal, the nucleated corpuscles (see below) disappear, and the patient is apparently restored to health and goes to work with a laugh at the doctor. I have followed one case through five such relapses in a period of three years before the fatal issue came. Frequently the patient feels so well during one of these remissions that he goes to work and is lost sight of, and, under such conditions, the incautious are apt to report "cure." The accompanying charts l show the three types usually met with ; No. II. being, of course, only a fragment of a case similar to No. I., while the steady progression of No. III. may have been preceded by a rise from a former downfall, though no such history was obtained. Looking over a considerable number of cases, one can hardly help being struck with the tendency of the count to remain near the figure 1,000,000. Cases rarely remain sta- tionary at, say, 2,000,000, and often die without sinking be- low 1,000,000. It seems as if some self-applying mechanism tended to arrest the destruction of corpuscles at or near this point (see Table I.). In counting the red cells some difficulty and error may result from 'the very small size of some of the cells. It is especially important that the diluting solution should be clean and freshly made, else without the aid of a stain it may be hard to distin- guish the dwarf cells or microcytes from bits of extraneous sub- stance. Quantitative Changes. White Corpuscles (see Table II.). The rule is a very consid- erable diminution in the number of leucocytes. Thus of sixty cases which I have examined forty -two were under 5,000, the average of all being 3,800. [I have excluded from this series counts made immediately after hemorrhages and counts in infants. The latter are very apt to show a leucocytosis in connection with any form of anaemia.] As the disease progresses the leucocytes fall even more rapidly than the red cells, and counts as low as 500 white cells per cubic millimetre are not uncommon. 1 The number of perpendicular lines represents the number of weeks. 138 SPECIAL PATHOLOGY OF THE BLOOD. Leucocytosis when present in the blood of adult cases is always due to some complication like hemorrhage or suppura- tion. TABLE II. WHITE CELLS FIRST EXAMINATION. No. White cells. No. White cells. No. White cells. 1 . . . 400 500 800 1,000 1,000 1,000 1,500 1,600 1,800 2,000 2,000 2,000 2,000 2,000 2,300 2,600 2,800 2,800 2,800 2,800 21 2,900 3,000 3,000 3,200 3,200 3,300 3,400 3,500 3,600 3,700 3,704 4,000 4,000 4,000 4,000 4,200 4,300 4,400 4,500 4,720 41 4,828 4,900 5,000 5,200 5,300 5,500 5,600 6,000 6,000 6,000 6,400 6,500 7,000 7, 200 7,500 7,600 9,000 9,600 10,000 10,100 = 3,8004- 2 22 42 3 23 43 4 24 44 5 25 45 6 26 46 7 27 28 47 8 48 9 29 30 49 10 50 11 31 32 51 12 52 13 33 53 14 34 54 15.. 16 35 36 55 56 57 17 37 18 38 . 58 19 39 59 20 40 60 Average As mentioned above, the blood plates and fibrin are much diminished. In four cases in which Dr. Lindstrom, of Boston, was kind enough to give massage, we were unable to see the slightest gain either in corpuscles or haemoglobin, such as can be produced temporarily in most healthy persons. The observations of J. Mitchell on this point we were unable to confirm. Haemoglobin. . The great majority of cases of pernicious anaemia have a relatively high percentage of haemoglobin (e.g., 1,000,000 red cells and 35 per cent, of haemoglobin, or a color index of 1.75). In some cases this is not so, and in others we cannot tell whether it is so or not, owing to the unreliability of the v. Fleischl instrument when used for very low haemoglobin per- centages. Of the 50 cases in the series on page 134, in which the haemo- THE BLOOD IN PERNICIOUS ANAEMIA. 139 globin was tested, a color index of over 1 was apparently present in 29, or 58 per cent, and a color index of less than 1 in 21, or 42 per cent, of the cases. How many of these haemo- globin estimations may have been wrong I cannot say. From the frequency with which we find the corpuscles well stained and larger than normal in pernicious anaemia (see below), we should expect that the haemoglobin would be relatively high, and in a larger percentage of cases than the v. Fleischl instru- ment indicated. An increased color index is probably a bad prognostic sign. In the remissions of the disease when the cells are increasing fast, the haemoglobin lags behind and the color index is low. As the relapse follows, the color index in many cases progres- sively increases. Cases ivhose color index is low and in which the average diameter of the red cells is normal are apt to be gaining at that time, while those with high color index are apt to be losing at that time. The average color index in the cases in which the haemo- globin and red cells were both tested was 1.04, the average percentage of haemoglobin being 26 and of corpuscles 24 (-1,200,000). QUALITATIVE CHANGES. 1. Red Corpuscles. (a) Increase in the average diameter of the cells is a very constant and striking feature of the stained specimens in this disease. In no other disease do so large cells or so many of them occur. Out of forty -eight cases in which I have looked for this point, forty-one showed the increase, as far as could be judged without actually measuring any large number of cells. This does not mean that every cell is larger than normal, but that those larger than normal outnumber those undersized; the "macrocytes" are more numerous than the "microcytes." Oc- casionally we see cells over 20 v in diameter, some with nuclei, some without. (b) Deformities in Shape. The eye soon gets used to the shapes assumed by the necrobiotic corpuscles and learns to dis- tinguish them from the distortions due to technique or to crena- 140 SPECIAL PATHOLOGY OF THE BLOOD, tion. Most of them fall under one or another of the types shown in Plate IV. Litten has laid particular stress on the horseshoe forms, and thinks them peculiar to pernicious anaemia. The battledore and sausage-shaped forms are very common. In one case I found all the red cells of the latter shape, so that they looked at first sight like a lot of gigantic bacilli. That this ap- pearance was not due to the technique 1 " (as I had at first sup- posed) is probable from the fact that the rod-shaped cells did not point all in one direction as they would have done if pulled out of shape by the process of spreading (see Fig. 30) . This ap- PIG. 30. Elongated or Oval Corpuscles in Pernicious Anaemia, resembling the blood of lower animals. pearance is only an exaggeration of what may be seen in most severe anaemias, namely, a tendency toward an oval shape like that of amphibian corpuscles. This is usually true of those cells (in pernicious cases) which are not more violently deformed. Occasionally we see cases with no considerable deformities whatever in the red cells. In nine cases out of sixty in which 1 Some writers advise the use of less heat than usual in dealing with cover-glass specimens of pernicious anaemia. I have not found this so, and heat as usual up to 150 C. and then stop. THE BLOOD IN PERNICIOUS ANAEMIA. 141 this point was observed, little or no deformity was noted. I cannot make out that such cases have any better or worse prog- nosis than others. I have never seen cases whose red cells were all undersized, but a normal average diameter was present in somewhat under one-quarter of the cases in which I have looked out for this point. (c) Staining Properties of the Red Cells. The white spots or streaks described by Maragliano, Hayem 3 and others are very often seen in the red cells of pernicious anaemia despite good technique. Some corpuscles are so pale in the centre that we see only the narrow ring of stained protoplasm at the periph- ery, a mere shell. Others are swollen up so as to show no sign of central biconcavity, and stain deeply and evenly all over. More common than in any other form of anaemia are the polychromatophiiic red corpuscles (see Plate IV.) which with the Ehrlich-Biondi mixture stain brownish, purple, or gray, either as a whole or in parts. In the nucleated red cells the pro- toplasm is very apt to show this change, so that it is often diffi- cult to distinguish them from lymphocytes. In difficult cases we have sometimes to fall back upon the appearances of the edge or periphery, which in most red corpuscles shows some thin place or crinkle characteristic of a flat cell, while the lym- phocyte gives us the more solid-looking outline of the spherical cell. All these microchemical changes can be better brought out with haematoxylon-eosin or eosin-methyl-blue stains, but all that is needed for clinical purposes can be made out with the ordinary Ehrlich-Biondi mixture. Nucleated Red Corpuscles. Nothing further needs to be said in description of these forms (see above, pages 89-94). We have no exact method of estimating the number of nucleated cells either in relation to the whole num- ber of red cells or in a cubic millimetre. All we can do is to note the number seen in such an area of a cover-glass specimen as is covered while counting a given number of white cells, say 1,000. Knowing the ratio of red to white corpuscles, we can calculate from this number of nucleated red cells their approximate re- lation to the whole number of red cells. Thus if the ratio of white to red be 1:1000 (1,000,000 red and 1,000 white) and we have seen two nucleated red corpuscles 142 THE BLOOD IN PERNICIOUS ANJEMIA. while making a differential count of 1,000 white cells, the total number of red cells passed over must be approximately 1,000,- 000 and the number of nucleated corpuscles about two in 1,000,- 000 red cells or two in a cubic millimetre. Of course where leucocytosis is present and the ratio is raised say to 1 : 150 (10,000 white and 1,500,000 red) finding two nucleated rec! cells while counting 1,000 white would mean that there were two nucleated cells in every 150,000 non-nucleated, or twenty in a cubic millimetre (or in 1,500,000 non-nucleated cells). Such calculations are inaccurate because we are never sure that the red cells and white cells are distributed in the dried speci- men exactly as they are in the blood. Part of the leucocytes may be accumulated at the edges of the cover-glass so that the ratio in the middle may be different from that in the circulating blood. Nevertheless we can get some idea of how plentiful the nu- cleated corpuscles are, and as their significance in prognosis depends far more on their kind than on their number, greater accuracy as to the latter is not at present important. For in- stances, two megaloblasts per cubic millimetre mean a worse prognosis than twenty normoblasts, provided there are no other kinds present in either case. It is the ratio of megaloblasts to normoblasts and not the absolute number of each, that is of im- portance. In all of the sixty cases of pernicious anaemia in which I have examined the blood, the number of megaloblasts has ex- ceeded the number of normoblasts, and as the cases grew worse the megaloblasts grew relatively more numerous (often abso- lutely as well). Further, in several hundred cases of severe secondary anaemia I have never yet seen the number of megalo- blasts exceed the number of normoblasts. The range of variation in the number of nucleated cells pres- ent has extended in my series from 6 per cubic millimetre to 7,100 per cubic millimetre (see Table III.). The calculation can be made by using the following formula. Let n = the number of white cells counted (by differential count). " m = " " nucleated red cells seen while counting these. " p =: " " white cells per cubic millimetre (Thoma-Zeiss). p x = x = number of nucleated red cells per cubic millimetre. The search for nucleated corpuscles in pernicious anaemia THE BLOOD IN PERNICIOUS ANEMIA. 143 is sometimes the most laborious undertaking in all blood ex- amination, but it is also one of the most important. We may search two or three hours before finding one nucleated cor- puscle, but on^ that corpuscle may hang the character of our prognosis. If it be a megaloblast and no other nucleated red corpuscles are seen, the prognosis is bad, and it is impor- tant that we should know it. This is particularly true when the case is seen during a remission, for under these conditions we might never suspect a case of pernicious anaemia but for the presence of megaloblasts. They are not always difficult to find; indeed, in one of my cases they were nearly as numerous as the white cells, but, as a rule, we do not get off with less than two hours' work. The following table (Table III.) shows the number of nucle- ated corpuscles per cubic millimetre in thirty of the cases ex- amined by the writer. TABLE III. NUMBER OF NUCLEATED RED CELLS PER CUBIC MILLIMETRE IN THIRTY CASES OF PERNICIOUS ANAEMIA. Case Number. Total nucleated red cells. Megaloblasts. Normoblasts. Microblasts. 1 . 7,100 5, 300 1,325 475 2 6,468 3,476 924 2,068 3 854 574 266 14 4 277 277 5 240 160 80 6 229 123 106 7 208 130 78 8 200 134 66 9 117 103 14 10 116 80 36 11 114 95 19 12 112 96 16 13 96 96 14 96 84 12 15 92 59 33 16 46 26 20 17 '. 45 36 9 18 39 33 6 19 35 32 3 20 28 26 2 21 28 21 7 22 28 28 23 18 12 6 24 14 14 o 25 11 11 26 11 10 1 27 11 9 2 28 9 6 3 29 8 7 1 30.... 3 2 1 144 SPECIAL PATHOLOGY OF THE BLOOD. White Corpuscles. Qualitative CJianges. Unless the cover-glasses are spread unusually thick, it may take a long time to find enough leu- TABLE IV. PERCENTAGES OF LEUCOCYTES IN PERNICIOUS ANAEMIA. LYMPHOCYTES, LARGE AND SMALL. EOSINOPHILES. Number of counts. No. Per cent. No. Per cent. 1 79. 77. 71. 61.6 58. 57.6 57.2 57. 56.9 56. 53.9 53.8 51.5 49.5 49.4 47.9 47.9 47. 46. 45.9 45.5 44.7 43.7 42.2 41. 40.8 40.5 39. 38. 38. 37.8 36.1 35.7 35.6 35.6 34. 1 9. 6.2 6. 4.7 4.6 4.5 4.4 4.3 4. 4. 4. 3.7 3.5 3.4 3.4 3.1 3. 2.8 2.7 2.6 2.6 2.6 2.6 2. 2. 2. 1.6 1.6 1.5 1.5 1.5 1.5 1.4 1.2 1.2 1.2 1 1 1 2 2 3 1 1 1 5 1 2 1 2 1 1 3 2 2 1 3 2 1 1 1 2 1 1 5 2 1 1 1 3 1 1 2 2 3 3 4 - . 4 5 5 6 6 7 7 8 8 9 9 10 10 ... 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 ... 23 24 24 25 25 26 26 27 27 28 2$ 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36.. 36.. cocytes for an accurate differential count, so great is the leuco- penia in many cases. It is worth while, therefore, to spread some cover-glasses more thickly than would be advisable if we had only the red cells to examine. Such preparations should be dried at once by artificial heat. Lymphocytosis is the chief feature (see Table IV.). THE BLOOD IN PERNICIOUS ANAEMIA. 145 TABLE IV. PERCENTAGE OF LEUCOCYTES IN PERNICIOUS ANAEMIA (Continued) . LYMPHOCYTES, LARGE AND SMALL. EOSINOPHILES. Number of counts. No. Per cent. No. Per cent. 37 33.6 33.1 33. 33. 31.8 29.4 28.7 28.4 27.3 27.2 26.5 24.2 22. 21.2 19.8 16. 37 1. 1. 1. .8 .8 .8 .7 .6 .5 ? .3 .2 .0 .0 .0 .0 1 2 1 1 1 1 2 1 1 1 38 38 . . 39 40 39 40 41 41 42 42 43 43 44 44 45 45 46 46 47 47 48 48 49 49 . 50 50 .. 51 51 52 52 In 52 cases examined by myself the lymphocytes (large and small) averaged 45.4 per cent. About nine-tenths of these were small forms. As the fatal termination approaches, the percentage of lymphocytes rises. An extreme case of this change has already been recorded on page 90. Two other cases showed respectively 71 and 79 per cent of lymphocytes a few days before death. The polymorphonuclear cells suffer proportionately as a rule. On the other hand, Ewing has observed a marked rise in the percentage of the poly- nuclear cells near death, although autopsy revealed no com- plication. Eosinophiles are occasionally increased, 9 per cent being present in one of my cases, 6.6 per cent in another. The aver- age of 78 examinations in my 52 cases is 2.7 per cent. Small percentages of myelocytes are the rule. They are present in 42 of my 52 cases. The following table shows the percentages : 10 146 SPECIAL PATHOLOGY OF THE BLOOD. TABLE V. No. Percentage of myelocytes. No. Percentage of myelocytes. No. Percentage of myelocytes. 1 9.2 8.8 8. 6.3 6. 4.6 4. 4. 3.6 3.4 3. 3. 2.7 2.5 2.2 2.2 2.2 2.0 19 1.8 1.5 1.5 1.5 1.4 1.2 1. 1. 1. 1. 1. .8 .8 .8 .6 .6 .6 .6 37 . 38 0.6 .6 .5 .4 .3 .2 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 : 2 per cont. 2 20 3 21 39 ... 4 22 40 . . 5 23 41 6 24 42 43 7 25 8 26 44 9 ... 27 45 .. 46 10 N 28 11 29 47 12 30 48 13 31 49 14 32 50 15 33 51 16 34 52 17 35 Average - 18 36 As has been explained above (page 121), the inyelocyte is found in a great variety of affections, although very sparingly in most, but, so far as my observations go, its presence is more constant and the percentages run higher in pernicious anaemia than in any other disease except leukaemia. I am speaking now of percentages. With a leucopenia such as is usually present in pernicious anaemia, 2 per cent of myelocytes means absolutely a very small number per cubic millimetre. Taking 3,800 leucocytes per cubic millimetre as the average for pernicious anaemia (see above, page 137) 2 per cent of mye- locytes amounts to only 76 per cubic millimetre. In leukaemia the absolute number of myelocytes is seldom under 150,000 per cubic millimetre. The more important characteristics of the blood of pernicious anaemia are as follows : 1: Red cells about 1,000,000 per cubic millimetre. 2. White cells much diminished. 3. Haemoglobin variable, sometimes increased relatively (= high-color index). 4. Deformities in size and shape of red cells in many cases. 5. Increase in average diameter of red cells. 6. Polychromatophilic red cells. THE BLOOD IN PERNICIOUS ANAEMIA. 147 7. Megaloblasts more numerous than normoblasts. 8. Lymphocytosis. 9. Small percentage of myelocytes. The items italicized are the most important and character- istic. Diagnostic Value. 1. Pernicious ancemia and chlorosis may be indistinguishable without the examination of the blood. The pallor of the two diseases is not always different either in degree or in kind, and the symptoms and physical signs may be identical. The differential diagnosis is easily made by the blood. The red cells rarely reach as low as 2,000,000 in chlorosis and the number and degree of degenerative changes are less than in pernicious anaemia. Megaloblasts have been seen in chlorosis (Hammerschlag) but have never constituted a majority of the nucleated red cells present. In the great majority of cases the pallor and other signs and symptoms of chlorosis are due to lack of haemoglobin per corpuscle (for the corpuscles are not only pale but very small-sized), and not to a lack of corpuscles. The high-color index and large size of the scanty cells in per- nicious anaemia constrast strongly with this. The white cells are about the same in both diseases, though usually fewer in pernicious anaemia. Lymphocytosis is common to both diseases. Myelocytes are occasionally found in chlor- osis, but much less commonly than in pernicious anaemia. 2. Pernicious Ancemia and the Anaemia of Malignant Disease. Not long ago I examined the blood of a gentleman who had gradually and without assignable cause acquired a " lemon-yel- low" pallor, without loss of flesh, vomiting, pain, or any localiz- ing sign or symptom. The diagnosis of pernicious anaemia had been made. To my great surprise I found over 4,000,000 red cells, with only 38 per cent of haemoglobin, and 18,780 white cells, 86 per cent of which were polymorphonuclear neutrophiles. One normoblast was seen. Fibrin was not increased. The anaemia was evidently secondary, and the autopsy ten months later showed cancer of the stomach. Malignant disease may bring down the count of red cells to 1,000,000 or lower, but in such cases leuocytosis is usually present. As will be seen in the chapter on malignant disease, 148 SPECIAL PATHOLOGY OF THE BLOOD. leucocytosis is by no means invariable in the anaemia of can- cerous growth, but in those cases which cause such an anaemia as to resemble the counts of pernicious anaemia, leucocytosis is the rule. This in itself is usually sufficient to exclude uncom- plicated pernicious anaemia. Where an increase in the whole number of leucocytes is not present in malignant disease, there is often an increased percentage of polymorphonuclear cells, contrasting strongly with the increased percentage of lym- phocytes in pernicious anaemia. Normoblasts and not megalo- blasts are the rule in malignant disease. If megaloblasts are present they are in the minority, while in pernicious anaemia they are in the majority. The average size and staining power of the red cells is increased in most cases of pernicious anaemia and decreased in most cases of malignant disease. 3. Pernicious Ancemia and other Secondary Ancemias. Most secondary anaemias which are severe enough to reduce the count of red cells below 2,000,000 follow the type of malignant dis- ease and show leucocytosis. The great pallor and dyspnoea seen in connection with some cases of tuberculosis and nephritis rarely mean a low count of red cells, but simply a loss of haemoglobin. I remember two cases in adjacent beds at the Massachusetts General Hospital, both with extreme yellow pal- lor without emaciation; one had 1,020,000 and the other 4,100,- 000 red cells, the haemoglobin in each being about thirty per cent. The first was pernicious anaemia, the second nephritis. Purpura, typhoid, lead poisoning, chronic malaria, and other diseases may reduce the red cells to a point as low as that seen in early stages of pernicious anaemia and may not be accompa- nied by leucocytosis; but the absence of changes most charac- teristic of the latter disease (a majority of megaloblasts, in- creased diameter and color index in the red cells) serves to make the diagnosis clear. 1 4. Pernicious Ancemia and Leukcemia. Occasionally in in- fants these two diseases seem to approach very near each other and are difficult to distinguish. In infancy, as is well known, any anaemia (primary or secondary) is apt to be accompanied by 1 Another point of difference emphasized by Grawitz is that the plasma of pernicious anaemia has a relatively larger amount of solids than that of anaemia secondary to the above diseases. This is hardly a clinically ap- plicable test, but is said to be a valuable one. THE BLOOD IN PERNICIOUS ANEMIA. 149 leucocytosis and an enlarged spleen. Further leukaemia, which in adults usually causes a relatively slight anaemia, affects the red cells much more strongly in infancy, and may reduce them to a number decidedly suggestive of pernicious anaemia. There- fore in both diseases we may have enlarged spleen, great anae- mia, and leucocytosis. The one characteristic point of leukaemic blood the abun- dance of myelocytes usually enables us to distinguish the two diseases, for although present in both diseases the myelocyte is much more plentiful in leukaemia. Unfortunately we have no way of fixing just how numerous myelocytes must be in order to constitute leukaemia. It is only in infancy and very rarely then that this difficulty arises, but at that period I am inclined to believe that we sometimes see conditions intermedi- ate between the two diseases, indicating the ultimate identity of the two. Their numerous clinical resemblances cannot here be discussed. (For further comment on this point see page 395.) PROGNOSTIC VALUE or THE BLOOD IN PERNICIOUS ANAEMIA. The prognosis is always very bad, but the following scheme indicates the presence of a severe or of a mild type: 1. Severe (rapidly fatal). 2. Less Severe (slower course), (a) Extreme progressive an- (a) Remissions. asmia. (b) Normal or low-color index, (ft) High-color index. (c) Normal-sized or small cells. (c) Increase in size of red cells. (d) No degenerative change. (d) Degenerative changes. (e) Numerous normoblasts. (e) Numerous megaloblasts. (/) Few megaloblasts. (/) Few or no normoblasts. (g) Normal percentage of poly- (g) Lymphocytosis. morphonuclear cells. It has been thought by some observers that the absence or great scantiness of nucleated corpuscles indicated lack of any effort at regeneration on the part of the blood-making functions and hence a peculiarly malignant type of the disease. I have never seen cases in which no nucleated corpuscles were present, but their scantiness has seemed to me as a rule to be associated with a more slowly fatal type of the disease. No significance has seemed to me to attach to the presence of larger or smaller percentages of eosinophiles. 150 SPECIAL PATHOLOGY OF THE BLOOD. Pernicious anaemia. Chlorosis. Secondary anaemia. Leukaemia in infancy. Red cells White cells.... Haemoglobin . . Megaloblasts . . Normoblasts . . Size of red cells Lymphocytes.. Polymorphonu- clear cells. Myelocytes.... About 1,000,000.... Usually decreased. Often relatively high. Constitute the ma- jority of the nu- cleated red cells. L e s s numerous than the megalo- blasts. Increased Rarely under 2,- 000,000. Usually normal.... Always relatively low. Rare May be 1,000,000 or less. Usually in- creased. Relatively low... Rare; never more numerous than normoblasts. May be under 2,000,000. Usually more in- creased than in any other dis- ease. Relatively low. Common . Common. Various; not in- creased. Usually increased. Usually d i m i n - ished. Usually more numerous than in other diseases. Occasional; always more numerous than megalo- blasts. Diminished Various; not in- creased. Usually d i m i n - ished. Usually in- creased. Rare Decreased . . Decreased Common Bare To illustrate the different size of the cells in chlorosis and pernicious anaemia I have had photographs taken of the blood of a case of two of these diseases and of normal blood, all on pre- cisely the same scale. See Figs. 31, 32, 33. FIG. 31. Normal Blood. Magnified 350 diameters, THE BLOOD IN PERNICIOUS ANAEMIA. 151 Fis. 32. Pernicious Anaemia. Magnified 350 diameters. Note the relatively large size and well-stained centres of the cells. FIG. 33. Chlorosis. Magnified 350 diameters. Note small size and pale centres. 152 SPECIAL PATHOLOGY OF THE BLOOD. 2. THE BLOOD IN CHLOROSIS. This has been already described for the most part under the heading of Secondary Anaemia. In many cases the two are in- distinguishable by the blood examination alone, the changes consisting simply in the presence of light, small-sized, pale, more or less deformed red cells ^ whose number may or may not be decreased, according to the severity of the case. Leucocy- tosis is rarely if ever present in uncomplicated chlorosis, but is often absent in secondary anaemia. Normoblasts may be pres- ent in both. The chief points of distinction are : (a) The red cells are more apt to be uniformly undersized and under-colored in chlorosis, while in secondary anaemia we more often find normal cells among the diseased ones. (b) The color index may be lower in chlorosis than is com- mon in secondary anaemia, and this lowering is more constant in chlorosis. (c) Lymphocytosis, which is very common in chlorosis, is not so common in secondary anaemia. (d) Nucleated corpuscles are less common in chlorosis than in anaemia secondary to malignant disease. (e) Coagulation is rapid, in contrast with the very slow clot- ting of pernicious anaemia and of many secondary anaemias. Yet fibrin is not increased. The Mood in Gross. The pallor of the drop is sometimes excessive, fully as great as in pernicious anaemia, and the liquid is very fluid and thin. Yet it coagulates very rapidly and our technique must be prompt. BED CELLS AND HAEMOGLOBIN. Quantitative Changes. Hay em has recorded cases whose count was as low as 1,662,- 000 and even 937,360 per cubic millimetre. Such figures are certainly rare in this country, and the striking fact is usually the slight numerical loss of red cells, considering the extreme pallor of the patients. THE BLOOD IN CHLOROSIS. 153 The lowest count in the Massachusetts Hospital series was 1,932,000, and in W. S. Thayer's 63 cases 1,953,000. The accompanying tables, from the Massachusetts Hospital records, show the range of red cells and haemoglobin in 109 cases as counted when the patients first came under obser- vation. The highest counts (7,100,000 and 5,884,000) are un- doubtedly due to some temporary stasis or concentration of the blood. The average of the 109 cases, 4,112,000 red cells per cubic millimetre, is remarkable in so nearly coinciding with Thayer's ' series above referred to, the average of which is 4,096,544. The average haemoglobin percentage of this series, 41.2 per cent, is also very close to Thayer's (42.3 per cent). This gives us on the average a reduction of the corpuscle substance to one- half the normal, or to the equivalent of 2,250,000 healthy red cells; 61 of the 109 cases have 4,000,000 or more red cells. These figures do not agree with those collected by v. Limbeck, in which only 99 out of 247 are over 4,000,000. But this prob- ably means simply that in this country the patients seek medical advice before their disease has advanced very far, while in Ger- many they wait longer before resorting to a hospital. For, as above explained, in all anaemias the individual corpuscles suffer in quality first and only after some time begin to decline in number. This is especially the case in chlorosis, although by no means peculiar to that disease. The color index is invariably low, as seen in the table, al though it is rare to see it fall below .30. In only four cases of the present series did it go below that figure, the average being about .50. v. Noorden 2 found that the color index was especially apt to be low in first attacks and less often in the recurrent or habitual cases, but Eomberg 3 in a study of one hundred and seventeen cases has not found this true, and I agree with Eomberg. One of the lowest color indexes in my series was in a woman over fifty who had a truly habitual chlorosis. 1 See Osier's article on Chlorosis in the "American Text-Book of Medi- cine," vol. ii., 1894. 2 Chlorosis: Wien, 1897 (Holder). *Berl. klin. Woch., June 28th, 1897. 154 SPECIAL PATHOLOGY OF THE BLOOD. TABLE VI. CHLOROSIS. Red Cells. Cases. Between 7, 000, 000 and 8, 000, 000 . 6,000,000 " 7, 000, 000 1 5,000,000 " 6,000,000 17 4.000,000 " 5,000,000 42 3,000,000 " 4, 000, 000 33 2,000,000 " 3,000,000 14 1,000,000 " 2,000,000 i Average of these 109 cases = 4,112,000 White Cells. Between 15,000 and 14,000 Cases. 2 \ Between kVhite Cells. 7, 000 and 6, 000 ... Cast 19 " 14,000 " 13,000 1 6,000 " 5,000... 11 13,000 " 12,000 3 u 5,000 " 4,000... 9 12,000 " 11,000 5 tt 4 000 '' 3 000 7 11,000 " 10,000.. 10,000 " 9,000.. 9,000 " 8,000.. " 8,000 " 7,000.. ... 9 ... 6 ... 7 ... 23 it 3,000 " 2,000.. 2,000 " 1,000... 1 1 104 Average, 7,400. PER CENT OF HAEMOGLOBIN IN CHLOROSIS. Between 10 aud 19 = 7 cases. 20 " 29 = 13 " 30 " 39 = 28 " 40 " 49 = 25 " 50 " 59 = 24 " 60 " 69= 7 " 70 " 79= 1 " 105 Average, 41 per cent. The striking contrast is with pernicious anaemia, rather than with secondary anaemia. In the former the color index, as above mentioned, averaged 1.04 in 68 cases. In secondary anaemia it is almost always below 1, but does not average so low as in chlorosis, although in individual cases it may be very low. For example, Osterspey quotes a case of gastric cancer with a blood count of 4,230,000 red cells, and only 22 per cent of haemoglobin, a color index of .26. THE BLOOD IN CHLOROSIS. 155 BED CELLS (CONTINUED). Qualitative Changes. (a) The stained specimen shows a greater or less degree of pallor of the corpuscle centres corresponding so accurately to the diminution in haemoglobin that a practised observer can tell ap- proximately how low it is simply from the stained specimen. The pallor, however, is to be taken in connection with the size of the cells, for the diminution in haemoglobin is not due simply to a bleaching out of the cells, but to their loss of size. Hence, (b) The diminution in the average diameter of the cells is a very important feature. Both in this respect and as regards the bleaching of individual cells, many cases contrast with most secondary anaemias, in that a large proportion of the cells are affected alike, i.e, are small and pale, while in secondary anae- mia there are apt to be well-stained and good-sized or over- sized cells in every field. These last occur also in chlorosis, but less frequently as a rule. Hence the usually lower color index of chlorosis. In certain cases this distinction does not hold and the two conditions are identical in so far as the size and color of the red cells are concerned. It is to the white cells that we must look for help in differential diagnosis. (c) Deformities in size and slmpe are very common in all ad- vanced cases, but often absent in mild or moderate ones. They present no special peculiarities except that macrocytes are rela- tively rare and microcytes relatively common. In the severest cases, however, the macrocytes begin to get more numerous and we approach the picture of pernicious anaamia. (d) Degenerative changes (Maragliano, see page 88) are not common but are occasionally present in severe cases. (e) Nucleated red corpuscles are very scanty even in advanced cases. Hay em never saw any, but most observers find them in small numbers after long search. They are almost always of the normoblast type, but megaloblasts have also been found. The scantiness of nucleated red cells is a point of contrast with the anaemia secondary to malignant disease, in which even in mildly anaemic states we readily find nucleated corpuscles, while in chlorosis, even in severe cases, a long search may show very few or even none at all. 156 SPECIAL PATHOLOGY OF THE BLOOD. Specific Gravity. Chlorosis is usually agreed to be one of the diseases in which specific gravity and haemoglobin run parallel, and as the inac- curacies and inconveniences of the v. Fleischl instrument are so great, it seems to the writer better to follow the specific gravity rather than the haemoglobin. The tables on page 42 (Part I. ) show how the inference from density to coloring matter can be made. A specific gravity of 1030 is not very rare. WHITE CELLS. A. Quantitative Changes. Leucocy tosis is absent in uncomplicated cases. In the series in Table YI. the occasional leucocy tosis may be due to digestive or to a variety of other influences (uterine troubles, etc.), which could not be excluded. The average in Thayer's 63 cases was 8,467; in the present series (see Table YI.) it is 7,485. As in pernicious anaemia, the worst cases are apt to have leucopenia, and as improvement progresses the white rise even faster than the red corpuscles. Thus in Romberg's careful study of 117 cases, 24 cases whose haemoglobin was under 40 per cent had an average of 6,350 leucocytes per cubic millimetre, while 52 cases whose haemoglobin averaged 60 per cent had an average of 9,250 leucocytes. He found the average in healthy girls of the same age 9,068 white cells per cubic millimetre. The absence of leucocytosis is the most important point in distinguishing chlorosis from secondary anaemia due to cancer, suppuration, etc. B. Qualitative Changes. Lymphocyte-sis is usually present, as in pernicious anaemia, wherever the disease is well marked, and sometimes even in mild cases. Thus Eieder found in 12 cases an average of 33 per cent of lymphocytes, the highest percentages being 53.7, 43.5, and 41.7. Either the small or the large lymphocytes may pre- dominate. In my own experience it has usually been the small forms. THE BLOOD IN CHLOROSIS. 157 The neutrophiles suffer proportionally, their low percentage contrasting often with that of secondary anaemia associated with leucocytosis. Eosinophiles are occasionally increased. In Eieder's 12 cases the average percentage was 3.5, the highest percentages being 9.6 and 7 per cent. Myelocytes are rare but have occasionally been observed in small numbers. Regeneration of the Blood. As the patients begin to mend under the influence of treat- ment, the blood changes are just the reverse of those seen dur- ing the development of the disease. First the corpuscles gain in numbers, the haemoglobin still remaining low; later and much more slowly the coloring matter, size, and weight of the cells are renewed. It seems as if the new-formed cells were of light weight and had to be replaced gradually by cells of normal stature. The nucleated corpuscles and deformities disappear and the leucocytes shoot up often a little above the normal. Slood Plates. "Usually considerably increased. Chlorosis without Known Blood Changes. Eomberg quotes the following facts : Three girls, nineteen, twenty, and twenty-five years of age, came to him with typical symptoms of chlorosis. Their blood counts showed : I. Eed cells, 5,246,000; Hb., 80 per cent. II. " " 5,376,000; " 83 III. " " 4,408,000; " 87 All improved markedly under iron treatment. I mention this because I have seen several similar cases and have heard of others from colleagues. Summary. 1. Blood as a whole : Very pale in marked cases, very fluid, but coagulates rapidly. Fibrin not increased. Specific gravity usually low, running parallel with the haemoglobin. 2. Ked cells : Average 4,000,000 when patient is first seen, 158 SPECIAL PATHOLOGY OF THE BLOOD. very rarely go below 1,000,000. The majority of them are small-sized, pale, often deformed. Nucleated corpuscles are rare (normoblasts as a rule). 3. White cells, not increased. Lymphocytosis, occasionally eosinophilia. 4. Blood plates increased. Diagnostic Value. 1. The points of difference from pernicious anaemia have been discussed. 2. It is important to distinguish it from simple debility, and from cases whose skin only is anaemic; in both of these conditions the blood is normal. 3. From secondary anaemia it may be indistinguishable in case the latter be without leucocytosis. Where leucocytosis is constantly present and the percentage of polymorphonuclear leucocytes is increased, chlorosis (uncomplicated) can be ex- cluded. Of course many of the complications which may occur in chlorosis are accompanied by leucocytosis. Other Forms of Ancemia. 1. Cases of acute fatal anaemia following purpura or severe hemorrhage of any kind are sometimes classed as pernicious anaemia. The blood certainly differs in many respects from that of ordinary pernicious anaemia. Ehrlich describes such a case following metrorrhagia in which the red cells were re- duced to 213,360 per cubic millimetre without much deformity of individual cells and with a decided decrease in the average diameter. Poly chroma tophilic forms numerous. No nucleated corpuscles whatever could be found even after many hours' search. The leucocytes were decreased to about 200 (!) per cubic millimetre. Eighty per cent of them were small lympho- cytes, six per cent large lymphocytes, and only fourteen per cent polymorphonuclear. Eosinophiles and myelocytes ab- sent. Autopsy showed no red marrow in the long bones except at the epiphysis, and an entire lack of effort on regeneration. W. S. Thayer has observed two similar cases, and Bignami and Dionisi have seen cases of the same kind following the THE BLOOD IN CHLOROSIS. 159 prolonged deglobularizing action of the malarial organisms. I have never seen just this type of anaemia. 2. Some cases of severe chronic chlorosis seem to belong in a separate category. Like the variety last described the blood is here of the microcyte type, the diameter of the corpuscles being greatly reduced. I have observed one case somewhat similar in a male of fifty-two, a carpenter whom I had noticed for several years at his work, the palest man I ever saw out of bed. About two years ago I had an opportunity to examine his blood, he feeling well, at work, and objecting to the bother of the FIG. 34. Chronic Secondary Anaemia due to Bleeding Piles. Magnified 350 diameters. Note the similarity to chlorotic blood (Fig. 33, page 151). examination. His blood showed: Eed cells, 2,530,000; white cells, 2,000 ; haemoglobin, 32 per cent. The red corpuscles were not at all deformed but were very small (see Fig. 34) and very pale in the centre. While counting 1,200 white cells 17 normoblasts were seen; no megaloblasts. The leucocytes showed : Polymorphonuclear neutrophiles, 54.5 ; small lymphocytes, 35 ; large lymphocytes, 9.4; eosinophiles, 1.1. The man is still well and hearty, complains of nothing, but is as pale as ever. A few days ago he disclosed to me that he had had bleeding piles for ten years. He had been hitherto concealing this fact. CHAPTEE II. I. LEUKAEMIA. THE distinction between leukaemia and leucocytosis has been sufficiently dwelt on above. The blood of the vast majority of oases of leukaemia falls clearly under one or the other of two distinct types, myelocytwm.ia on the one hand, lymphcemia on the other. Myelocytaemia is only found in cases with great hypertrophy of the spleen, marked marrow changes, and little or no enlargement of the other lym- phatic tissue. Such cases are usually chronic (two to five years). Lymphaemia, on the other hand, may be associated either with acute or chronic forms of the disease, and while in all cases of lymphaemia we have some set of lymphatic glands enlarged there may be no externally visible glands enlarged, and the spleen may be as big as in cases associated with myelocytaemia. The diagnosis of leukaemia can easily be made by the blood alone, but we cannot say from the blood whether or not the spleen or visible lymph glands are the organs chiefly involved. In acute cases the lymph glands of the alimentary tract (cer- vical, faucial, gastro-enteric, mesenteric) may be the only set involved. All of the thirty-one cases associated with myelocytaemia which have come under my observation have run a chronic course, while of the cases showing lymphaemia five were chronic, three acute, and two subacute. All showing myelo- cytaemia had very large spleens without enlargement of visible lymph glands, but two of the lymphaemias had spleens almost filling the abdomen. The disease leukaemia, then, is associated with three types of blood. 1. Chronic myelocytaemia. 2. Chronic lymphaemia. 3. Acute lymphaemia. LEUKEMIA. 161 1. MYELOCYT^EMIA. (Splenic-myelogenous leulccemia. ) The drop as it emerges from the puncture looks somewhat opaque in color, but is neither whitish nor chocolate colored. It flows very sluggishly, however, and is difficult to spread between cover-glasses owing to the masses of white cells con- tained in it. Coagulation is slow. KED CELLS. In early stages of the disease there is no anaemia. Later the diminution in red cells is moderate, averaging about 3,120,000 in the thirty -nine cases of Table VII., A. The patients are often not pale and may feel perfectly well. The haemoglobin is usually diminished, the color index being about 0.6 in my cases. It is difficult to read the v. Fleischl instrument in leu- kaemia, as the presence of so many leucocytes gives a muddy tint to the liquid, not easy to compare with the red of the glass. TABLE VII. LEUKAEMIA. A L B No. Red cells. No. White cells. 1.. 2 .. Highest . . . 5,000,000 4 877 000 1 2 Highest . . 1,072,222 980 000 3 4 800 000 3 820 000 4.... 4, 592, 000 '4 ... 800 000 5 4, 288, 000 5 756 000 6.... 4,016,000 6 748 000 7.... 3,760,000 7 716 000 8 3 635 570 8 656 000 9.... 3 605 000 9 626 600 10.... 3 400 000 10 570 000 11.... 3,292,000 11.... 500 000 12 ... 3 200 000 12 492 000 13.... 3,080,000 13 454 000 14.... 3 078 000 14 448 000 15.... 3 010 000 15 430 000 16.... 2 996 000 16 428 000 17.... 2 960 000 17 405 000 18.... 2, 938, 000 18 400 000 19.... 2 921 600 19 394 000 20.. 2 868 000 20 386 000 21.... 2,792,000 21.. 340,000 11 162 SPECIAL PATHOLOGY OF THE BLOOD. TABLE VII. LEUKAEMIA (Continued). A B No. Red cells. No. White cells. 22 2, 738, 000 2,715,000 2,576,000 2, 520, 000 2, 322, 222 2,320,000 2,256,000 2,140,000 2,112,000 2,060,000 2,016,000 2,010,000 1,866,664 1,420,000 1,386,000 1,358,000 1,200,000 408, 000 3 = 3,131,000 + 22 320,000 290,000 260, 000 220, 500 213,000 188,000 183, 000 175, 800 170,000 139, 600 138,000 134,400 132,000 111,000 98, 000 age = 428, 000 23. .. 24.... 25 23 24.... 25.... 26 26 27 27.... 28 28.... 29 ... 29 . 30.... 30 .. 31 31 32 32 33 33.... 34.... 35 34.... 35.... 36 36. ... Lowest . . . Aver 37.... 38 39. ... Lowest .... Averag Qualitative Changes. The striking point is the presence of very numerous nucleated red cells even in the absence of any sign of ancemia. With over 4,000,000 well-formed and well-colored red cells, we may have hundreds of nucleated ones in every cover-glass. They are as numerous in this form of leukaemia as in the worst forms of pernicious anaemia, even tho.ugh the patient may be feeling nearly well. Both normoblasts and megaloblasts may be seen, but in most cases the latter are in the minority. Many of the normo- blasts show fragmentation in their nuclei, and occasionally true karyokinetic figures are to be seen. In the anaemic cases we find all the other changes in the red cells characteristic of anaemia, but the nucleated cells are always more prominent than in any other form of anaemia of a like severity. This shows that nucleated corpuscles are not to be thought of as evidence (like deformities in shape) of regenerative or degener- ative conditions only. A special connection to the bone marrow is very clearly indicated, all the more so as in the lymphatic PLATE II FIG. 1. Both this and Fig. 2 are intended to be fac-similes of actual microscopic fields. (a) Note the cell between those labelled 8 and 9 apparently a " mast cell." Such cells are often seen in this form of leukaemia. With Ehrlich's stain they present this appearance. Basic stains bring out coarse blue granules on the periphery of the protoplasm. (5) Note also the cell at the extreme upper right-hand corner of Fig. 1, which it is almost impossible to classify either as a myelocyte or as a polymorphonuclear neutrophile, since it appears to be intermediate be- tween the two varieties. (c) Both the nucleated corpuscles are normoblast *, 9 has polychromato- philic protoplasm. The red cells show scarcely any deformities and very slight deficiency in coloring matter. F IQi 2. (a) Note the deformities in size and shape of red corpuscles, owing to the anaemia present. (6) No lymphocytes are figured, as they made up only two per cent of the white cells in this case. Eosinophiles were absent. (c) Note that the contrast between this figure (leucocytosis) and the one above it (leukaemia) is not in the abundance of white cells but in the kind of white cell predominating among those present. Examination of the Blood. PLATE II. Figure I = Splenic-myelogenous Leucaemia Figure II = Leucocytosis (cancer of kidney) Cells stained yellow = Red corpuscles 1. 2. 3. 4 a. 5 = Polymorphonuclear neutrophiles 6 = Lymphocyte 7 a. 8. = Eosinophiles 9 a, 10 = Nucleated red corpuscles All others = Myelocytes ~,* Figure I Leucaemia. Cells stained yellow = Bed corpuscles All others = Polymorpho- nuclear- neutrophiles Figure II Leucocytosis. Scale of R. C. Cabot fee. Lith. Anat. T. E. A. Kunk, Leipzig. LEUKAEMIA. 163 form of the disease in which the bone marrow is usually much less affected, nucleated corpuscles are much less numerous, ap- pearing in relatively small numbers in the very acute anaemic cases and not at all in those who are not anaemic. Other qualitative changes are not marked and correspond to the degree of anaemia present ; often there are none at all. As the count of the white cells rises, that of the red may fall, and vice versa ; or the red cells may remain at a comparatively high figure despite the progress of the white. WHITE CELLS. Quantitative Changes. The average number per cubic millimetre in the thirty-six cases of Table VII., B (the lymphatic cases being excluded), was 438,000 at the time when the cases first came under obser- vation. The highest count in this series is 1,072,222 and the lowest 98,000. Cases are on record in which the white cells were actually more numerous than the red. The average ratio in my series is about one white to seven red. The highest ratio is 1 : 2, and the lowest 1 : 37. It is best to use the " red counter" with a dilution of 1 : 200 in counting the white cells, otherwise they are often too crowded for convenience. The hsematokrit is useful in this disease and in any condition where the white cells are much increased, not to supersede the Thoma-Zeiss or to give us the absolute number of cells, but for comparative obser- vations as to the length of the column of white cells from day to day in a given case. Hayek ' has shown that the count of leu- cocytes may vary enormously in a very few hours ; e.g., 10 A.M., 122,500; 4 P.M., 235,000; or again, 10 A.M., 730,000; 4 P.M., 547,500. In the fresh specimen we notice that a large proportion of the white cells are not amoeboid, a point of marked contrast with leucocytosis, in which nearly all the leucocytes are amoe- boid. This is due to the fact that the myelocytes which form so large a portion of the leucocytes in this disease possess little if any faculty of amoeboid motion. We should expect there- fore to find their nuclei free from the twists and distortions 1 Hayek: Wien. klin. Woch., 1897, No. 20. 164 SPECIAL PATHOLOGY OF THE BLOOD. characteristic of the amoeboid (polymorphonuclear) cells. And this is in fact the case (see below) . TABLE VIII. MYELOCYTJEMIA (SPLENIC-MYELOGENOUS LEUKAEMIA). 6 fc 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 g 1 2,010,000 1,720,000 4,125,000 4,016,000 4,592,000 2,960,000 3,184,000 3,156,000 3,400,000 3,670,000 3,100,000 3,080,000 2,520,000 4,016,000 2.792,000 2,715,000 2,256,000 4,288,000 2,921,600 3,010,000 2,996,000 4,800,000 3,060,000 2,016,000 2,576,000 2,448,000 2,120,000 2,528,000 5,120,000 5,000,000 4,800,000 I i -2 716,000 732,000 708,000 253,900 200,000 646,000 448,000 175,800 264,000 276,000 111,000 183,090 430,000 405,000 510,000 528,000 570,000 560,000 800,000 26,000 139,600 394,000 340,000 213,000 492,800 188,000 134,400 220,500 274,000 260,000 748,000 168,800 190,000 188,600 159,000 134,000 137,800 138,000 PERCENTAGE. Normoblasts seen while counting them. Megaloblasts seen while counting them. Date. d M W 1 Polymorphonuclear neutrophiles. 1-1 Small lymphocytes. cc 1 1 1 j>> 10 1 1 4 1 | >> g 42 1 1 CO 1 1 "fl 1 1 3.8 Many Many Many Nov. 1st, 1897.1 Nov. 3d, 1897. Nov. 12th. 1897. July 16th, 1895. July 25th, 1895. Jan. 21st. 1897. Feb. 8th, 1897. Feb. 17th, 1897. July 15th, 1897. July 29th, 1897. Aug. 5th, 1897. Aug. 10th, 1896. Aug. 31st, 1896. Jan. 22d, 1896. Jan. 22d, 1896.2 Jan. 23d, 1896. Jan. 24th, 1896. June 4th, 1894. Aug. 10th, 1894. April, 1893. Later the count of leucocytes was normal for sev- eral months. Jan. 22d. 1896. June, 1897. Feb. 22d, 1896. Feb. 25th, 1896. Feb. 28th, 1896. 55 60 42 38 in 35 48 40 50 56.5 46 46.6 17 7 1.5 1.4 3 4.5 14.5 10.4 11 4 1.5 2.2 19 28 33.5 36.7 49 Many Many 3 2 3 13 1 31 27.2 31 10.5 15 26 8 3 5 25 32 Many Many 72.2 51 2 .6 3.6 2 2 4 17.4 42.4 Many Many 5 9 58 55 5 4 4 34 Many Many 87 41 42 45 78 50.4 44 62.3 53.8 61 37 46 62 33 26 46 45 74.2 18.9 1.5 3.8 2.2 3 23 1.5 1.5 10 8 2 3 4 0.2 1 1.8 2.8 6 0.5 1 11.5 1 0.5 3 4 6.1 2.5 1.8 14.4 3 8 4 2.5 3 5 1.5 3 2.8 24.4 51 30.3 18 33 26 48 33 42.5 60 50 46 15 8.8 4 1 ;; 61 32 49.6 30 54 6 4 3.6 10 8 24.5 4 5 6 2.2 4 2.2 5 5 4 1.5 3 2.5 5 6 5 5 1.5 28 3.7 1 28 55 38.6 45 34 27.5 28 36 47 10 1.5 2 36 57.5 42 * Av. 3,120,000 348,000 52 46 4.6 5.1 35 5 1 Many cells on border-line between large lymphocytes and myelocytes and between these and polymorphonuclear neurophiles. 2 Cerebral hemorrhage. Death January 25th, 1896. LEUKEMIA. 165 With or without the influence of therapeutic agencies the white cells may fall gradually to normal and remain there for some time, the patient feeling greatly improved. Such a case occurred under my observation, and the patient, a washer- woman, went back to work and afterward passed through an attack of lobar pneumonia in safety. At such a time, when no increase in the white cells is pres- ent, we should never suspect leukaemia, seeing the case for the first time, unless we chance to make a differential count; then the characteristic qualitative changes (see below) would be seen. QUALITATIVE CHANGES. 1. Myelocytes. The enormous number of myelocytes is the chief point of interest. The average in my 28 cases was 35 per cent (see Table VIII.), rising in one case as high as 60 per cent and only twice falling lower than 20 per cent. Taking the average total number of leucocytes as 428,000 per cubic millimetre, the absolute number of myelocytes would be over 150,000 per cubic millimetre. So far as I am aware the highest count of myelocytes in any other disease is that men- tioned on page 306 in a case of malignant disease, namely, 4,514 per cubic millimetre. The contrast is sufficiently striking. I wish to insist upon this point, namely, that the blood of splenic- myelogenous leukaemia is absolutely peculiar and characteristic, and could not be confused with that of any other disease. Cer- tain writers of late years have concluded that because myelo- cytes do occur in a great variety of diseases as well as in leu- kaemia, therefore there is nothing peculiar about the blood of the latter affection. It would be as logical to say that because albumin and casts occur occasionally in the urine of persons practically well, therefore there is nothing characteristic about the urine of acute nephritis. Between the largest number of myelocytes ever recorded in any disease other than leukaemia, and the smallest number ever found in the latter disease, there is as great a difference as there is between the minute traces of sugar to be found in normal urine and the marked glycosuria of diabetes mellitus. 166 SPECIAL PATHOLOGY OF THE BLOOD. At the first glance the stained specimen of leuksemic blood seems to be composed mostly of myelocytes, but this is because they are on the average so much larger than the other forms of white cells, which, being packed away in the interstices between the large myelocytes, do not appear prominently at first sight. Although (as just mentioned) the average size of the mye- locytes is greater than that of any other kind of leucocyte, there is a great range of variation in their size, and some are hardly, if at all, larger than a red cell. (This is equally true of the mye- locytes as seen in the bone marrow. See above, page 71 ,) The individual characteristics and variations in the myelo- cytes have been already sufficiently described on page 69. 2. Polymorphonuclear Cells. Although absolutely the number of these cells is greatly in- creased, the number in each 1,000 leucocytes is considerably diminished. The average percentage in the 28 cases of Table VIII. is 46, the figures ranging between 17 and 72 per cent. The individual cells show a much greater range of variation in size, staining properties, and the size and shape of the nucleus than in any other condition. In most forms of leu- cocytosis, for example, one adult cell looks very much like another, but in this form of leukaemia we are often struck by (a) Very small cells or very large cells. (b) Dark stained or very pale stained cells. (c) Unusual shapes in the nuclei. Besides these variations we often see cells apparently be- longing to this type, but whose protoplasm shows no color what- ever. Such a cell is figured to the right of Plate II., Fig. 1. Other cells show a few granules scattered about against a per- fectly white background. The outer rim of the cell is usually stained faintly, so that we can hardly make out its outline. Such cells usually contain basophile granules beside tlie neutrophile. (d) There are always some cells on the border-line between the polymorphonuclear and the myelocyte, and in regard to which decision must be arbitrary. We cannot help getting the LEUKEMIA. 167 impression that at any rate in this disease the two varieties are only different stages in the development of the same cell. (e) More than one kind of granule is sometimes seen in the protoplasm, i.e., eosinophilic or basophilic as well as neutro- philic. 5. Lymphocytes. It is here that the greater relative diminution occurs, to make room for the incursion of the myelocytes. In percentages they are reduced from their normal, 20 to 30 per cent, to an average of 10.6 per cent, as in leucocytosis. But still their absolute number is always increased. Thus the lowest percentage present in Table VIII. (namely, two percent) would mean 8,760 out of the average 438,000, the total leucocyte count per cubic millimetre, and 8,760 is three or four times as many lym- phocytes per cubic millimetre as are present in normal blood. The proportion of large and small forms among the lym- phocytes varies a great deal. Sometimes the lymphocytes of this form of leukaemia do not differ from those of normal blood, but in most cases we find one or more of the following atypical varieties : (a) Lymphocytes with a protoplasm so darkly stained that it is difficult to distinguish them from myelocytes. Indeed in some cases where hints of a granular look appear in the violet- stained rini we find it impossible to be sure whether we are deal- ing with a large lymphocyte or a myelocyte. The personal equation alone decides. (b) Cells like lymphocytes except that they contain from three to ten widely separated granules of one or more varieties (basophilic, acidophilic, or neutrophilic). 4. Eosinophiles. Like all the other varieties these are absolutely much in- creased. Relatively by percentages they may or may not be so. In my series they ranged from 1.5 to 28 per cent, averag- ing 5.1 per cent, a slight increase over the normal. Many writers, wrongly interpreting Ehrlich's observations on this point, have stated that an increased percentage of eosino- philic cells was the distinguishing mark of leukaemia, and even 168 SPECIAL PATHOLOGY OF THE BLOOD. recent writers (e.g., Gilbert, Strumpell) continue to repeat this false statement. The cell characteristic of splenic-myelogenous leukcemia is not the eosinophile but the myelocyte. We distinguish several types of eosinophiles in leuksemic blood. (a) Ordinary (polymorphonuclear) eosinophiles. (b) Eosinophilic dwarf cells. (c) Eosinophilic myelocytes. (a) Needs no comment; (b) is simply a very small cell with eosinophilic granules ; sometimes such cells are not over 5 P- in diameter. They are not uncommon in this form of leukaemia and are very rare in any other disease. The same is true of (c), the eosinophilic myelocytes which are very rare in any other disease, except pernicious anaemia, where they are occasionally seen. These cells are like myelocytes except that their granules are eosinophilic instead of neutrophilic (see Plate I. and Plate II.). They are found in the marrow in considerable numbers and constitute the majority of the eosinophilic cells seen in this form of leukaemia. Occasionally we see eosinophiles with a few basophilic or neutrophilic granules as well. 5. Basopliiles. (a) The lymphocytes may contain basophilic granules as in any ordinary blood. (b) Certain of the myelocytes contain fine basophilic granules in addition to their usual neutrophilic granules. (c) " Mastcellen" or coarsely granular basophiles, usually with a trilobed nucleus, are almost always seen in specimens stained with dahlia or methylene blue. With the triple stain their protoplasm is nearly unstained, but usually a number of round white spots can be made out against a faintly stained background. These are the basophilic granulations. Mast- cells make up from one to ten per cent of the leucocytes in most cases of myelocytaemia. LEUKAEMIA. 169 6. Polymorphous Condition of the Blood. Weiss has rightly insisted on the fact that in this type of leukaemia the blood preparations show a very polymorphous condition, that is, there are no fixed types, but every variety shades through intermediate forms into some other variety. No two cells are alike. Precisely the same conditions obtain in the normal marrow, and we can scarcely resist the impression that in this form of leukaemia we see in the blood unfinished cells of various kinds which usually do not appear in the circu- lating blood. As Charcot-Leyden crystals have no diagnostic value and are not peculiar to any disease, no description of them will be given here. They appear to be present wherever eosinophiles are plentiful, e.g., in asthma, gonorrhoea, in the bone marrow, etc. Kanthack considers that a diminution in eosinophiles (pro- gressive) is a bad prognostic si^n. During remissions, when the leucocyte count may fall to normal, the percentage of myelocytes remains large and the diagnosis could usually be made even if we saw the case then for the first time. This I have observed in two cases, and Thayer has had the same ex- perience. II. LYMPH^EMIA. (Lymphatic Leukaemia.) Although Fraenkel once maintained that all cases of lym- phatic leukaemia are acute, and that therefore the difference be- tween the various forms of the disease rests simply on the rapidity of the process in the blood and clinically, there is no doubt that chronic lymphatic leukaemia exists. Fraenkel was enabled to maintain his position only by ex- tending the term acute to cover all cases whose symptoms last not more than four months. Six weeks is the limit agreed upon by most other observers. The writer has watched five cases of typical lymphatic leu- kaemia for periods of from seven months to two years. One was as little sick as any case of leukaemia that I have ever seen, and came over thirty miles from time to time to report at the Out- Patieut department. His blood showed little variation from the following figures: Ked cells, 2,300,886; white cells, 112,000. 170 SPECIAL PATHOLOGY OF THE BLOOD. The differential count always showed the overwhelming ma- jority (over ninety per cent) of small lymphocytes characteristic of the disease. The lymph glands were all much enlarged, the spleen just palpable. The patient kept about his work as a gardener for over two years. Grawitz has watched a similar case for over four years. The blood of acute lymphaemia differs as a rule in many respects from that of chronic cases. These differences will be referred to later on. RED CELLS. The count of red cells is often somewhat lower than in the splenic-myelogenous form of the disease, averaging 2,730,000 in my cases. In acute cases it is usually very low and the an- aemia progresses rapidly. In chronic cases the red cells behave about as in myelocythaemia, except as regards nucleated forms. Here the point of interest is the comparative rarity of nu- cleated red cells, the abundance of which is so marked a feature of splenic-myelogenous leukaemia. They follow the grade of anaemia present. Cases occurring in children show more abun- dant nucleated corpuscles (the same is true of all leukaemia in children) than those occurring in adults', and the megaloblasts, usually scanty, may equal the number of normoblasts. In very acute cases the number of nucleated forms is greater and may be as great as in myelocytaemia. Two cases recently reported by Herrick x exemplify this. WHITE CELLS. Quantitative Changes . As a rule the numerical increase is not nearly so marked as in the splenic-myelogenous form. The average ratio of white to red cells is about 1 : 50 instead of 1:7, and we rarely see counts reach the height common in the other form of the disease. The highest count of my series was 1,480,000 at the patient's first visit, and the lowest 30,000, the average being 141,000 as compared with 438,000 in the other form. These figures refer to uncomplicated cases. 1 Journal of the American Medical Association, July 24th, 1897. PLATE III. (a) Chronic Lymphcemia with Excess of Small Lymphocytes. One polymorphonuclear cell is present. All the rest are lymphocytes and exemplify the variations in the morphology of the cell occurring in this and other diseases as well as in health, e.g., variations in the stain- ing of the protoplasm and nucleus, indentation and even division of the nucleus. Note that the scale of the whole of Plate III. is larger than in the other plates (see scale of p.) . (b) Acute Lymphcemia with Excess of Large Lymphocytes. Note the lack of chromatin in both nuclei and protoplasm of large lymphocytes. The plasma around them or their extreme edge took most of the stain. The brown tint of the red cells is due to underheating. Compare the colors with those in the figure above (a) in which the prep- aration was properly heated. Examination of the Blood. Lymphatic Leucaemia a. Small Lymphocytes in excess b. Large Scale of |A R. C. Cabot fee. Lith. Anst. v. B. \. Kunke, Leipzig LYMPHATIC LEUKAEMIA. 171 Qualitative Changes. 1. Lymphocytes (small forms, large forms, or a mixture) make up usually over ninety per cent of all the leucocytes present. In some cases they are all nearly of one size, while in others we find every gradation from the smallest to the largest, so that it is absolute^ futile to attempt to separate them into " large" and "small." Four of my cases were made up wholly of the small forms all under 10 // in diameter, two were composed largely of forms over 15 v- in diameter, while six showed every inter- mediate size. TABLE IX. LYMPHATIC LEUKEMIA. No. Red cells. White cells. Per cent haemoglobin. Small lymphocytes. | t &f 1 Polynuclear neutrophiles. Eosinophiles. Myelocytes. Normoblasts. Megaloblasts. | Remarks. 1 2 4,877,000 91 000 132,000 23 000 17 75.8 15 16. 82 4.6 2 4 1.6 1 2. 5 4 1 Subacute ; ten weeks. Jan 24th 1896 1 440 000 43,600 ?S Jan. 26th, 1896 1,336,000 1 100 000 92,000 120000 20 25.4 73.2 .5 .1 .8 3 1 Acute ; two weeks. Jan. 27th. Jan. 28th. Death- 3 3,000,000 31,600 685 28. 3.2 autopsy. April 3d, 1896. 3 500 000 31 500 55 78 156 56 fi April 5th, 1896 3,608,000 28,500 40000 55 95.3 4.3 .3 ... April 6th, 1896. April 7th, 1896 4 700 000 31,500 40 000 95.5 * 4. .5 2 2 Acute: five weeks. April 8th, 1896. April 12th 1896 3 100 000 3 400 39 52 9 April 22d (sepsis 800 94 7 * 53 semicomatose). \pril 29th Death 4 2,960,000 1,480,000 87.9 12. .1 IB March 21st 1897 5... 4 160 000 80000 80 5 2 1 172 Chronic. Oct 26th 1896 2,768,000 77,500 51,800 50 88.7 1.6 9.4 .2 .1 Chronic. Nov. 5th, 1896. Nov. 7th 1896 90 4 1 4 g 2 Nov 15th J896 79,500 Nov 17th Died 6... 3 520,000 64,000 60 94 6 December, 1897. 7 94 # 5 7 87 8 9 97.9 862 7. 1.4 9 8 .8 2 Acute. 10 72 28 11... 99 6 '* ' 4 Few History unknown 12... 92 2 * 78 13 700,000 600,000 10 998 * 8 3 2,500 cells counted. Average. 2,900,000+ 59,000+ 40 80.2 15. 4.2 .2 .4 * Large and small forms counted together on account of the impos- sibility of differentiating them in these cases. 172 SPECIAL PATHOLOGY OF THE BLOOD. In acute cases, where the large cells usually predominate, the staining is often very faint throughout the nucleus and proto- plasm (see Plate III., 6), so that at first sight we should think something was wrong with our technique. Other forms of leu- cocytes in the same preparation, however, will stain normally, showing that the trouble is in the lymphocytes and not in the technique. These large lymphocytes are identical in their ap- pearance with those found at the " germ centres" of all adenoid tissue, and probably are the mother cells of the small lym- phocytes. Benda has termed them "lymphogonien." They have often been mistaken for myelocytes, from which they are to be distinguished by the absence of any neutrophile granula- tion. They often show evidences of degeneration (see above, page 63). The protoplasm may be entirely unstained as in most of the cells in Plate III., b, or it may stain pale gray or pink. In other specimens, especially those of the small- cell type (Plate III., a) the lymphocytes stain well. Their nuclei are frequently indented or even divided in two (this occurs also in normal blood, but less often) . Fraenkel believes still that lymphaemia, though not always acute, is usually so, and that if a chronic case takes on acute symptoms the blood becomes more lymphsemic, while if a case starts acute and becomes chronic the lymphocytes decrease. Thus in a case reported by v. der Wey 1 a chronic myelo- 'cytsemia six weeks before death began to have fever, hemor- rhages, great increase in the total leucocyte count and in the ansemia. No complication. The lymphocytes increased 30 per cent, and the polymorphonuclear neutrophiles dropped from 30 to 3 per cent. Gerhardt 2 watched a case which began acutely with a large percentage of large lymphocytes and then became chronic with a predominance of small lymphocytes. In acute cases Litten 3 has noticed fatty degeneration in the leucocytes. The following figures illustrate the influence of a septicaemia (from suppurating cervical glands) which ended the life of No. 3 in the above Table X. 1 Deut. Arch, f . klin. Med. , vol. 57. 2 15th Cong. f. innere Med., 1897. s llth Cong, f. innere Med., 1893. LYMPHATIC LEUKEMIA. 173 Date. Number of leucocytes. Percentage of lymphocytes. April 3d ... 31,600 96.5 4th 31,000 6th 28,505 93.6 8th 44,000 10th 31,500 95.5 12th 40,000 13th Sepsis began. 20th ... 5,661 21st 4,000 22rl 3,400 92. 24th . . 3,222 28th 800 " 29th 471 94.7 Death on the 29th. Zeissl's case, also of the lymphatic form, showed the follow- ing: Date. White cells. Percentage of lymphocytes. Percentage of adult cells. September 9th . 80,000 96. 4. 24th 113,000 26th 119,000 " 29th 122 000 97 8 2 October 6th . . 140 000 9th Pneumonia began 99. 1 10th 119 000 llth 98 000 12th 68,500 13th 43,500 88.7 11.3 14th 50,000 15th 9,350 85.4 14.6 16th (A.M.) 16th (P.M.) 133,200 172,000 75. 25. Polymorphonuclear neutrophiles are often so scarce that one has to look through several thousand leucocytes before finding one. There is nothing abnormal about them. Eosinopliiles and myelocytes are equally rare. Summary. The leading characteristics of leuksemic blood are as follows : (a) Myelocytcemia. 1. Red cells about 3,000,000, nucleated forms very numerous. 2. White cells about 450,000, of which 3. Myelocytes form about thirty per cent. 174 SPECIAL PATHOLOGY OF THE BLOOD. 4. Every possible form of cell intermediate between the ordi- nary varieties is to be seen. ("Polymorphous blood.") (b) Chronic Lymplicemia. 1. Red cells about 3,000,000 or lower; nucleated forms rare. 2. White cells about 100,000 or lower, of which 3. Small lymphocytes usually form over ninety per cent. 4. Myelocytes and eosinophiles very scanty. (c) Acute Lymphcemia. 1. Red cells much diminished; nucleated forms not infre- quent. 2. Large forms of lymphocytes usually predominate; many of them often show signs of degeneration. 3. Neutrophiles and eosinophiles very scanty. Diagnostic Value. Leukaemia is distinguished by the blood examination from 1. Hodgkin's disease: (a) splenic, (b) glandular. 2. Tumors of the spleen and vicinity (e.g., kidney or retro- peritoneal glands). 3. Enlargements of the lymphatic glands from tuberculosis, syphilis, malignant disease. 4. Hydronephrosis. 5. Huge leucocytosis from any cause. 6. Chronic malaria. 7. Amyloid disease. 1. Leickcemm and Hodgkin's disease (lymphadenoma or pseudo-leukaemia) . The pathology of the two diseases is iden- tical but for the blood count. In Hodgkin's disease the blood is normal, or shows only a moderate anaemia or leucocytosis (poly morphonuclear cells alone increased) , and the diagnosis is easily made. 2. Tumors of the spleen and epecially of the kidney are very apt to be mistaken for leukaemia. Within a year I have been asked to examine the blood in three cases of "leukaemia," all of which turned out to be malignant disease of the kidney. In all of these there was a large tumor resembling the spleen in the left hypochondrium and a very large increase of white cells. In two of them the blood was examined fresh and the great number of white cells in the slide taken as evidence confirmatory of leukaemia. The stained specimen, however, showed only LYMPHATIC LEUKAEMIA. 175 marked leucocytosis with ninety per cent of adult cells of the ordinary type and no myelocytes. Other large tumors of this region showed similar results. Occasionally cases of leukaemia with numerous metastases are described as " sarcomatosis, " and then it is asserted that the blood of leukaemia is identical with that of sarcoma. The source of the mistake is obvious. 3. Adenitis with hyperplasia due to tuberculosis shows usually normal blood ' and is thus easily distinguished from leukaemia. Leucocytosis is often present in syphilitic cases and still more marked in those due to cancer or sarcoma, but the counts rarely reach 30,000 and myelocytes are absent or very scanty. 4. One case of hydronephrosis, in which the distention of the sac was so great that it presented as a hard, solid tumor on the right hypochondrium, was taken for leukaemia by a competent observer some years ago. The normal blood examination re- vealed the mistake, and excluded also malignant disease in all probability. The diagnosis was only reached, however, at the autopsy. 5. Huge leucocytosis in pneumonia or malignant disease may often cross the old boundary line of 100,000 white cells, beyond which none but leukaemic cases were supposed to ven- ture. The differential count sets us right instantly, showing ninety per cent or so of the increase to be made up of ordinary polymorphonuclear leucocytes. 6 and 7. The large spleen and cachectic appearance asso- ciated with chronic malaria and long-standing suppurations may be easily distinguished from leukaemia by the absence of any- thing more than anaemia and leucocytosis in the blood. Red cells. White cells. Lympho- cytes. Poly- nuclear leucocytes. Myelo- cytes. Nucleated red cells. Leukaemia (splenic- myelogenous). Leukaemia ( 1 y m - phatic). Hodgkin's disease . . Tumors of or near the spleen. Leucocytosis in gen- About 3,000,000 About 3,000,000 About normal. Usually diminished. 450,000 100,000 7,500 20,000 to 40,000 May be over About 7.6 per cent. About 96 per cent. Normal. Greatly decreased. Greatly About 50 per cent. About 3 per cent. Normal. Greatly increased, Greatly About 37 per cent. Absent. Absent. Few if any. Few if Very numerous. Rare. Absent. Few. Few at eral. Chronic malaria. . . . Amyloid disease Hydronephrosis . Much diminished. Usually diminished. Normal. 100,000 Somewhat increased. Usually increased. Normal. decreased. Usually increased. Usually decreased. Normal or decreased. increased. Usually decreased. Usually increased. Normal. any. Few if any. Absent. Absent. times. Few. May be a few. Absent. 1 Sometimes marked leucopenia. 176 SPECIAL PATHOLOGY OF THE BLOOD. EFFECT OF INTEBCUBBENT INFECTIONS. There are on record about thirty cases in which leukaemia (acute or chronic) has been complicated with some intercurrent infection, with marked effect upon the blood in all but one. This single case was an acute rheumatic arthritis reported by Bichter in the discussion of Fraenkel's article in the Deutsche medicinische Wocliemchrift for 1895 (Nos. 39, 43, and 45), p. 639. Here the blood remained unchanged. Miiller's ' case of lymphatic leukaemia was complicated by a septicaemia, and the count of white cells rose from 180,000 to 400,000 per cubic millimetre, with a marked increase in the per- centage of polymorphonuclear cells. Here was a genuine leu- cocytosis added to a leukaemia. With the exception of these two cases, all those hitherto published have shown a marked progressive decrease in the total number of leucocytes without any change in the percentages of the different varieties in twelve, while eight showed like Miiller's an increased percentage of the polymorphonuclear cells despite the decrease in the total leucocyte count. Marischler 2 in a case of lymphatic leukaemia with cancer of the kidneys found : 1. At First. 2. Later. Red cells 3,450,000 2,400,000 White cells 96,000 48,000 Haemoglobin 50 per cent. 30 per cent. Polymorphonuclear cells 15.6 " 57.5 Small lymphocytes 83.3 " 40 Large lymphocytes 1.8 " 1.6 Eosinophiles 18 " .16 Myelocytes .16 Various infections miliary tuberculosis, pneumonia, grippe, erysipelas, abscess of kidney, septic lymph glands alike de- creased the leucocyte count. In one case a rise just before death was observed. Thus in Henck's case the leucocytes fell from 400,500 to 89,000, in one of Miiller's from 246,900 to 57,300, in Kovacs' from 67,000 to 17,000, in Zeissl's from 140,000 to 9,350. I 'Muller: Deut. Archiv fur klin. Med., 1892, vol. 50, p. 47. nVien. klin. Woch., July 23d, 1896. LYMPHATIC LEUKEMIA. 177 have already mentioned a case of lymphatic leukaemia (page 173) in which the leucocytes fell from 40,000 to under 500, this last being on the day of death. In this case the percentages of the different varieties of leucocytes remained entirely unchanged. Herrick l reports a case complicated by acute streptococcus infection in which the white cells were 60,000 at the time of death. How high they may have been earlier is not known. It appears, therefore, that when an infection complicates leukaemia we may have 1. No effect (see case of rheumatic fever as a complication, just mentioned). 2. A genuine leucocytosis on top, so to speak, of the leu- kaemia, with an increased percentage of polymorphonuclear cells. 3. A decrease in the leucocyte count with or without an in- crease of polymorphonuclear cells. This decrease is by far the most common result and may go far below normal as death ap- proaches. Goldschneider 2 found that by the injection of splenic extract and other substances he could bring about a similar diminution in the number of leucocytes, but that, as in the case of intercur- rent infections, this diminution was not accompanied by any im- provement in the patient's condition and death followed as usual. Abscesses occurring in leukaemic patients are filled with adult leucocytes as ordinary abscesses are, and do not contain mye- locytes. HODGKIN'S DISEASE. (Pseudo-Leukcemia, Lymphoma). The diagnosis of this disease is impossible without the blood count. Its pathology is identical with that of leukaemia and even post mortem the two diseases are indistinguishable so far as the lesions outside of the blood are concerned. Yet the blood is in no way peculiar, but presents in most cases all the characteristics of the normal tissue. Its value is as negative evidence, telling us in a given case that leukaemia is absent even though all the other signs and symptoms may be those of leu- kaemia. 1 Loc. cit. * Discussion of Fraenkel's article. 12 178 SPECIAL PATHOLOGY OF THE BLOOD. (I.) Transitions from Hodgkin's disease to leukaemia are said to have taken place under the eyes of competent observers, but they are very rare. Only three such cases are on record so far as I know, that of Fleischer and Penzoldt, 1 that of Hosier, 2 and one reported by Senator, 3 where two sisters came under obser- vation, both suffering form Hodgkin's disease. One died of it; in the other the blood changed to that of leukaemia before death. 4 Doubtless many of the other cases supposed to exemplify a similar transition were really cases in which a leucocytosis arose owing to some inflammatory complication, as not uncommonly occurs (see below, Table X.). From the existence of these very rare cases of a transition to leukaemia it has been supposed, especially by French observers, that Hodgkin's disease is simply an early stage of true leu- kaemia and that this would always become apparent were it not that the patients die of some intercurrent disease before the signs of leukaemia have time to show themselves in the blood. One difficulty with this view is that there occur chronic cases which last from eight to ten years without any change in the blood. Another difficulty is that the transition is in fact rare despite the relative frequency with which the disease is met with. (II.) Undoubtedly many cases diagnosed as Hodgkin's dis- ease are in fact cases of glandular hypertrophy due to syphilis or tuberculosis, and this fact has led many to the belief that all cases called Hodgkin's disease are in reality only syphilitic or tubercular adenitis. In a considerable number of cases, how- ever, tuberculosis has been disproven by careful inoculation ex- periments with the glandular tissue, and there is no reasonable doubt that some cases at any rate are not due to tuberculosis or syphilis. Probably the diagnosis can never be made with ab- solute certainty during life. (III.) The frequent occurrence of fever and other symptoms characteristic of an infectious disease has led some writers to class it as such. In a certain percentage of cases the disease 1 Deut. Arch, f . klin. Med. , vol. 17. 2 Zierassen's "Handbuch d. Path, and Therap.," vol. 8. 3 Berl. klin. Woch., 1882, p. 533. 4 It is noteworthy that all these cases are of some years' standing before Ehrlich's methods were much used. HODGKIN'S DISEASE. 179 i (like leukaemia) has run an acute course, lasting not more than six weeks from the first symptom to death. In some chronic cases the same sort of evidence of an infectious nature has been brought forward. Ulcerations occur in the mouth and intes- tine through which morbid products might gain admission. Various bacteria (pyogenic and others) have been found in the blood and tissues from time to time, but numerous negative examinations for micro-organisms are also on record, and the evidence is insufficient to establish the infectious nature of the disease. None the less, there is a growing tendency among the leading writers and observers in Germany and elsewhere, to be- lieve that the disease will ultimately be shown to be infectious. (TV.) Meantime most surgeons continue to regard it as a form of sarcoma 'and to treat it like malignant disease. The Blood. Whatever the nature of the disease, we find in the earlier stages of most cases normal blood as will be seen in Table X. (cases 7 to 23 inclusive). As the disease progresses the haemoglobin soon begins to fall, later the red cells, until, as at the end of Case 10 of the pres- ent series, the blood may reach the severest grade of anaemia. In acute cases the anaemia may develop very rapidly. The usual qualitative changes characterizing severe secondary anae- mia may be present. TABLE X. HODGKIN'S DISEASE. Red White Per cent X cells. cells. haemo- Remarks. GO globin. 28 F. 5,500,000 64,000 75 Polymorphonuclear cells, 95 per cent. Lymphocytes, 5 per cent. IT 3 848 000 39 200 48 A TVA i enn r u 95.2 per cent. Lymphocytes, 4. 6 per cent. 24 F. 4,886,000 32,000 53 19 F 5,528,000 22,200 Diff 200 cells Polymorphonuclear cells 86 5 5,160,000 25,400 per cent. Six weeks later. Lymphocytes, 12. per cent Eosinophiles, 1.5 19 M. 2,480,000 20,200 33 Stained specimens normal. 1 There are no reliable differentia between sarcoma of lymph glands and "benign" lymphoma histologically. 2 Diff. = Differential count of. 180 SPECIAL PATHOLOGY OF THE BLOOD. TABLE X. HODGKIN'S DISEASE (Continued). Age. 1 Red cells. White cells. Per cent haemo- globin. Remarks. 37 -\I 5,990 000 13 500 Polymorphonuclear cells 95 per cent Lymphocytes, 5 per cent. 25 M. 5,440,000 9,500 59 Death; autopsy. 19 F. 5,724,000 C.,800 42 Polymorphonuclear cells, 60 per cent. Lymphocytes, 40 per cent. Adult. M. 3,652,000 5,800 Diff. 300. Polymorphonuclear cells, 50.0 per cent. Lymphocytes, 45.3 per cent Eosinophiles, 1.3 Myelocytes, 1.7 Big spleen, pallor, nosebleed, debility. 29 M. 5,210,000 3 840000 5,000 5,600 1 000 000 58 M. 2,820,000 4,800 60 Polymorphonuclear cells, 80 per cent. Lymphocytes, 17 per cent. Eosinophiles, 3 31 M. 4,560,000 4,000 5,800 23 AI 4 210 000 3 332 Eosinophiles, 4 M. 3,800,000 1,440 67 Diff. 500. Polymorphonuclear cells, 71.25 per cent. Lymphocytes, 28.0 percent. Eosinophiles, .75 " One normoblast. No Diff 200 Polymorphonuclear cells 63 5 per leucocy- tosis. cent. Lymphocytes, 36.5 percent. Eosinophiles, 1 Many of the lymphocytes have two nuclei. No Diff 300 Polymorphonuclear cells 41 7 per leucocy- tosis. cent. Lymphocytes, 48.4 per cent. Eosinophiles, 9.3 Myelocytes, .6 4 -\I No Diff 500 Polymorphonuclear cells, 60.2 per leucocy- tosis. cent. Lymphocytes, 36 percent. Eosinophiles, 5.6 ' Myelocytes, 2. Two normoblasts. F Diff 500. Polymorphonuclear cells, 92.6 per cent. Lymphocytes, 5.2 per cent. Myelocytes, 2.2 No eosinophiles. No Diff. 313. Polymorphonuclear cells, 62.3 per leucocy- tosis. cent. Lymphocytes, 37 per cent. Myelocytes, .6 " 28 M. 5,218,000 11,800 85 Polynuclear, 51 per cent. Small lymphocytes, 35 Large Eosinophiles, 7 HODGKIN'S DISEASE. TABLE X. HODGKIN'S DISEASE (Continued). 181 Age. 1 Red cells. White cells. Per cent haemo- globin. Remarks. 30 M. 5,280,000 6,800 55 Diff. Polymorphonuclear cells, 76 per cent. Lymphocytes, 22.3 " Eosinophiles, 1.4 " Myelocytes, .3 No nucleated red cells. 32 M. 4,616,000 2,400 70 2,200 Diff Polymorphonuclear cells, 69 per cent. Small lymphocytes, 19 " Large " 18 Eosinophiles, 4 ?ew normoblasts. Wliite Cells. When inflammation arises in the glandular tumors and some- times when none is found, the white cells may be greatly in- creased, even up to a ratio of 1 : 80 red cells, as in Case 1 of the present series. There is, however, no more resemblance to leu- kaemia than in any other form of leucocytosis, the polymor- phonuclear cells alone being increased. There is no reason for supposing, as Eeinert * does, that relative diminution of the lymphocytes is owing to the diseased condition of the lymph glands, for, unless oome septic process gets a foothold in the glands, the lymphocytes present a normal number or even (as in Case 16) considerably increased percentages. Pfeiffer 2 has recently reported a case of the cutaneous form of the disease with sixty per cent of lymphocytes out of a total leucocyte count of 6,500. As in any other cachectic condition, small numbers of mye- locytes may be found. They were seen in six of our cases out of eighteen in which a color analysis was made, the highest per- centage being two per cent. Eosinophiles are usually decreased when leucocytosis is present. Summary. Normal blood in early stages. Later often marked anaemia. Sometimes leucocytosis. 1 "Die Zahlung der Blutkorperchen, " Berlin, 1891. 2 Pfeiffer: Wien. klin. Woch., 1897. 182 SPECIAL PATHOLOGY OF THE BLOOD. Diagnostic Value. The only help given us by the blood is in excluding leukae- mia. Syphilis, tuberculosis, or malignant disease might cause similar blood changes or lack of changes. EFFECTS OF SPLENECTOMY ON THE BLOOD. Twenty-three cases are on record in which the blood has been studied after operation, but only some half-dozen of these are carefully recorded. They explain themselves : TABLE XL _c O op Si. 5 1 1 1 A j Red cells. White cells. *! Polymorp nuclea neutrophi Small lymphocy Large lymphocy Eosinophil Averag diameter red cell: Remarks. *1 4,570,000 8,000 63 Before operation. 4, 970, 000 30,000 64 Three days after. 5, 180, 000 65,000 77 Six days after. 4, 800, 000 17, 500 66 Forty -eight days after. 4, 353, 000 11,700 85 Four months after. 3, 300, 000 11,600 85 Five years after. 3 3,200,000 53,000 65 ( 1893, for abscess. ) Three weeks after. 4,500,000 13,800 80 Four months after. t 1,634,000 12,000 45 61 16 20 3 8.ij* Operated (April 9th, 1893) for malaria] hypertrophy with twisted pedicle. April 23d. 2,460,000 20,000 87 49 18 32 1 May 6th. 4,530,000 27,000 110 66 18 15 1 7.T> May 13th, 1894. 3,977,000 8,000 100 62 21 11 6 October 2d, 1895. f4 4,850,000 30,000 108 83 8 8 1 .... Operation for hypertro- p h i e d , wandering spleen. Before opera- tion. 4,700,000 39,000 100 91 5 4 .... Seven days after. 3,630,000 18,000 105 78 15 6 1 .... Two months after. 2,750,000 20,000 63 84 5 10 1 Three years after. *Czerny : Cited in Laudenbach ; Arch, de Physiol., 1896, p. 724. f Hartman and Vaquez : Soc. de Biol., February 5th, 1895. PART II. ACUTE INFECTIOUS DISEASES. CHAPTEE in. INFLUENCE OF FEVER ON THE BLOOD. SOME of the blood-changes found in acute infections are to be regarded as due simply to the fever associated with the dis- ease. It is worth while, therefore, to consider what fever per se can do to the blood. Maragliano ' and others have shown that during fever from any cause a contraction of the peripheral vessels occurs. When fever disappears, whether spontaneously or from the action of antipyretics (phenacetin, quinine, etc.), a dilatation of the ves- sels follows. Following the laws to which we have so often alluded, the contraction of the vessels causes a concentration of the blood with rise in specific gravity and in the number of blood cells per cubic millimetre. This concentration is still further increased by the greater loss of water which the organism suffers during fever than under normal conditions. The effect of these two influences in increasing the number of red cells per cubic millimetre is, however, counteracted to a considerable extent by the sharing of the blood in the general tissue destruction which goes on with increased rapidity during fever. Many corpuscles are thus destroyed, but until the tern- perature falls the anaemia is covered up by the concentration. When the fever leaves the patient there is a sharp fall in the number of cells per cubic millimetre, due partly to the destruc- tion of corpuscles (hitherto masked by concentration) and partly to the dilution of the blood which is the result of the post-febrile dilatation of the peripheral vessels above mentioned. The sud- 1 Zeit. f. klin. Med.. vols. 14 and 17. 184 SPECIAL PATHOLOGY OF THE BLOOD. denness of this fall in the count is proportional to the sudden- ness of the fall in temperature. The alkalinity of the blood has been often said to be dimin- ished in fever, but recent research tends to show that these re- sults were obtained by faulty technique, and it is doubtful whether the reaction of the blood shows any constant changes in fever. Leucocytes and fibrin show no constant changes, though in the majority of infectious fevers they are increased. PNEUMONIA. The Blood as a Whole. (a) Bacteriology. The diplococcus lanceolatus has been found in the blood of pneumonic patients repeatedly, especially in those in whom there has been some secondary diplococcus infection (e.g., diplococcus endocarditis); but such findings are rare and have generally been in fatal cases with very severe gen- eralized infection. For example, Sittmann ' out of 16 cases found diplococci in the blood of 6, most of which were complicated with lesions in other organs, and 4 of which died, while of the 10 whose blood was sterile, 9 recovered. Boulay 2 found the organism in 2 cases shortly before death. Belfanti 3 found it but 6 times out of a large number of cases, and f these 6, 5 died. Goldschneider 4 and Grawitz 5 got similar results. Cohn 6 in 32 cases found the organism in 9; 7 of these died. The other 2 had empyema and other evidences of metastatic action of the pneumococci. Fraenkel has ob- tained over 300 colonies from one puncture. Their virulence was less than that of those in the sputa, showing apparently the effects of the blood's antitoxic power. Nevertheless, it is obvious that the presence of pneumococci in the blood is a bad prognostic sign. 1 Deut. Archiv f. klin. Med., 1894, p. 323. 2 Paris Thesis, 1891. 8 Riforma Medica, Naples, 1890, No. 37. 4 Deut. med. Woch., 1892, No. 14. 5 Grawitz: Charite-Annalen. vol. 1. 6 Cohn: Deut. med. Woch., 1897, No. 9. PNEUMONIA. 185 (b) Coagulation is remarkably rapid and in fresh specimens the fibrin network is very thick and appears within a few minutes. (^' ... -I ...v ; . V V * 97 80,000 p .1 . * c a V a n / ^ 80,000 / \ *- - ^ 18,000 f 16,000 j \ \ 14,000 ?\ / 12.000 \ f \ V - 10,000 4 / 8,000 * * 6,000 4,000 3,000' The upper chart shows the course of the temperature, the lower that of the leucocytes. in the (fatal) cases in which leucocytosis is absent, data are scanty. Bieganski thought the polymorphonuclear varieties decreased, Eieder found them increased, while Billings finds them normal. No general law can be stated on this point as yet. In one remarkable case occurring at the Massachusetts General Hospital in 1894, the conditions were entirely different from those just stated. The patient, a girl of six, had at en- trance 72,100 leucocytes per cubic millimetre. Two days after PNEUMONIA. 189 the count was 94,600. A differential count made at the same time showed that the small lymphocytes made up 66 per cent of all the 94,600 leucocytes per cubic millimetre. The poly- morphonuclear cells were reduced to 30 per cent. Lymphatic leukaemia was thought of, but the leucocytosis was gone in ten days, and within a fortnight the patient left the hospital well. I have seen one reference to such a condition. " In a certain number of cases the leucocytosis is characterized by the great number of the youngest forms of leucocytes. This condition persists during convalescence." 1 Diagnostic and Prognostic Value. 1. In cases of so-called "central pneumonia" in which the symptoms but not the physical signs of the disease are manifest, the presence of a well-marked leucocytosis is often of great diag- nostic value. It excludes malaria, typhoid, and uncomplicated grippe as causes of fever, and if scarlet fever and suppuration can be excluded by other evidence, it makes pneumonia very probable. I have repeatedly seen the diagnosis of pneumonia made in the absence of physical signs and largely on the evidence of the blood count, the diagnosis being confirmed several days later by the appearance of typical signs of consolidation. In a case of Dr. F. C. Shattuck's, sick five days, yet showing no signs of consolidation of the lung, the presence of a marked leucocytosis excluded typhoid, the only other likely diagnosis, and led Dr. Shattuck to treat the case as pneumonia, the wisdom of which course was later demonstrated by the appearance of signs of consolidation. 2. Between pneumonia and capillary bronchitis the condition of the blood is of no help, as the latter also causes leucocytosis, and some cases affecting the larger tubes do the same. 3. In cases of pneumonia occurring in very old or very 3 r oung people, in which the fever and symptoms may be very slight, the presence of leucocytosis may be the first thing to direct our attention to the lungs, dyspnoaa and cough being ab- sent. 'Stienon: Jour, de Med., de Chirurg. et de Pharm., Bruxelles, 1895, t. iv., fasc. 1. 190 SPECIAL PATHOLOGY OF THE BLOOD. In prognosis, the important point is that the absence of leu- cocytosis is a very bad sign, ivhile its presence is neither good nor bad. It must be remembered also that in the very mildest cases we may find the same absence of leucocytosis which in any other but the mildest would be almost surety fatal. This last point, which appears to rne of great importance, is illustrated by the following figures : Halla reported 14 cases ; 2 had no leucocytosis, and both died. Billings reported 22 cases ; 1 had no leucocytosis and died. Laehr with 16 cases, and Bieder with 26, got similar results. Ewing in 101 cases found leucocytosis absent in 6 ; 6 died. Yon Jaksch and Kilodse likewise maintain that the absence of leucocytosis is usually fatal. In the Massachusetts General Hospital 329 cases have been studied. In general they entirely confirm the results obtained by Billings and summarized above; 32 of them presented no leucocytosis at any time, and of these 32, 30 died, another one seemed moribund but finally recovered, while the remaining case was a very mild one. The evidence, therefore, is overwhelmingly in favor of the view that where leucocytosis is absent in any but the mildest cases the prognosis is almost fatal. The presence of leucocyto- sis, on the other hand, is no guaranty whatever of a favorable issue. The series of cases at the Massachusetts Hospital is too large to exhibit in tabular form. Their results may be summarized as follows : Cases with leucocytes under 10,000 = 32 (30 of these fatal) between 10,000 to 15.000 = 38 15,000 " 20,000 = 72 20,000 " 25,000 = 65 25,000 " 30,000 = 37 30,000 35,000 40,000 45,000 50,000 35,000 = 22 40,000= 4 45,000= 7 50,000= 4 55,000= 5 but not accurately counted = 43 Average = 24, 000 - 329 TYPHOID FEVER. 191 TYPHOID FEVER. Bacteriology, Although the bacilli of Eberth are occasionally to be found in the blood by culture, it is only in the marked cases that they occur, and then but rarely, so that at present we derive no help in doubtful cases by the bacteriological examination of the blood. Kiihnau, using 5-10 c.c. of blood, found the organism in 10 out of 41 cases, from 2 to 9 slow-growing colonies in each. Other observers have been successful in only 7 out of a total of 176 cases examined. Block has recently succeeded in isolating the organism twice during the life of a patient who sub sequently died. In 6 other cases he was unsuccessful. It has been asserted that when the bacilli enter the blood the serum reaction (vide infra) does not appear. Toward the end of th- disease, when the temperature is apt to be very irregular (so-called " period of steep curves"), pyogenic cocci are occasionally to be found in cultures made from the blood, and doubtless account for many of the recrudescences and temporary febrile attacks, with or without chills, which are so common in early convalescence. The Blood as a Whole. 1. Coagulation and fibrin are normal. 2. Specific gravity follows the course of the haemoglobin. 3. The general effects of fever (see above, page 183) are in part accountable for the changes next to be described, while some of them are more peculiar to typhoid fever. Red Cells. During the first two weeks there are no consid- erable changes, except in so far as a certain amount of concen- tration of the blood with apparent increase of cells may be brought about by diarrhoea or sweating. Baths have a like effect if the blood is examined just after the immersion. 1 In the third week the red cells usually begin to decrease and in ex- treme cases may get as low as 1,300,000 at the beginning of con- valescence i.e., when body weight begins to increase. Hay em considers that the diminution begins rather suddenly in the middle or end of the third week of severe cases, but according 1 Antipyrin and acetanilid have no effects on the red cells. 192 SPECIAL PATHOLOGY OF THE BLOOD. to Thayer the diminution is gradual, though at first sight, grow- ing more rapid at the time of defervescence, and continuing often into convalescence. The lowest point is reached about the first week of convalescence. The following figures (Thayer) illustrate this. First week, Second week, Third week, Fourth week, Fifth week, Sixth week. 2 counts. 10 counts. 9 counts. 6 counts. 7 counts. 5,636,000 4,960,599 4,951,535 4,038,333 3,856,786 4,364,250 His later counts show a gradual increase. He finds that the amount of anaemia bears, as a mde, a direct relation to the severity of the case, but in one of his cases a grave anaemia (1,300,000) followed a mild attack. " The anaemia may be severe enough to form of itself a dangerous complication of the proc- ess." Henry has likewise recorded counts as low as 1,306,000 and 804,000 in two convalescent typhoids. Hcemoglobin. The loss of coloring matter roughly parallels that of the red cells, but is always relatively greater and is slower in reaching normal. In the case just noted it was 20 per cent, color index .7. Leucocytes. The absence of anv increase oi the white cells is the most important point. Starting with an approximately normal count, the number falls during the fever, often below 2,000, according to Hay em, and sometimes below 1,000 per cubic millimetre. Khetagurow finds the lowest counts (2,500-3,000) about the end of the third week. Thayer 's figures are as follows : First week, Second week, Third week, Fourth week. Fifth week, Sixth week, 21 counts. 50 counts. 40 counts. 28 counts. 16 counts. 5 counts. 6,984 6,468 6,260 5,877 6.621 7,000 In the four hundred and ninety-one cases counted at the Massachusetts General Hospital, the course of the leucocytes has unfortunately not been followed by weeks with sufficient accuracy to make comparisons of value. In a general way, however, they corroborate all of Thayer's positions. At the be- ginning of the cases the count was often high (11,000), owing probably to concentration of the blood by starvation and diar- rhoea. The high count of red cells confirmed this, the ratio of red to white remaining normal. The counts of leucocytes then gradually diminished, as in Thayer's cases. TYPHOID FEVER. 193 The range of the counts was as follows : Between 1,000 and 2,000= 7 cases. 2,000 " 3,000= 33 " 3,000 " 4,000= 59 " 4,000 " 5,000= 108 " 5,000 " 6,000= 82 " 6,000 " 7,000= 72 " 7,000 " 8,000= 47 " 8,000 " 9,000= 37 " 9,000 " 10,000= 29 " 10,000 " 11,000= 10 " Over 11,000 = 7 " 491 cases. From these figures I have excluded all cases counted only under circumstances likely to concentrate" the blood (cyanosis, after baths, after severe diarrhoea). There is no doubt that leucocytosis does occasionally occur when no complication exists so far as loe can ascertain during life. Four of the cases over 11,000 (see the above table) were counted repeatedly and complications were carefully sought for, but none were found. The most striking case showed the fol- lowing counts : October 3d 13,100 4th 13,000 5th 16,500 7th 13,300 8th 11,200 " 10th 10,600 13th 13,500 15th 17,700 17th 15,500, death ; autopsy. The autopsy showed typical typhoid lesions and nothing else. 1 Another and much milder case showed 11,000-12,000 white cells constantly for over two \v^eks, and no cause could be found to account for it. The great rarity of such cases and constant association of leucocytosis with any of the numerous complications which we can recognize, rather inclines me to the belief that in all the cases in which leucocytosis exists constantly, some complication 1 Thrombosis of internal veins and osteomyelitis were not carefully searched for at autopsy and may have existed. 13 194 SPECIAL, PATHOLOGY OF THE BLOOD. really is present though unrecognized. The possibility of a secondary septic infection, of an osteomyelitis, or phlebitis of internal veins cannot be excluded without further evidence. Examples of the effect of complications are as follows: Perforation. Case I. (a) Five days before perforation, 8,300 (b) At time of the perforation, . 24,000 Case II. At time of perforation, . . 18,500 Phlebitis. Case I. (a) Two days before onset, . . 6,400 (b) At time of the onset, . . . 12,900 (c) One week later, .... 10,100 Case II. (a) One week before onset, . . 4,800 (b) At time of onset, .... 16,200 Otitis Media. Case I. (a) At entrance, 5,300 (6) Mastoid abscess, .... 16,400 Case II. (a) At entrance, 8,400 (b) Two weeks later, after open- ing drum membrane (sero-pu rulent clis- charge), 11,200 Case III. (a) At entrance, 7,320 (b) Otitis, 14,000 A freely discharging otitis soon ceases to cause leucocytosis, e.g., a case of serous otitis media seven days after puncture, but still freely discharged, showed but 5,320 white cells per cubic millimetre. An abscess of the buttock raised the count from 8,000 to 11,200, and a hemorrhage from 8,000 to 11,300. General bronchitis has usually no effect in augmenting the leucocyte count unless the disease invades the smallest tubes (capillary bronchitis). Thus two cases of this affection showed 9,000 and 8,000 leucocytes respectively. Cystitis had no effect in two cases. 1 In two cases whose symptoms simulated otitis (deafness, rise of temperature, pain in the head, and in one a convulsion) but whose blood counts were normal, the trouble turned out to l>e functional and nothing came of it, the symptoms disappear- ing within twenty -four hours. 1 1 have an impression based on rather fragmentary evidence that com- plications directly due to the Eberth bacilli, e.g., Eberth cystitis or Eberth pneumonia, do not raise the leucocyte count. I hope to investigate this point later. TYPHOID FEVER. 195 Some observers l have noted a slight leucocytosis at the be- ginning of convalescence. Thayer did not find this, and I have been equally unsuccessful. It occasionally happens in very exhausted patients that com- plications fail to produce any leucocytosis, the patient (as in some fatal cases of pneumonia or purulent peritonitis) being unable to react against the infection. For example, I have seen a large ischio-rectal abscess develop in a moribund typhoid patient without producing any effect on the leucocyte count. Von Limbeck has noticed the same lack of reaction in typhoid patients after a hemorrhage and bronchopneumonia, and Eieder in croupous pneumonia occurring as a complication. These cases, however, are exceptional, and in many of them the percentage of adult leucocytes rises, though no increase in the total leucocyte count is present. This increased percentage of polymorphonuclear forms generally betrays the presence of the complication, since during most of the disease (if uncompli- cated) the polymorphonuclear forms are diminished. In normal cases the blood begins to return to normal as soon as the fever is gone and reaches the normal in the sixth or seventh week. Qualitative Changes. Red Cells. The condition is either normal or shows the changes common to all varieties of secondary ansemia. White Cells. All observers are agreed upon the following changes : 1. The polymorphonuclear cells progressively diminish with a corresponding increase in the lymphocytes. This change is but slight in the first two weeks, but grows marked in the latter part of the illness, the polymorphonuclear cells falling below 50 per cent. Among the lymphocytes, the larger forms pre- dominate. 2. It is not until after the disappearance of fever (from three to ten days after it, according to Ouskow) that the polymorpho- nuclear cells begin to increase again and their normal percentage 1 E.g., Aporti and Radaeli (llth Congress for Medical Science, Rome, March 29th, 1894) . 196 SPECIAL PATHOLOGY OF THE BLOOD. is not reached until the tenth or eleventh week. Thayer's differential counts show : Second week, Third week, Fourth week, Fifth week. Sixth week, 5 counts. 1 count. 3 counts. 1 count. 2 counts. 71. 7 per cent. 66. 5 per cent. 65. 3 per cent. 58. 5 per cent. 53. 4 per cent. 3. Eosinophiles are present in small numbers. Summary. 1. Post-febrile anaemia, sometimes very intense. 2. No leucocy tosis ; in late weeks leucopenia. 3. Increased percentage of lymphocytes at the expense of polymorphonuclear forms, especially marked in later weeks. 4. Most complications cause leucocy tosis. Diagnostic Value. There are few diseases (outside of those known as diseases of the blood itself) in which the blood count is so often of value in diagnosis. The diagnosis of typhoid fever is to be made by ex- clusion exclusion of other causes of fever and of local inflam- matory processes in particular. 1. Now in this process of exclusion, the blood is a most powerful adjuvant, inasmuch as almost all local inflammatory processes have leucocytosis, while typhoid (uncomplicated) does not. I have seen two cases in which the chart and symptoms pointed to typhoid but in which the persistent marked leucocytosis directed attention to the search for an inflammatory focus. Both were at first unattended with pain, tenderness, or other localiz- ing symptom, but later signs and symptoms began to point to the liver, from which pus was evacuated by puncture. These cases of abscess of the liver are typical of the value of blood ex- amination for any deep-seated suppuration. I have seen good clinicians puzzled for twenty -four hours over the diagnosis be- tween appendicitis and typhoid, but the indication of the blood count was always fulfilled. All pyaemic or septicsemic processes are distinguishable from typhoid by the same test the pres- ence of leucocytosis in the former. Of the value of the blood in distinguishing certain cases of pneumonia from typhoid I have already spoken on page 189. 2. Aside from local or general pyogenic infections perhaps the disease most often confounded with typhoid is malaria. TYPHOID FEVER. 197 This is especially the case in the southern part of this country, where for want of proper blood examination the confusion of the two diseases is indicated in such a term as " typho-malarial fever." Malaria and typhoid are alike in having no leucocy- tosis, but the presence of the malarial parasite is an absolute test and in marked cases is always decisive. Very mild cases of malaria may show so few organisms in the peripheral circu- lation that without prolonged search they cannot be found, and in the severest types of all, the organisms are not very abun- dant. In the vast majority of cases, however, the organism can be readily found and our diagnosis made certain. 3. Tuberculosis, if uncomplicated by any pyogenic organ- isms, cannot be distinguished from typhoid by the examination of the blood alone, as neither disease shows leucocytosis. A large proportion of lymphocytes is commoner in typhoid than in tuberculosis, but it may occur in either disease. In the majority of cases, however, tuberculosis is complicated with septicaemia from a secondary pyogenic infection, and is then easily distinguished by the existence of leucocytosis. 4. Typhus fever has not been well studied and the reports of its blood condition are contradictory. At present we cannot say whether or not it can be distinguished from typhoid by the blood examination. In most cases the absence of a serum reac- tion will exclude typhus. 5. Two cases of erythema nodosum with fever between 101 and 103 gave me trouble in diagnosis lately. In both the blood was normal and differed from typhoid only by the absence of a serum-reaction. The occurrence of complications in typhoid may mask its characteristic blood changes so as to make the blood useless in diagnosis ; but in most early cases, in which the diagnosis is es- pecially important and difficult, the blood shows no leucocytosis and is therefore of great value in the exclusion of other diseases. DIPHTHERIA. Bacilli of diphtheria in the circulating blood are practically never to be found. The specific gravity, according to Grawitz, is above normal at the height of the disease. He obtained the same result experi- 198 SPECIAL PATHOLOGY OF THE BLOOD. mentally by injecting cultures of the Klebs-Loffler bacillus into dogs and rabbits. He concludes that the poison of the disease is lymphagogic and so concentrates the blood. Red Corpuscles. Morse's 1 investigations show an average of 5,100,000 in twenty cases counted during the first week of the disease and of 5,150,000 in 10 cases during the second and third week of the disease practical^ normal figures. These are the first systematic 2 investigations of the red cells in diphtheria and are confirmed by the reports of Ewing, Engel, and Billings. The latter observer in counts made in seven cases during the first five days of illness found an average of 5,600,- 000-h red cells per cubic millimetre. During the first five to ten days after this, the same cases showed an average loss of 510,000 cells per cubic millimetre ; five out of the seven showing consid- erable losses, two remaining about the same. These were cases treated without antitoxin. The two cases showing no loss of red cells were both very mild, one having no membrane at any time. The diminution ranged from 470,000 (third day) to 2,040,000 (sixth day). As a rule no diminution can be made out until after the third or fourth day. Out of twenty-three cases treated with antitoxin and each counted several times over, only three showed any considerable diminution in the red cells and these lost less than 400,000 each, not much beyond the limit of error (200,000) allowed for by the investigator, and all of them severe cases. Six patients who were anaemic when admitted (average =4, 640, 000) showed a steady rise in the red cells as the disease (treated with anti- toxin) progressed. It is evident from these figures that antitoxin largely pre- vents the anaemia which usually develops in the first five to ten days. In cases not treated with antitoxin the regeneration from the resulting anaemia is slow. Healthy individuals in- jected with antitoxin showed a very moderate reduction in the red cells in about one-half the cases, the greatest loss being 932,- 000 per cubic millimetre (fifteen cases counted by Billings) . Qualitative CJianges. Billings' careful study of stained speci- mens showed no deformities in size or shape and no nucleated red cells. Polychromatophilic red corpuscles were very few in 1 Boston Medical and Surgical Journal, March 7th, 1895. 2 Earlier reports are faulty as to technique. DIPHTHERIA. 199 the cases in which antitoxin was used, but more numerous where it was not used. Haemoglobin. Here again the most thorough investigations are those of Billings. In cases treated without antitoxin there was an average loss of ten per cent, regained in part during con- valescence, but as usual reaching normal later than the count of corpuscles. When antitoxin was given, the diminution of haemoglobin was less marked, but where the decrease did occur the return to normal was slow compared to that of the red cells, even when the patients were up and about and apparently well. White Corpuscles. Leaving out the older observations in which the technique was probably faulty, the principal investi- gators are Morse, Ewing, Gabritschewsky, and Billings. All agree that a considerable leucocytosis is present in most cases 34 out of 36 of Billings' cases, 26 out of 30 of Morse's (the latter made but one count in each case), 49 out of 53 of Ewing' s. In a general way, the severest cases show the greatest leucocy- tosis, but it does not follow the pulse, temperature, nor the ex- tent of the membrane, and " the ordinary clinical examination of the patient is of much greater value in ... prognosis . . . than any information to be gained from the examination of the blood. The latter is simply confirmatory, never indispensable" (Billings). Morse's conclusions are the same, although he con- siders that with notable exceptions the amount of membrane is a rough measure of the degree of leucocytosis. He finds no correspondence between the glandular swellings and the degree of leucocytosis, though he noted that " in the fatal ' septic' cases with greatly enlarged glands, " very high counts were present. Other cases with little or no enlargement of glands showed equally high counts, however. Swing's 4 cases without leucocytosis were all mild, but of Billings' 2 cases without leucocytosis one was the severest of his whole series, while the other was rnild. Of Morse's 4 cases without leucocytosis 3 were mild and 1 severe. Gabritschew- sky 's 14 cases all showed leucocytosis. Putting the results of these four observers together we see that when leucocytosis is absent the cases are either very mild or very severe, conditions analogous to those to be noted in pneu- monia and septicaemia. The counts in recent epidemics range from normal to 48,000 (Morse) or to 38,600 (Billings). Felsen- 200 SPECIAL PATHOLOGY OF THE BLOOD. thai l found 148,229 per cubic millimetre in one case, and Bouchut's 2 counts are often over 75,000. In a general way the counts rise while the disease progresses and fall gradually as improvement goes on, disappearing after the membrane. " The leucocy tosis is well marked by the third day and very likely earlier" (Morse). Billings found an in- crease after one day's illness, but usually less than was present later in the disease ; one of his cases, however, had a higher count on the first day of the disease than on any subsequent day, though no antitoxin was given. The injection of antitoxin has apparently no effect upon the leucocyte count (strange to say) except in the first twenty- four hours after its use. Immediately, i.e., within thirty min- utes after an injection, the leucocytes are stated by Ewing to be considerably diminished, but the leucocyte curve does not reach normal any sooner than in cases in which no anti- toxin is given, although it begins to fall in the majority of cases after the injection. The same thing (according to Billings) takes place without antitoxin. The leucocytes of healthy persons are likewise unaffected by antitoxin injections. Qualitative Changes. All authors agree that in most cases the neutrophiles are increased. Morse found an average of 80 per cent in 26 of his 30 cases. Of the other 4, 1 was normal and 3 subnormal (58, 59, and 59 per cent) ; 2 of these were con- valescent, the other had been sick a week and had 12,000 white cells per cubic millimetre. A similar lymphocytosis was present in 1 of Ewing' s 53 cases, and in 1 of Eieder's during convales- cence. Billings thinks such a lymphocytosis may be present in perfect health, mentioning cases with 32, 33, and 35 per cent of small lymphocytes in sound persons. Such a condition did not occur in any of his diphtheritic cases except in the single fatal case without leucocytosis. Here the polymorphonuclear cells were reduced to 55 per cent and the lymphocytes (large and small) made up the remaining 45 per cent, 28 per cent being large forms. 3 In the rest of his cases the polymorphonuclear 1 Archiv f. Kinderheilk. , vol. xv., p. 78, 1893. 2 Comptes Kendus, 1877, Ixxv. , No. 3. 3 In Rieder's case above referred to, aged three years, the lymphocytes rose from 19 per cent during the fever to 64 per cent in convalescence. DIPHTHERIA. 201 varieties averaged 80 per cent and the lymphocytes 19 per cent, the eosinophiles being reduced to 1 per cent on the average and often being entirely absent. With Morse eosinophiles averaged 2 per cent. The proportion of polymorphonuclear cells is usually directly proportional to the total increase of leucocytes. Ewing thinks that " the staining reaction of the leucocytes is an accurate measure of the severity of the diphtheritic infection," and this staining reaction he finds increased in favorable cases by the injection of antitoxin. Billings did not find any such changes in " staining reaction," though he claims to have carefully followed out Ewing' s pro- cedures. Engel 1 found that antitoxin at first slightly increased the percentage of lymphocytes, and sometimes this increase was very marked. In one case the lymphocytes increased from 24 to 65 sixty-five per cent after antitoxin. The point on which he specially insists is the presence of considerable numbers of myelocytes in fatal cases. Of the cases examined by him 17 died, and 9 of these had from 3.6 to 16.8 per cent of myelocytes in every one hundred leucocytes. Myelocytes were also present in some of the cases which recovered, but in smaller numbers (1.3 to 1.5 per cent.) In one case he found on the third day of the disease 4.3 per cent of myelocytes, and from this point the percentage gradually rose to 13.8 per cent, and then fell, there being 1.7 per cent present at the time of death. An abscess occurring in the case showed only the usual polymorphonuclear leucocytes in its con- tents. He concluded that a large percentage of myelocytes is a bad prognostic sign in any case. Myelocytes are not mentioned in any of the numerous differ- ential counts made by Gabritschewsky, Ewing, Morse, and Bill- ings, so that Engel' s observation is so far unique. Summary. 1. Moderate anaemia, especially in cases treated without an- titoxin. Regeneration is slow. 2. Leucocytosis, very roughly parallel to the severity of the disease, unaffected by antitoxin treatment, gradually decreas- 1 Gesellsch. f . innere Med. , Berlin, July 6th, 1896. 202 SPECIAL PATHOLOGY OF THE BLOOD. ing the disease passes off, sometimes absent in very mild or very severe cases. 3. Polymorphonuclear leucocytes much increased during fe- brile stages, often diminished in convalescence. 4. Myelocytes numerous in some severe cases. The blood examination has no diagnostic value so far as I can see ; in prognosis the absence of leucocytosis (except in ob- viously mild cases) and the presence of many myelocytes are ap- parently bad signs. CHAPTEE IV. ACUTE INFECTIOUS DISEASES (CONTINUED). SCARLET FEVER. HEUBNEB* noted hsemoglobinsemia in one case. Fibrin is not increased even at the height of the fever, provided inflam- matory complications are absent. Red Cells. Very little is to be found in literature upon the subject. Kotschetkoff 2 noted a gradual diminution of the red cells to about 3,000,000, regeneration taking place in the course of not less than six weeks. Other observers have found little or no ansemia. Hayem 3 estimates the average loss of red cells at 1,000,000. In mild cases he finds the lowest figures on the first day of normal temperature. In severer cases in which the fever comes down slowly, the red cells may not reach their minimum till twenty-four hours after reaching the normal temperature. Felsenthal 4 in six cases found the count to be 4,500,000 to 5,500,000 no considerable variation from normal. Zappert 5 in six cases found it to be from 3,920,000 to 4,- 500,000, an average of 4,150,000. White Cells. Most observers are agreed that leucocytosis is the rule, contrasting in this respect with measles, in which no leucocytosis occurs. The increase may be present even six days before the rash appears and attains its maximum two or three days after the eruption. In light cases it may sink to normal even before the fever is gone, while in severer cases it may per- sist several days after a normal temperature is reached. Von Limbeck had a case in which the leucocytosis persisted for 1 Dent. Arch. f. klin. Med., vol. 23. 2 Ref. in Petersburg, med. Woch., 1892, 1. 3 Loc. cit., p. 914. 4 Arch. f. Kinderheilk., 1892, p. 80. 6 Zeit. f. klin. Med., 1893, p. 292. 204 SPECIAL PATHOLOGY OF THE BLOOD. twelve days after the temperature had become normal. Forty thousand per cubic millimetre is not unusual in well-marked cases. Rieder's ten cases averaged 17,500; Felsenthal's six counts were between 18,000 and 30,000. My own are simi- lar. In a general way the severest cases are apt to have the high- est leucocyte counts ; the figures have no direct relation to the amount of fever, glandular swelling, or to complications in the ear or kidney. Qualitative Changes. The polymorphonuclear forms are in- creased, often to 90 per cent, soon falling except in the worst cases. The peculiar characteristic of the disease is the persis- tence of eosinophiles in all but the severest cases despite the increase of polymorphonuclear forms. They may run as high as 5 per cent during the fever, and are still more numerous in convalescence, remaining increased for six weeks. According to Kotschetkoff, disappearance of eosinophiles is a bad prog- nostic sign except at the very beginning of the fever, when they may be temporarily absent in favorable cases. Presumably they have some connection with the exanthem, eosinophilia being so common in connection with skin lesions. They may number 15 per cent of the leucocytes in convalescence. Felsenthal's average is 5 per cent; Zappert's, 3 per cent. The lymphocytes are decreased proportionately to the severity of the case, the worst cases showing only 2 to 4 per cent. An increase of eosinophiles during a scarlatinal nephritis is regarded by Neusser and his pupils as a favorable sign, and their absence as ominous. In ordinary cases without nephritis they reach their maximum in the second or third week and are not normal till the sixth. Summary. Moderate anaemia. Leucocytosis beginning before the eruption and often lasting into convalescence. Eosinophiles said to be increased in favorable cases, absent in bad cases. MEASLES. Diagnostic and Prognostic Value. 205 1. The chief importance of the blood examination is in dis- tinguishing the disease from measles and the eruptions of other diseases. Measles has no leucocytosis. 2. Whether the prognostic significance attached by Neusser and others to the percentage of eosinophiles is genuine or not cannot as yet be positively stated. MEASLES. In mild cases the blood shows no changes at all. Where bronchitis, coryza, and conjunctivitis are very marked, fibrin may be increased. Bed Cells. In mild cases no change never over 400,000 or 500,000 red cells are lost (Hayem). Felsenthal's eight cases showed counts of 5,000,000 to 5,500,000. White Cells. In most cases there is no increase. Felsenthal in eight cases found the count normal or diminished. Pee found but 4,000 in a case with a fever of 102.7. Eieder's eight cases averaged 7,500, being lowest at the height of the disease and increasing as fever passed off. Complication with catarrhal pneumonia or a very bad bronchitis and coryza may slightly raise the count. The eosinophiles, contrary to the example of scarlet fever, are often absent during fever. The Massachusetts Hospital records furnish the following counts : TABLE XII. MEASLES. Age. Sex. Red cells. White cells. Per cent haemo- globin. \ Remarks. 38 8 M. M 4,700,000 9,000 9,000 65 " Black measles" petechise. Differential count normal. 9,8 F 8,000 68 104 ; eruption out. 4 10 M. M 5,000,000 7,000 6,000 60 Eruption just out. 103, three days before the erup- 53 33 F. F 6,000 6,000 3,500 3,500 67 tion ; differential count normal. Eruption out one day. Eruption out three days. Felsenthal found the polymorphonuclear cells much in- 206 SPECIAL PATHOLOGY OF THE BLOOD. creased and eosinophiles never over one per cent. In my own cases the differential counts were normal. The value of the blood examination is considerable in excluding scarlet fever, diphtheria, and syphilitic roseola, all of which show leucocyt- osis. It cannot apparently be distinguished by the blood count from Eotlidn (German measles), in two cases of which, seen at the Massachusetts Hospital, the white cells were 6,000 and 8,000 respectively. MUMPS. Five cases of mumps under my care showed no leucocytosis. WHOOPING-COUGH. A girl of six recently seen showed at the height of the dis- ease 12,600 leucocytes with 78 per cent of haemoglobin. SMALL-POX (VARIOLA). Red Cells. According to Hay em no other fever Is so de- structive of red cells. During the fever the count is normal or increased, but when the temperature falls permanently the num- ber of red cells falls suddenly, whether because the blood is diluted (see above, page 158) or by a real destruction. From this time on the cells are slowly regenerated; even at the fifteenth day Hayem found them considerably below normal. In hemorrhagic cases the anaemia comes on more quickly, its degree depending on the amount of hemorrhage. In one case, dying on the seventh day of the eruption, Hayem found but 2,000,000 red cells, in another at the same stage, 4,600,000. Fibrin is not increased until the stage of suppuration is reached. White Corpuscles. Pick, who carefully studied 42 cases, found that the very lightest cases, such as occur in vaccinated persons, may cause no leucocytosis. In a woman of twenty- two on the third day of illness with a temperature of 105, the count was only 4,200 and on the fifth day (temperature 99) 3,600. This patient had been vaccinated. Severe cases if without complication show no leucocytosis till the pus appears in the vesicles, and after this period the leu- cocytosis slowly sinks again. For example, on the fifth day of the illness, leucocytes 4,200; at the beginning of suppuration, ACUTE ARTICULAR RHEUMATISM. 20? 11,600 (eighth day) ; at the height of suppuration (tenth day), 17,200; at the thirteenth day, pustules drying up, leucocytes 7,600. In the severest types, the leucocytes follow about the same course, there being no leucocytosis whatever in the initial or eruptive stages. Only when the infection with pus organisms begins do the leucocytes rise, the poison of variola itself having apparently no tendency to increase the count. The amount and duration of the increase at the stage of suppuration is in a gen- eral way proportional to the severity of the case. Widal ' has recently found virulent streptococci in the blood in six cases of variola. VARICELLA (CHICKEN-POX). The only observation of which I am aware is that reported by Engel. 2 In a child of five he found during the height of the pustular stage a moderate leucocytosis, with 67 per cent of neutrophiles (high for a young child), and no eosinophiles. Three days later as the pustules were healing the neutrophiles had sunk to 47 per cent (normal for that age) and the ecsino- philes had risen to 16 per cent. The same conditions obtain after vaccination. ACUTE ARTICULAR RHEUMATISM. According to Hayem and Garrod 3 the blood constitutes as in syphilis a most valuable measure of the intensity of the sickness, which is parallel to the severity of the blood-changes rather than to the number of joints affected. The fever, the intensity of the lesions, and the state of the blood run parallel, in a gen- eral way, but the degree of anaemia is a more delicate index of the patient's condition than even the temperature chart (Garrod). The Blood as a Whole. Fibrin is greatly increased. In no other disease except in pneumonia is the network thicker or more rapid in formation. According to Maclagan, this is to be explained by an increase of 1 Widal : Centralb. fur. allg. Path. , etc. , 1896, p. 569. 2 15th Cong, fur innere Med., 1897. 3 British Medical Journal, May 28th, 1892. 208 SPECIAL PATHOLOGY OF THE BLOOD. tissue metamorphosis. Coagulation, on the other hand, is not quicker but slower than usual. Lactic acid is present in excess, but cannot be clinically es- timated, nor is its excess peculiar to this disease. The alkalinity of the blood had been reported diminished, but the technique is not considered reliable by the best observers. Red Cells. Hay em ' and Osier 2 state that the poison of acute rheumatism is a powerful and rapid destroyer of red cells. In acute cases, according to Hayem, the red cells lose at least 1,000,000 of their number and in cases which drag along and re- lapse the loss is from 1,500,000 to 2,000,000. When an attack is cut short by salicylate treatment the drain on the corpuscles is stopped. So far as can be judged from the figures in Table XIII. of the Masachusetts Hospital cases this diminution does not seem to occur in all cases. Many of these cases had been sick some weeks before the time when the count was made, yet the counts are not very low. In the eight cases which have been sick over twenty days, the average of red cells is 4,462,000; in those sick between one and twenty days, 4,540,000; and in the whole group of cases, 4,400,000. The lowest count was 3,608,000. Accord- ing to Haj^em 4,000,000 is the usual count in acute cases and 3,000,000 to 3,500,000 in those which drag on and relapse. Qualitative Changes. Maragliano's so-called degenerative changes in the red cells have been observed in this disease, but are not very marked. Deformities and nucleated corpuscles appear only when the anaemia is very marked. Haemoglobin. As in all secondary anaemias the corpuscles get thin and pale before they die, and hence the coloring matter is diminished more than the count. The average haemoglobin percentage in this series is sixty -seven, and the color index .76. Hayem noted that, in some cases during convalescence, as the red corpuscles slowly increase tfae color index remains low or even goes lower still. Leucocytes. All observers agree that leucocy tosis is the rule and that its degree is roughly parallel to the acuteness and severity of the attack (the individual's vigor of reaction is al- ways a factor) and the amount of fever. The following tables illustrate the variations of the leucocytes in a fairly typical way : 1 Loc. cit., p. 917 "Practice of Medicine," 1895. ACUTE ARTICULAR RHEUMATISM. 209 TABLE XIII., A. ACUTE ARTICULAR RHEUMATISM. 1 c fc Age. W % Duration. Degree of inflammation. Red cells White cells. Per cen haemo- globin. Remarks. - 50 F 17 days. Red and hot. 9 39,000 65 j 21 M. 5 weeks. ? 4,160',000 31,500 65 Knees and 1 ankle. 59 F. ? 9 5,476,000 27,000 94 Patient pale. ^ Adult 31. 9 9 9 25,900 ? { 33 31. 2 weeks. Red and hot. 4,852,000 24,500 76 b 20 F. ? ? ? 22,400 ? Acute endocarditis also. J Adult 31. ? ? 4,216.000 21,000 56 8 23 31. 4 weeks. Tender and hot. 5,192,000 18,300 70 Temperature 102. 9 10 28 19 31. 31. 3 " 4 days. ? Red and hot. 9 ? 17,800 17,400 ? ? Many joints affected. 11 49 31. ? ? 4,800,000 17,700 ? Paronychia also. 12 49 31. ? ? 9 17,100 ? Dec. 2d. 13 21 F. ? Red and hot. 3,944^000 17.000 45 Cheeks rosy. 14 24 31. 2 days. ? 4,600.000 16,000 68 18 24 31. ? ? 4,670,000 15,500 68 16 35 31 15,200 45 17 18 13 12 F. 31. 1 day.' 2 weeks. Red and hot. 4,880,000 4,400,000 15,200 15,000 65 56 Temperature 102. 19 19 31. 4 days. " " 4,760,000 14,500 75 Severe case. 20 9 F. ? v 4,240,000 14,386 - 60 81 9 F. " Red and hot. 9 14,050 9 22 47 31. Iday. 4,750',000 14,000 72 One joint only af- fected. as 25 F. 3 days. Tender and hot. 4.850,000 14,000 75 24 18 F. 2 months No redness or 4,156,000 14,000 54 heat. 25 19 31. ? 9 4,172,000 14,000 70 as 19 31. ? ? 4,580,000 13,500 64 Nov. 10th, 1895. 27 21 F, Iday. Red and hot. 9 13,500 ? XJK 29 M. V ? 4,320^000 12,750 68 29 9 ? ? ? 4,128,000 12,650 65 Dec. 1st, 1895. 30 87 F. 1 mouth. Swollen, tender. 5,320,000 12,500 64 SI 28 31. ? " u 5,000,000 12,500 65 33 32 31. ? ? 9 12,100 ? Purpura also. 3S 30 F. 9 9 4,160',0 12000 9 31 35 47 27 31. F. 4 weeks. 3 days. Very slight. 4,288,000 3,880,000 12.000 11,600 65 65 3<; 17 3[. 1 week. " " " 4,600,000 11,500 70 Mild case. 37 27 31. 10 days. Hot and red. 4,200,000 11,500 60 ss 33 31. 4 weeks. ? 5,480,000 11,000 80 Hands alone i n - volved. 39 18 31. ? Not red and hot. ? 10,000 .... One joint only af- fected. 40 28 31. 3 weeks. u u u 3,608.000 7,000 40 41 Adult. 31. ? 9 3,768,000 6,800 42 43 29 30 31. F. 9 weeks. ? Some joints hot. ? 4,104,000 3,440,000 5,500 4,700 '58 26 Tourth relapse. Specific gravity 1040. Average 4,400,000+ 6,800+ 67 TABLE XIII., B. SUBACUTE ARTICULAR RHEUMATISM. No Age. Sex. Red cells. White cells. Per cent haemoglobin. 1 2 3 4 5 6 25 30 28 28 Adult. M. F. F. F. M. F. 4,750,000 4,644.000 9 4,684,000 4,016,000 4,188,000 15,000 13,000 10,600 8,000 6.200 5,750 60 63 ? 75 41 73 Ave rage = 4,400,000 9,760 62 14 210 SPECIAL PATHOLOGY OF THE BLOOD. TABLE XIII. , C. CHRONIC RHEUMATISM, CHIEFLY ARTICULAR. No. 1 2 3 4 5 6 Age. Sex. Red cells. White cells. Per cent hemoglobin. 78 19 32 58 30 20 F. F. F. M. F. M. 9 5,248,000 ? 4,744,000 5,576,000 7,200 8,300 6,400 6,500 6,100 9,800 ? 45 ? 60 ? 62 Average = 7,400 TABLE XIII., D. MUSCULAR RHEUMATISM. No. 1 2 3 4 5 6 Age. Sex. Red cells. White cells. Per cent haemoglobin. Remarks. 46 54 38 54 27 35 M. M. M. M. F. M. 4,580.000 4,360,000 ? 3,820,000 ? 7,500 7,500 6,600 14,000 6,000 5,700 70 75 ? 58 ? During febrile attacks. Lumbago. Average = 7,500+ The average leucocytosis in the acute cases is 16,800 ; in those mild and more chronic, so-called " subacute" cases the leuco- cytes range lower, averaging 9,760; while in chronic rheuma- tism, whether articular or muscular (including lumbago), there is no increase at all (average =7,450). In five cases of arthritis deformans treated at the Massa- chusetts Hospital the blood was normal except for a slight de- ficiency of haemoglobin in two cases. Summary. Anaemia with leucocytosis, the degree of which is a measure of the severity of the infection. Fibrin much increased. Diagnostic Value. The blood tells us little if anything that could not be learned in other ways. It does not differ at all from that of a septic arthritis, or from that of acute gonorrhoeal arthritis. The only cases that I remember in which a blood examina- tion has been valuable are the following : ASIATIC CHOLERA. 211 CASE I. The patient had muscular pains, fever, and a his- tory of a malarial attack some months earlier. The question to be decided by the blood examination was between malaria and " rheumatism." The leucocytes were 23,600 per cubic milli- metre, which made it clear that the case was neither malaria nor "rheumatism," since the former never increases the leuco- cytes and the latter could only give so high a count in case genuine articular inflammation were present. The case turned out to be croupous pneumonia which the high leucocyte count strongly suggested. CASE II. Patient presented symptoms and signs of acute polyarticular rheumatism with fever. The fever came down under salicylates, but soon rose again, and the man became wildly delirious. His delirium persisted after the salicylate was stopped. Several joints continued swollen and tender. The fever was very moderate, ranging between 99 and 101. There were no rose spots and no spleen. The question arose as to whether it was a case of sepsis with localization in the joints, or whether it was a case of typhoid supervening on an arthritis of some kind. The blood count, which was repeated several times, always showed a perfectly normal blood except for a slight an- aemia. The subsequent course of the case, during which he re- mained for nearly three weeks more or less delirious, made it clear to Dr. F. C. Shattuck, under whose care the patient was, that the diagnosis was typhoid. Chronic rheumatism (muscular or articular) produces no constant blood changes appreciable by clinical methods (see Table XIII. , C and D). ASIATIC CHOLERA. In no other disease so far as I am aware has an acid reaction in the blood been reported. This is at the end of life. All observers agree that the alkalinity is at least greatly reduced. Our knowledge of the corpuscles is best summed up in Biernacki's ! study of thirty -eight cases. Red Cells. In the stadium algidum, or stage of collapse, most of the symptoms are due to the great concentration of the blood from the loss of serous fluid in the stools. Hay em found the increase of red cells from this concentration to amount to from 1,000,000 to 1,500,000 per cubic millimetre. Biernacki ' found 7,662,500 in one case twenty-four hours 'Deut. med. Woch., 1895, No. 48. 212 SPECIAL PATHOLOGY OF THE BLOOD. after the beginning of the disease. The specific gravity may be as high as 1071 or 1072. White Cells. Leucocytosis is present, not merely as the result of concentration, but as a genuine increase to at least double the normal count. Biernacki found that cases with par- ticularly high counts (40,000 to 60,000) were soon fatal, so that he considers a marked leucocytosis in the algid stage as a bad prognostic sign, although patients also die with low leucocyte counts in this period. Such a leucocytosis does not occur in ordinary diarrhoea or dysentery. Leucocytosis is present as early as twelve hours from the first symptom and lasts at least as late as the sixth day. In the stage of reaction it usually decreases. In one very mild case reported by Biernacki there was not only no increase, but leu- copenia (4,375 per cubic millimetre). The differential count shows from eighty -two to ninety-five per cent of adult cells and a corresponding diminution of the young forms. ERYSIPELAS. Halla, Pee, Beinert, Bieder, and v. Limbeck agree that leu- cocytosis is usually present in well-marked cases. Yon Lim- beck finds the "leucocyte curve" to run roughly parallel with the temperature chart, sometimes beginning to fall a little before the latter. The counts rarely run very high, yet Beinert counted 39,627 in one case. Pee noted that the leucocyte count increases only while the process is spreading and that the size of the count was a tolerably accurate measure of the severity of the case. Bieder found in seven cases an average of only 15,000 per cubic millimetre despite very high temperatures. In one case the leucocyte count remained high after the temperature had fallen, but in the others it anticipated the temperature. In one mild case he found no leucocytosis. Polymorphonuclear cells are greatly increased as in other forms of leucocytosis. Hay em noticed the same dependence of the leucocyte count upon the severity of the process. . In six cases at the Massachusetts General Hospital I found 17,000, 14,000, 13,000, 12,700, 7,250, and 6,200, the last two TONSILLITIS (FOLLICULAR). 213 very mild cases. The count of leucocytes seemed proportional to the severity of the affection. When the disease occurred in "scrofulous" cases, Hayem found only 7,000-8,000 leucocytes per cubic millimetre, while in cases with very extensive process and high fever 12,000-20,000 were present. He found also a loss of 500,000-1,000,000 in the count of the red cells, according to the severity of the case. This showed itself particularly just before the fall of the tempera- ture. I have seen no reference by other writers to the condition of the red cells in this affection. TONSILLITIS (FOLLICULAR). Halla, 1 Pick, 2 and Pee 3 found leucocytosis as a rule in un- complicated follicular tonsillitis; Rieder found it in a case complicated with acute nephritis. The following table confirms these observations in the main, though in mild cases no leucocytosis was present. TABLE XIV. TONSILLITIS. o" Age. 1 Red cells. White cells. Per cent haemo- globin. Remarks. 1 22 F. 4,368,000 19,200 35 2 21 F. 18,000 8 27 M. 18,000 4 21 F. 16 800 5 25 F. 16,200 6 30 F. 4, 750, 000 16,000 80 Temperature 101. r? i 27 M. 4,556,000 15,500 67 Six days ; slight. 8 Adult. F. 4,860,000 14,000 Follicular. 1) 30 M. 4,730,000 13,500 76 Convalescent. 10 24 F. 5,000,000 13,500 68 Follicular. 11 Adult. M. 13,500 12 M. 4,952,000 12,250 94 18 '24' F. 5,816,000 11,900 65 Streptococcus ; slight ar- ticular rheumatism. 14 M. 5,000,000 11,800 90 Follicular. 15 19 F. 4,552,000 11,600 52 16 18 M. 5,150,000 11,500 83 Chronic recurrent; out in two days. 17 22 F. 5,016,000 9,600 IS Adult. F. 4,200,000 5,800 60 Follicular. 19 23 F. 7,925 52 Follicular ; slight ; t e m - perature 99 next day. 20 45 F. 6,800 1 Zeitschrift f. Heilkunde, 1883, p. 198. 2 Prag. med. Woch., 1890, p. 303. 3 Pee: Inaug. Dissert., Berlin, 1890, p. 8. 214 SPECIAL PATHOLOGY OF THE BLOOD. The blood examination has no diagnostic value so far as I am aware. It is worth knowing that a simple tonsillitis can cause leucocy tosis, to the end that if such is discovered on blood examination we need not suppose that some other process is present to account for the increase. GRIPPE. The references in literature to the blood of grippe are very scanty. Orion (Archiv. d. Med. milit., 1890, p. 280) found fibrin increased during the early days of the disease. Eieder (Munch, med. Wocli., 1892, XXXIX.) found no leucocy tosis in grippe and but little in the " catarrhal pneumonia" following it. The following table shows that the leucocytes are normal in at least five-sixths of the cases. Only eleven of the sixty-seven cases showed leucocytosis, and in one or more of these some complication was very possibly present. This is of importance in excluding pneumonia and focal inflammatory conditions. The leucocyte count does not help us to distinguish the disease from typlioid. In this decision the serum reaction is our mainstay (see below, page 400). From malaria it may be distinguished by the absence of malarial organisms. In one case after an operation for traumatic epilepsy, the temperature rose to 104, with a chill, and the question of meningitis was considered. The absence of leucocytosis excluded the meningitis, and the at- tack turned out to be grippe, which was just then very prevalent. TABLE XV. GRIPPE. White cells. Between 8,000 and 3,000 lease. 3,000 " 4,000= 3 cases. 4,000 " 5,000= 6 " 5,000 " 6,000= 5 " 6,000 " 7,000= 14 " 7,000 " 8,000= 5 " 8,000 " 9,000= 6 " 9,000 " 10,000= 9 " 10,000 " 11,000= 7 " 11,000 " 12,000= 3 " 12,000 " 14,000= 8 " . 67 " SEPTIC^MIA. 215 TABLE XV. GRIPPE. Red cells. Between 3, 000, 000 and 4, 000, 000 = 2 cases. 4,000,000 " 5,000,000= 9 " 5,000,000 " 6,000,000= 14 " 25 " SEPTICAEMIA. Puerperal septicaemia, infected wounds, septic arthritis, septic endocarditis, general infections with pyogenic bacteria, "pyaemia," are all identical so far as their effects on the blood are concerned, and will be considered together under the general head of Septicaemia. Bacteriology of the Blood. Cocci can be demonstrated in cultures from the blood of sep- ticaemia more frequently than in any other class of infections. Rosenbach 1 in 1884 found streptococci and staphylococci in sepsis. Garre 2 in 1885 found the last-named coccus in a case of osteomyelitis. In 1890 v. Eiselsberg 3 found staphylococci in ten cases of septic wounds and one case of osteomyelitis, and streptococci and staphylococci together in five more cases whose wounds had become septic. Czerniewsky, 4 Stern and Hirschler 5 found the same organ- isms in puerperal fever, the former observer in five cases. Brunner, 6 Hoff, 7 and Blum 8 found pyogenic staphylococci in pyaemia and sepsis, and Saenger, 9 Eoux and Lannois, 10 Cantu/ 1 and Bommers 12 had equal success, each in a single case. 1 " Microorganismen b. d. \Vundinfectionskrankheiten," pto. . Wies- baden, 1884. 2 Fortsch. d. Med., 1885, No. 6. 3 Wien. klin. Woch., 1890, No. 30. 4 Archiv f. Gynakol., 1888, No. 33. 5 Wien. med. Presse, 1888, No. 28. 6 Wien. klin. Woch., 1891, No. 20. 7 Dissert, Strassburg, 1890. 8 Munch, med. Woch., 1893, No. 16. 9 Deut. med. Woch., 1889, No. 8. 10 Revue de Med., 1890, No. 12. 11 Rif. Med., 1892, No. 96. 12 Deut. med. Woch. , 1893, No. 16. 216 SPECIAL PATHOLOGY OF THE BLOOD. Canon ' and Sittman 2 investigated large numbers of cases with many positive results, and Grawitz 3 and Petruschky 4 and Cohn 5 were successful in finding pyogenic cocci in the blood of cases of ulcerative endocarditis as well as in other septic infec- tions. Herschlaff 1B found them in erysipelas, acute tuberculosis, perforated typhoid ulcer, etc. Kiihnan, ly on the other hand, was unable to find anything in the blood of twenty-three severe pyaemic cases, and was successful in only one out of twelve cases of ulcerative endocarditis. Taking the results of all these investigators together, it seems evident that in many cases of septicaemia, blood cul- tures, taken according to the directions on page 35, show the presence of pyogenic organisms, and that in many obscure septic cases the 'diagnosis may be greatly facilitated by such an exam- ination. Negative results are of course very far from excluding septicaemia, but positive ones are sometimes of great value if proper precautions are taken in the technique of the examina- tion. In the diagnosis of malignant endocarditis, often a most difficult one, Grawitz thinks blood cultures are especially im- portant and likely to prove positive when the disease is present (see Diseases of the Heart, page 293) . Almost all observers agree that the finding of pyogenic cocci (except the staphylococcus albus) in the blood makes the prog- nosis almost surely fatal. The toxicity of the blood is doubled. Bed Cells. All observers agree that very marked anaemia is present in severe cases. Roscher's 1 investigations tend to show that the diminution in red cells in septicaemia is greater than in any other infective disease, and appears in a shorter time. He found such a diminution present no longer than a few hours from the beginning of the illness. He finds the amount of anaemia proportional to the severity of the case, and (reckoning by means of the estimated solid residue) concludes that whenever the blood .has lost one-quarter of its substance or more, death follows. 'Deut. Zeit. f. Chirurg., 1893, p. 571. 2 Deut, Arch. f. klin. Med., 1894, p. 573. 3 Charite-Annalen, 1804, vol. 10. 4 Zeit. f. Hygiene, 1894, pp. 59 and 413. 5 Deut. med. Woch., 1897, No. 9. 6 Sem. Med., 1897, p. 105. 7 Deut. med. Woch. , 1897, No. 25. 8 Inaug. Dissert, Berlin, 1894. SEPTICAEMIA. 217 He considers, therefore, that help as to prognosis is given us by the blood examination in septicaemia. The serum becomes very watery, partaking of the general atrophy of the blood tissue. In a case of intensely acute puer- peral sepsis Grawitz found the red cells reduced to 300,000 ( ! ) although the patient had been sick less than twenty-four hours. The case seems almost incredible, but is reported in great detail in the author's recent text-book, to which reference has so frequently been made. He accounts for it by the combi- nation of blood destruction and dilution. In the nine cases of puerperal sepsis seen at the Massa- chusetts General Hospital in recent years the red cells averaged 3,780,000, which is very low, considering the shortness of the illness in most cases. (The influence of hemorrhage during parturition must of course be taken into account.) In most of the septic wounds which I have seen the counts have not been low. But in one case of septicaemia from a sup- purating fibroid of the uterus the red cells numbered only 1,800,000. In a case of puerperal sepsis of only a few days' duration, in a woman not previously anaemic, Hayem ' recently reports the following figures : December 3d Red cells 1, 450, 000 White cells 7, 500 Haemoglobin 20 per cent. December 6th Red cells 2, 578, 000 White cells 8,000 Haemoglobin 40 per cent. December 24th Red cells 4, 231, 000 White cells 7,200 Haemoglobin 65 per cento (Recovery. ) Such cases are the best examples we have of an acute ancemia (hemorrhage excepted) . The hemoglobin is usually diminished about as much as the corpuscles. According to Bond it tends to crystallize about the edge of a slide and cover-glass preparation of the fresh-blood. Deformities in the shape and size of the corpuscles are not usually present except in the severest cases. 1 La Med. Moderne, January 13th, 1897. 218 SPECIAL PATHOLOGY OF THE BLOOD. TABLE XVI. PUERPERAL SEPTICJEMIA. d fc Age. Sex. Red cells. White cells. Per cent haemo- globin. Remarks. 1 31 F. 77, 500 Autopsy. 2 21 F. 2,300,000 26,000 3 29 F. 3,900,000 23,900 68 * Two days before delivery. 21,000 . Day of delivery. 9,500 One day after delivery. 15,500 . . Five days after delivery; breasts caked. 15,000 Ten days after delivery. 4 28 F. 3,784,000 11,800 22,000 55 Twenty-six days after delivery. Miscarriage five days before ; septic ; curetted. 13,600 Three days later, temperature falling. 8,300 Seven days later, temperature normal. 15,800 Fourteen days later, tempera- ture up ; curetted again. 14,900 Fifteen days later, temperature falling. 15,000 Sixteen days later, temperature falling. 9,500 Thirty-two days later, temper ature falling. 5 25 F 20 800 55 6 34 F. 15,900 September 2d, 1897. 35,600 September 9th, chills. 33,000 September 13th. 35, 600 . . September 16th. Recovered. 7 25 F. 2,936,000 20,000 50 Ar*il 1st, 1894. 21,000 April 3d, 1894. 8 32 F. 4,904,000 19,300 Curetted. 9,300 One week later, well. 9 24 F. 3,556,000 18,400 10 F. Marked Polymorphonuclear cells, 94$ ; increase. lymphocytes, 6$. 11 26 F. 5,368,000 5,600 Died. TABLE XVII., A. SEPTIC WOUNDS. "A Age. 1 Red rplls. White cells. Per cent haSino- globin. Remarks. 1 fl 37 28 F. M 5,880,000 7 600 000 48,400 25 400 Sloughing breast ; bedsore. Septic wound of foot. 3 31 F 5,680 000 15 300 i Sloughing breast after cancer 6,700 operation. 5 840 000 23 200 One month later wound clean. 4 27 M 4 450 000 10 500 Septic hand. 5 M 5,600 000 8 800 Septic finger SEPTIC^MIA. 219 TABLE XVII., B. SEPTICJEMIA WITH ARTHRITIS. Age. M > 02 Red cells. White cells. Per cent haemo- globin. Remarks. 8 M 25 000 Pus in 6lbow "joint * no in "jury 43,000 Two days after operation, vent 24, 000 not free ; opened further. Seven days after operation 20, 700 Eight days after operation. 6,700 Sixteen days after well. 34 59 M. M 4,520,000 19,000 18,500 65 Gonorrhoeal, pus in knee. Pus in shoulder joint, no trauma* 22 M 13,800 Gonorrhoeal ankle 39 M. 8,940 Gonorrhoeal ankle ; cultures neg- ative. TABLE XVII., C. GENERAL STREPTOCOCCUS SEPTICAEMIA. Age. 1 Red cells. White cells. Per cent haemo- globin. Remarks. Adult, u M. F. 5,248,000 1,800,000 41,400 46,000 Suppurating fibroid. 22 F. 3,776,000 25,800 52 A fatal case, yet no fever ! Hcemoglobincemia with reddish staining of the serum is often noticeable in the dried and stained cover-glass specimen where the plasma is deeply stained. Leucocytes. Considerable controversy has taken place as to the changes in the white cells effected by septicaemia ; some ob- servers finding leucocytosis, while others find none. The results of experimental infections referred to above (see page 110) and the parallelism of the leucocyte changes in pneu- monia, peritonitis, and diphtheria fully explain these appar- ent divergences, which perfectly exemplify the rules stated on page 106. Leucocytosis occurs only when the struggle between the pa- tient and his disease is intense, and whichever is victorious. When either side wins without any difficulty, i.e., in the mildest and in the severest cases, leucocytosis is nearly or entirely ab- sent ; indeed, leucopenia may be found (as for instance in a case 320 SPECIAL PATHOLOGY OF THE BLOOD. of septic endometritis reported by v. Limbeck only 3,000 leu- cocytes). Von Limbeck and Krebs l found no leucocytosis in cases of perpetual septicaemia, but these were all fatal cases or very mild ones. Bieder, on the other hand, and the great major- ity of other observers (Sadler, 2 Boscher, 3 Kanthak, 4 Grawitz, etc.) find leucocytosis. This means that in most cases ob- served by these writers the infection was of moderate severity. Only two of the twenty-one cases in Tables XVI. and XVII. showed no leucocytosis. One was very mild, the other died on the day of the count. Summary. 1. Bapid development of severe anaemia. 2. Leucocytosis marked, except in very mild or very severe cases. 3. Blood cultures often contain pyogenic cocci. Diagnostic Value. The advantage of a positive bacteriological examination is obvious. Of the value of the blood count in distinguishing septic from non-septic wounds and estimating the degree of sepsis and the importance or needlessness of operative interference, not much is known. The subject deserves to be carefully worked out from a surgical point of view. The following cases, how- ever, tend to show that we might utilize blood counting far more than we do to determine questions of this sort : CASE I. Frank B was a case of appendicitis operated on by Dr. M. H. Bichardson at the end of an attack. A little pus was found, the appendix was excised, and. the wound nearly closed, a small strand of gauze, however, being left in. Several days after the operation, there being at the time no external dis- charge, the temperature rose. The wound seemed perfectly clean. The man was very nervous about himself, and much 1 Krebs : Dissert. , Berlin, 1893. 2 Sadler : Lot. tit. 8 Roscher: Dissert., Berlin, 1894. * Kanthak : Brit. Med. Journal, June, 1892. SEPTIC^MIA. 221 stirred up at each dressing ; and as the temperature never went higher than 101, there seemed to be considerable doubt as to what the cause of the temperature was. The blood count in this case showed 52,000 leucocytes. On opening the wound a large amount of broken-down blood clot was evacuated, and the temperature came down to normal. CASE II. Mrs. S was a case of pus tube shelled out and sewed up tight. Ten days after the operation the temperature began to look: as if pus were present. Here again the patient was exceedingly nervous ; and, as so often happens, the question was asked and re-asked, whether she was keeping up her own temperature by the state of her mind. The blood count, how- ever, showed marked leucocytosis, which led to a careful ether examination, revealing a fluctuating mass behind the uterus, from which pus was obtained by puncture. CASE III. Mr. R entered the Massachusetts General Hospital in December, under the service of Dr. C. B. Porter, with a compound fracture of the thigh. Some days after it had been put up, the temperature began to suggest the presence of pus, the wound, however, remaining perfectly clean. I counted the blood, and found a marked leucocytosis. A more thorough exploration of the wound revealed a pocket of pus, the evacu- ation of which brought down the temperature. I was not sure in this case whether the absorption of the blood clot, such as takes place, I suppose, after any compound fracture, would be sufficient to cause leucocytosis. I therefore counted several cases in which there was fever and presumably blood-clot ab- sorption, namely, a hsemothorax, a pelvic hsernatocele, two com- pound fractures, and a crushed foot ; in none of these was any leucocytosis present. CASE IV. Mr. S was operated on by Dr. J. C. Warren for traumatic epilepsy. Nothing special was found, and the wound was closed. Ten days after the operation the temperature rose to 104, and the patient complained of severe headache and pain in the back. I counted the blood, and found no leucocytosis. Next day the temperature was down. The patient apparently had the grippe. Several cases in which an old malaria was supposed to be " brought out" by a surgical operation, the patient having irreg- ular fever and chills after the operation, have shown, on exam- ination of the blood by the writer, no malarial organisms but marked leucocytosis. In these cases the symptoms of " malaria" ceased when the wound was more thoroughly drained, and I have no doubt that many cases of " malaria" after surgical oper- ations are really wound sepsis. 222 SPECIAL PATHOLOGY OF THE BLOOD. It is difficult to make inferences from a leucocytosis in such albuminoids are drained out of it, leaving it watery and poor in corpuscles. A patient of Grawitz after years of chronic dysen- INTESTINAL OBSTRUCTION. 283 tery had but 1,880,000 red cells per cubic millimetre, while the serum had twice the normal amount of water and half the normal amount of solids. I have seen the count fall as low as 1,928,000 in a case of prolonged colitis, with final recovery. In another case the red cells reached no lower than 2,440,000, but the haemoglobin was but ten per cent. A differential count of this man's blood showed the following: Polymorphonuclear neutrophiles 66.3 per cent. Lymphocytes (small) 24.9 " Lymphocytes (large) ... 6.0 Eosinophiles 1.4 " Myelocytes , 1.4 " While counting 400 leucocytes I saw 8 normoblasts and 5 megaloblasts. The total leucocyte count was 9,800 per cubic millimetre. Cases 1, 3, 4, 12, and 14 of the series in Table XXVIII. show similar conditions. The haemoglobin, however, usually suffers most, and the color index is low. Leucocytosis is rare, but does occasionally occur, possibly owing to some complication or auto-intoxication. TABLE XXIX. INTESTINAL OBSTRUCTION. Age. g 02 Red cells. White cells. Per cent haemo- globin. Remarks. 3 120 000 20 800 Cancer 52 M 5 568 000 18 860 9th of May cancer Adult. M 18,800 14,666 17th of May, cancer. No fseces three days. 12 400 No urine two days. One day later no fseces * urine 4 100 drawn by catheter. Three days later bowels moved 35 M 3, 504 000 12,000 six times. Chronic obstruction with hemor 21 M 5 150 000 12 000 rhage. Obstruction (hv *\ h^ncH 56 57 Adult. 72 Adult. M F. F. M. M. M. M. 4,440,000 4,272,000 5,800,000 4,850,000 5,200,000 5,540,000 12,000 11,000 6,800 6,000 4,000 4,000 52 75 Cancer. Cancer. Cholera is discussed on page 211. For appendicitis see Abscess, page 222. 284 SPECIAL PATHOLOGY OF THE BLOOD. INTESTINAL OBSTRUCTION. The only point brought out by Table XXIX. is that the white cells may be increased, especially where the obstruction is cancerous. Hence the blood count cannot be relied on to help us in the diagnosis between obstruction and peritonitis. It is more likely that the examination of the amount of fibrin will be useful, as it is said to be increased in peritonitis and not in ob- struction. DISEASES OF THE LIVER. CATARRHAL JAUNDICE. The serum is colored yellow or greenish-yellow and contains bile pigments in solution. It has been asserted that jaun- dice can be recognized here before it shows in the skin or urine. In mild cases, i.e., where some bile goes to the intes- tine and the obstruction is not long standing, the blood is practically normal, as the cases in Table XXX. show. No TABLE XXX. CATARRHAL JAUNDICE. Age. Sex. Red cells. White cells. Per cent haemoglobin. 21 M. 10.500 81 44 M. 10,200 25 30 42 26 21 30 F. F. M. M. M. M. 4,310,000 2,896,666 4,800.666 10, 000 10,000 8,000 9,600 8,775 7,500 7,500 7,400 77 90 68 47. Alcoholic gastritis. 65 64 30 53 29 M. M. M 4,240,666 7.300 6,793 6,200 79 82 35 29 19 M. M. F. F 4,996,000 4,350,000 6,000 4,900 4,200 9,600 4,000 78 85 85 one of these cases shows any leucocytosis, and the red cells and haemoglobin have not suffered except in the alcoholic case in which other causes for anaemia were present. This is con- DISEASES OF THE LIVER. 285 trary to the observations of Grawitz, who found constantly leucocytosis, but agrees with those of v. Limbeck and Hay em, who never found any increase of leucocytes or any other changes in the blood count. Coagulation and the amount of fibrin are normal. Yon Limbeck noticed an increased resistance of the red cells to the influence of distilled water and dilute saline solu- tions which in normal blood dissolve the haemoglobin. He noticed also that the size of the red corpuscles was greater than normal, their volume in a given amount of blood being seventy- seven to eighty-one per cent (i.e., they take up seventy -seven to eighty-one per cent of the room occupied by the drop) while the normal is about forty-four per cent. This, was in cases with only from 4,000,000 to 5,200,000 red cells per cubic millimetre, so that it was evidently due not to an overcrowding of the drop with red cells but to a true increase of size in the individual cells. The same fact has been attested from a different point of view by the investigations of v. Noorden, who found the solid residue increased, and of Hammerschlag ; and Grawitz has noted an increase in the specific gravity of the whole blood, though that of the serum remained normal. Normal red corpuscles put into the serum of icteric patients increase their diameter con- siderably, so that apparently the serum is responsible for the change. Qualitative Changes. Grawitz noted in severe cases that crenation took place much more rapidly than usual in freshly drawn blood, and that the rouleaux formation did not take place. This latter point was also noticed by Hofmeier ' in icterus of the new-born. Silbermann 2 noticed in the same disease great deformities in the size and shape of the cells. In severe febrile icterus Weintraud noted in the red cells the white spots and streaks with active (molec- ular) movements described by Maragliano (see page 87) as en- doglobular degenerative changes. Summary. Normal blood, except for increased size of the red cells and some degenerative changes in severe cases. 1 "Die Gelbsucht der Neugeborenen," Stuttgart, 1882. 2 "Die Gelbsucht der Neugeborenen." Arch. f. Kinderheilk. , 1887, p. 401. 286 SPECIAL PATHOLOGY OF THE BLOOD. Diagnostic Value. The constant presence of leucocytosis excludes an uncom- plicated " catarrhal" jaundice, and points to the probability of malignant disease or inflammation (cholangitis, abscess). Syph- ilis and cirrhosis of the liver might show the same condition of the blood unless the characteristics of syphilitic blood were very marked (see page 269). From a severe cholaemia the absence of any marked anaemia distinguishes a purely catarrhal case. (For the changes in cholsemia see page 290.) CIRRHOSIS OF THE LIVER. 1. ORDINARY (ATROPHIC) CIRRHOSIS WITHOUT JAUNDICE. In the early stages (according to Hay em) neither the red cells nor the haemoglobin fall considerably. Most other observers (perhaps thinking chiefly of the later stages) report marked anaemia. Wlajew 1 counted from 3,000,000 to 4,000,000 red cells ; v. Limbeck had a case with only 1,500,000. He noted that the count might be increased after a tapping in cases with ascites, owing to the concentration of the blood from the rapid refilling of the belly with serum. Grawitz, on the other hand, noticed precisely the opposite effect in a case whose blood before tap- ping had been concentrated by cyanosis, the heart's action being embarrassed by the ascites. After tapping, when the heart's action had become easier and stronger, the cyanosis disappeared and the blood count fell from 4,700,000 to 4,300,000. In v. Lim- beck's case it rose from 4,680,000 to 5,160,000. The moral is that we should draw no inferences from the count of red cells soon after a tapping. The fifty -two cases in Table XXXI., A., were all advanced and their red cells averaged only 3,580,000 + per cubic milli- metre. They steadily decrease as the disease progresses, one case getting as low as 1,300,000; but the anaemia may be con- cealed by cyanosis and concentration. Qualitative Changes. Hay em noticed a curious stickiness of the red corpuscles, a great tendency to adhere to each other. Yon Limbeck looked 1 Ref. in Petersburger med. Woch.. 1894. No. 43 CIRRHOSIS OF THE LIVER. 28-7 for it, but could never find it. Hay em and Maragliano noticed degenerative endoglobular changes in the red cells (" etat cribri- forme"). TABLE XXXI., A. CIRRHOTIC LIVER WITHOUT JAUNDICE. Age. Sex. Red cells. White cells. Per cent, haemo- globin. Remarks. 53 F. 2,950,000 16,000 Recent hemorrhage. 41 M. 4,300,000 12,750 55 Liver enlarged ; ascites. 48 M. 4,992,000 9,000 62 Recent hemorrhage. 53 M. 2,120,000 9,000 23 March 15th. 1,300,000 7,500 22 April 8th. 15 April 18th. 15 April 29th. 2,350,000 6,000 20 May 10th. 2,375.000 5,300 26 May 12th. 2,450,000 5,200 20 June 10th. 4, 500, 000 7,800 25 June 16th. 50 M. 3,440,000 8,320 46 Liver enlarged. 34 M. 7,900 95 56 M. 7,500 68 53 M. 6, 200 60 38 F. 5,700 65 5,720,000 5,200 46 Differential count normal. 54 M. 5,400 64 56 M. 4,680,000 5,000 48 Liver atrophic, July 12th. 4,312,000 4,000 62 Julv 25th. 42 M. 2,920,000 4,500 56 October 30th. 13,400 15,300 November 7th, during digestion. November llth, during digestion. 53 M. 3,800 72 54 F. 3,300 65 63 M. 3, 844', GOO 3,000 50 M. 3,568,000 2,400 50 52 M. 3,440,000 2,400 50 Hemoglobin. Usually the color index is low; the average was .66 in the ten Massachusetts Hospital cases. White Cells. Except after recent hemorrhage none of our cases showed any leucocytosis, and the average count was 7,240, some cases hav- ing notably low figures (2,400, 3,000, 4,500). Hay em's results agree with this. Yon Limbeck makes no definite statement on this point. Eosenstein and Wlajew found leucocytosis, the latter 12,000 to 17,000. Possibly their cases 288 SPECIAL PATHOLOGY OF THE BLOOD. include the forms of cirrhosis with jaundice in which (see Table XXXI., B) the white cells are more often increased. The forms of hypertrophic cirrhosis without jaundice (fatty infiltrated liver) are here classed with the atrophic cases whose blood has just been described. TABLE XXXI., B. CIRRHOTIC LIVER WITH JAUNDICE. & Age. Sex. Red cells. White cells. Per cent haemo- globin. Remarks. 1 2 3 4 5 42 38 45 35 57 M. M. M. M. F 1,024,000 3,400,000 4,568,000 5,016,000 19, 600 19,500 14,000 12,000 36 50 65 Autopsy. Liver enlarged. Liver enlarged. Adult cells 83 per cent * young 6 7 36 50 M. M. 2,064,000 2,904,000 4,300 2,400 50 54 cells, 17 per cent. Jaundice only transient. Autopsy (hypertrophic cirrhosis). 2. HYPEETEOPHIC CIEEHOSIS WITH JAUNDICE. Red Cells. True (biliary) hypertrophic cirrhosis witli jaundice has ac- cording to Hayem an intense anaemia in many cases. In othe 1 ^ it has no more effect on the blood than ordinary atrophic cir- rhosis. The six cases in Table XXXI., B, averaged a little lower in the count of white cells than the ten atrophic cases, 3,200,000 as contrasted with 3,580,000. Hcemoglobin. In a single case of this variety of cirrhosis Hayem found in four successive blood examinations a color index of more than 1. His counts are as follows : Date. Red cells. White cells. Per cent haemoglobin. Color index. January 9th 1 599 600 41 1 27 u llth 1 884 000 21,803 50 1.39 u 12th 1 798 000 18,082 50 1.46 It 15th 1,971,000 15,500 53 1.40 Dried specimens showed an increased average diameter of the cells as in pernicious anaemia. The patient died January 15th and the autopsy confirmed the diagnosis of hypertrophic cirrhosis. ACUTE YELLOW ATROPHY OF THE LIVER. 289 The observations of v. Limbeck of the increased volume of the red cells in jaundice may perhaps be another example of the condition here noted by Hayem. The presence of bile in the blood makes all haemoglobin estimations unsatisfactory. Only one of our six cases showed this same condition Case 5 in Table XXXI., B. The corpuscles numbered 2,064,000, or forty per cent, and the haemoglobin fifty per cent, a color index of 1.25. This case was jaundiced at the time of the exami- nation. I have seen no confirmation of Hay em's observation by any other writer. White Cells. Leucocytosis is commoner in this than in the other variety of cirrhosis. Hanot and Mennier found from 9,000 to 21,800 leucocytes per cubic millimetre in five cases of hypertrophic cirrhosis and an average of 6,600 in ordinary cirrhosis. Leu- cocytosis was present in four of the six cases of the Massachu- setts Hospital series, the average of all six being 9,000. Diagnostic Value. The blood of either form of cirrhosis has no diagnostic value, so far as I know, except to exclude abscess and hydatids. If no leucocytosis is present, abscess and hydatid cyst can usually be excluded. HYDATID CYST OF THE LIVER. The only observations which I have met with are those of Hayem and Neusser. Hayem states that the blood shows leu- cocytosis and increased fibrin. Neusser considers that the in- crease of eosinophiles which he finds in hydatids serves to dis- tinguish them from hydronephrosis, dilated gall-bladder, etc. ACUTE YELLOW ATROPHY OF THE LIVER. Grawitz records a case with 5,150,000 red cells and 16,000 white cells. A single case with autopsy was studied at the Massachusetts General Hospital in 1894, the blood showing 5,520,000 red cells, 12,000 white cells, and sixty per cent of haemoglobin. 19 290 SPECIAL PATHOLOGY OF THE BLOOD. PHOSPHORUS POISONING. Taussig, 1 v. Jaksch,' J Badt, 3 and v. Limbeck 4 note an increase in the normal number of red cells per cubic millimetre. Taussig found 8, 650, 000 per cubic millimetre; Badt, 6,400,000, 6,500,000, and 6,800,000 in three successive cases; v. Limbeck, 6,500,000 and 7,900,000. That this increase is not due to concentration of the blood through vomiting of liquid is proved by v. Lim- beck's last case, in which no vomiting whatever took place. The count usually falls to normal within a few days. All these changes were verified in thirty-three cases at the Stock- holm Hospital in 1892 (see Stockholm Hospital reports for 1892) . The white cells in v. Limbeck's second case were increased to 12,500. In v. Jaksch's five cases the counts were 58,750, 48,000, 8,000, 4,070,and 3,400. When jaundice is intense and long standing, as in complete obstruction of the bile ducts by gall-stones or tumors, the blood is weakened very notably, and haemoglobin and the count of corpuscles fall steadily. Very little is to be learned upon the subject from the literature, but the qualitative changes men- tioned under catarrhal jaundice are much more marked, and leucocytosis is apt to be present. I have studied the blood in a case of fatal chronic jaundice without fever and for which at autopsy no cause was found. The leucocytes ranged between 12,000 and 14,000. GALL-STONES. Netter 6 and Sittmann 6 have found pyogenic organisms in cultures from the blood of patients with gall-stones, as have also -Gilbert and Girode. 7 1 Arch. f. experiment. Path, und Pharm., vol. xxx. 2 Deut. med. Woch. , 1893, p. 10. 8 Dissert., Berlin, 1891. 4 LOG. cit. , p. 34. * Progres Medical, 1886, No. 46. 6 Deut. Arch, f . klin. Med. , 1894, p. 323 7 La Semaine Med. , 1890, No. 58. CHOLANGITIS. 291 Of the 22 cases of this disease examined at the Massachusetts General Hospital 2 were complicated with cholangitis (see Table XXXII., A). Excluding these 2, leucocytosis was present in only 4 of 20 cases. The red cells were low in 2 cases (2,800,000 and 3,900, 000). The absence of leucocytosis helps us to distinguish the dis- ease from peritonitis and appendicitis, and excludes suppurative cholangitis. TABLE XXXII., A. GALL-STONES. Age. Sex. Red cells. White cells. Per cent haemo- globin. Remarks. 39 30 63 9,9 F. F. F. M 4,768,000 4,820,000 4,610,000 24,400 20,000 18,800 16, 200 Cholangitis also. Cholau gitis . Autopsy . December 16th. 40 60 F. F 4,520,000 13,200 10,000 13,000 12, 500 72 December 18th. December 21st. Temperature, 100.5. 40 49 M. F. 11,500 10,250 10,000 70 90 Jaundice. 47 M 9 800 85 45 38 25 22 25 54 F. F. F. M. F. F. 5,'d72,00'o 3,288,000 4,900,000 9,200 8,900 8,800 8.000 8,000 7,600 100 60 80 Distended gall-bladder. Jaundice. Jaundice. 29 57 F. F 2,844,000 7,400 7,400 85 37 F 7,300 October 1st. 58 M. 8,200 6,000 64 October 5th. 57 F 5 400 68 51 24 F. M. 4,' 320, 000 5,300 4,000 63 Recurrent pain and jaundice. CHOLANGITIS. Here the leucocytosis is well marked whenever the inflamma- tion has got beyond the catarrhal stage (see Table XXXII., B) and helps us to exclude simple impacted gall-stone, with or with- out colic. Cancer may or may not produce leucocytosis, but does not usually increase the fibrin network; it is said by Hay em that cholangitis does increase it. 292 SPECIAL PATHOLOGY OF THE BLOOD, TABLE XXXII., B. CHOLANGITIS. No. Age. Sex. Red cells. White cells. Remarks. 1 2 3 4 5 2l' 65 F. F. F. M. M 4,800,000 6,400,000 4,960,000 4,976,000 50,000 30,000 22,000 14,800 14, 186 Suppurative cholangitis. Jaundice and cholaemia. Gall-stones ; chills. 6 11,000 October 20th. Operation October 22d 7 8 28 34 M. F. 6,640,000 5,592,000 4,770,000 9,000 6,800 4,400 Abscess of liver. Catarrhal. Catarrhal. ABSCESS OF THE LIVER. In all but one of the cases seen by the writer (see Table XXXIII.) the leucocytosis has been very marked. I have never been able to account for its absence in that case. The blood does not differ from that of cholangitis with sup- puration. From cancer it may often be distinguished by the absence of increased fibrin network in cancer, while it is always increased in suppurations. TABLE XXXIII. ABSCESS OF THE LIVER. Age. Sex. Red cells. White cells. Per cent haemo- globin. Remarks. 20 15 60 98 M. F. F. M 4,533,000 3,750,000 4,460,000 33,200 48,000 26,800 18,000 12, 600 January llth. January 14th. Operation. Operation. Operation. 33 F 11,000 November 3d. 26 51 M. F. 2,664,000 3,440,000 17, 500 19,200 20, 600 10, 200 12,000 15,000 9,600 33 November 4th, 11 A.M. November 4th, 2 P.M. November 5th, 10 A.M. Operation ; autopsy. October 19th. October 20th. October 21st. October 25th, au- topsy ; streptococci. CANCER OF THE LIVER. (See Malignant Disease, page 346.) ENDOCARDITIS. 293 GUMMA OF THE LIVER. Von Jaksch in a single case found red cells, 2,756,000; white cells, 6,100. DISEASES AFFECTING THE HEART. PERICARDITIS. (See Inflammation of Serous Membranes, page 247.) ENDOCARDITIS. In many cases of acute endocarditis the blood shows no changes. In others, whatever alterations there may be are cov- ered up by those involved in the rheumatic arthritis associated with the endocarditis. ULCERATIVE ENDOCARDITIS. In idcerative or malignant endocarditis, we may find the signs of a pyogenic infection (see page 216). Sometimes pyogenic cocci can be cultivated from the blood and if present may be of the greatest value in a diagnosis always difficult to make. Grawitz goes so far as to say that in doubtful cases repeated negative results of cultures from the blood make it unlikely that ulcerative endocarditis is present. Sittmann ' considers that important help may be given as to the position of the primary focus of infection by the nature of the organism, present in blood cultures i.e., the pneumococcus pointing to the lung, the colon bacillus to the intestine, etc. Red Cells. As in all forms of septicaemia marked anaemia rapidly de- velops, more rapidly probably than in any other disease. The haemoglobin loses about equally with the corpuscles, according to most observers that is, the blood destruction is so rapid that the red cells do not get thin before they die, as is usually the case, but are cut off in the prime of health. Further evidence of rapid blood destruction is seen in the haemoglobinaemia often present. 1 Loc. cit. 294 SPECIAL PATHOLOGY OF THE BLOOD. Koscher (loc. cit.) records counts of 4,400,000 and 2,750,000, both fatal cases. In one cas& seen by the writer the count was 3,792,000 with fifty-eight per cent of haemoglobin. White Corpuscles. Bieder reports a single case showing these variations : Temperature. White cells. January 2d, 1891 105 17,000 3d, 1891... 99 13,700 8th, 1891 103 15,500 10th, 1891 101.5 18,000 12th, 1891 101.5 21,300 18th, 1891 101 18,800 22d, 1891 104.5 . 13,000 February llth, patient died. Pee found leucocytosis. Eoscher in two cases found : Case I.: 8,800 leucocytes; patient died in two days. Case II. : 16,- 800 and 12,000. Krebs in one case found: October 27th, 15,- 500; October 28th, 44,200; the patient died same day. Nine cases were counted at the Massachusetts Hospital with the following results. In three only the fresh blood was ex- amined and showed marked leucocytosis ; in the others : ULCERATIVE ENDOCARDITIS. Case. Red cells. White cells. Per cent haemoglobin. Remarks. 1 30,100 15,800 18,100 25,700 27,840 18, 100 22,000 20,400 12, 600 14,500 20,400 24,000 10,000 8,900 ' 5> 51 f J May 27th. May 30th. June 17th. May 22d. May 24th. May 26th. May 28th. Autopsy. January 13th. January 14th. January 16th. January 18th ; died. Autopsy. 9 3.. 4 5 3,792,000 6 Practically the same are the counts in the following cases of apparently " benign" endocarditis with fever and rapidly shift- ing murmurs, the first complicating chorea in a boy of thirteen, the other in an adult. MYOCARDITIS. 295 " BENIGN" ENDOCARDITIS. Age. Sex. Red cells. White cells. Per cent haemoglobin. Remarks. 13 M. .... 20,600 62 May 26th. Temperature 102- 104. 17,900 May 29th. 18,700 May 31st. 16,800 June 3d. 21,200 June 4th. 27,400 June 8th. 22,700 June llth. 24,200 June 13th. 21,900 June 15th. 26, 100 June 17th. 26,800 June 19th. 17,400 June 23d. 28,700 June 26th. 21,200 . . July 2d (outdoors). 21,300 ... July 4th. Left the hospital July 19th. 56 F. .... 50,100 November 24th. 35,800 November 27th. 36,600 November 30th. 22,800 ... December 7th. Diagnostic Value. (a) Blood cultures should never be omitted in cases of sus- pected malignant endocarditis. When positive they are of great value, (b) In excluding typhoid the presence of leucocy- tosis is important. I saw within a few months a case in which several consultants had made the diagnosis of typhoid, but in which the presence of marked and persistent leucocytosis and the absence of a typhoid serum reaction convinced me that the case was one of ulcerative endocarditis. This has since been verified. MYOCARDITIS. Whenever stasis and disturbance of the circulation result from weakness of the heart wall, blood changes identical with those described under Valvular Heart Disease are present. Otherwise the blood is normal. 296 SPECIAL PATHOLOGY OF THE BLOOD. VALVULAR HEART DISEASE. Grawitz divides valvular heart disease into three stages with corresponding blood conditions : 1. Stage of full compensation : blood normal. 2. Stage of acute failure of compensation: blood diluted (Oertel's "plethora serosa"). 3. Stage of chronic stasis and cyanosis : blood concentrated for the most part ; at times diluted as well. 1. A valvular lesion per se has no effect on the blood. 2. When compensation fails and blood pressure is lowered, we find (especially in the venous blood) that the fluid from the surrounding lymph spaces has made its way into the vessels and dilated the blood. The specific gravity falls, red cells and haemoglobin are lower than before, while the white cells are un- altered, and the plasma is shown to be more watery than before as well as of increased quantity per cubic millimetre. All these changes are less marked in capillary blood, and hence are rarely observed. 3. If the heart adjust itself partially to the increased work it has to do, and to the chronic passive congestion of the internal organs and at the periphery, the blood is concentrated, probably in part by transudations into serous cavities and lymph spaces, and in part by the increased excretion of moisture by the lungs. The specific gravity and the number of red cells are increased, especially in the capillaries, and to a lesser extent in the venous blood (the conditions being just the reverse of those in acute heart failure, stage No. 2) . This is the condition usually found in heart disease with chronic venous stasis (passive congestion) . But this concentrated condition of the blood may be offset from time to time by fresh weakening of the heart and lessen- ing of blood pressure, and the combination of the two conditions may result in a normal blood count. The condition of concentrated peripheral blood with the count of red cells above normal, is that most commonly seen in chronic heart disease with stasis. Von Limbeck finds that aortic lesions are more apt to show a normal or diminished blood count, while mitral disease is more apt to be accompanied by the temporary dilutions and long- standing concentration, above described. He does not explain VALVULAR HEART DISEASE. 297 the cause of this. One of his patients with double mitral le- sion showed a decrease of 1,170,000 red cells (from 7,500,000 to 6,330,000) after exertion. When the patient was quiet, the le- sion was compensated; on exertion compensation temporarily failed, blood pressure was lowered, and the blood diluted. Sadler 1 found considerable anaemia in three out of four cases of aortic disease, while only two of seven patients with mitral lesions showed anaemia. Schneider's 2 results were similar in that he found the red cells normal in the aortic cases and increased in the mitral ones. Hayem found anaemia most common in aortic regurgitatiou, especially in young people. In the Massachusetts Hospital records out of twelve cases of mitral disease five had less than 4,000,000 red corpuscles per cubic millimetre. Of three cases of aortic disease all were over 4,000,000. I think these figures simply mean that the mitral cases are more apt to come to the hospital in the stage of acute failure of compensation therefore (see above) with diluted blood while the aortic cases often come while compensation is still good and therefore with practically normal blood. White Corpuscles. Almost all writers whom I have consulted agree that the leu- cocytes are normal unless some complication occurs. Yet in a certain number of the Massachusetts Hospital cases mostly (but not exclusively) those with cyanosis, the leucocytes were increased, the counts ranging sometimes as high as 15,000, while the red cells were normal. I suppose this is to be accounted for by the fact that in any case in which the circulation is feeble and slow, the white cells accumulate at the periphery even more plentifully than the red. This is evidently so in the cases of congenital heart disease next to be mentioned, in which the red cells are increased only about forty per cent, while the white are often one hundred per cent more numerous than normal. The apparently normal count of red cells in some of our cases was probably due to the covering up of an anaemic or diluted condition of the blood by concentration, the resultant of the two forces being an apparently normal count. 1 Loc. cit. , p. 33 2 Inaug. Dissert. , Berlin, 1888. 298 SPECIAL PATHOLOGY OF THE BLOOD. Koblank (loc. cit.) gives the following cases illustrating this condition : Red cells. White cells. 1. Mitral leakage 5,461,250 28,000 -f ; autopsy. 2. Aortic leakage 4,716,600 13, 000 -j- This leucocytosis must be taken into account in making in- ferences from cases whose circulations are feeble, and no deeper underlying cause (e.g., abscess, cancer) need be assumed to ac- count for the increase. (Edema and diuresis have in themselves little or no constant effect upon the blood, as a recent observation of Petrowsky's has demonstrated. CONGENITAL HEART DISEASE. In the cyanosis accompanying this affection very high blood counts are reported. Gibson found : Case. Red cells. White cells. Per cent haemoglobin. 1 8 470 000 12 000 110 2 6 700 000 12,000 92 Carmichael reports, red cells, 8,100,000, white cells, 16,000, in a single case, and Toeniessen counted 8,820,000 and 7,540,000 in two similar cases. In one case entirely without evidence of any stasis I counted 8,431,000 red cells per cubic millimetre. How such cases are to be explained I do not know ; the ordinary explanation of concentration of the blood will not hold in cases in which no stasis or lack of compensation exists, yet the skin is blue and the blood counts are enormous. There is no doubt that the peripheral capillaries always con- tain more corpuscles per cubic millimetre than do the veins. Numerous reports from various observers agree upon this. Whether this is on account of the loss of water by perspiration and consequent drain of blood from the skin capillaries is uncer- tain, but in congenital heart disease both capillary and venous blood is overcrowded with corpuscles and the explanation is difficult. Hay em in a case of this sort reports 7,000,000 red cells with a decrease in the average diameter. DISEASES OF THE KIDNEYS. 299 The most important practical deduction from these data is that a blood count in a patient suffering from poorly compen- sated heart disease has no value in determining whether or not anaemia is present. The actual number of corpuscles in the body is not measured by the number contained in a drop of periph- eral blood, since anaemia may be effectually masked by con- centration or simulated by dilution. This holds good equally for any condition involving general stasis and cyanosis either from embarrassment of the heart's action or otherwise (for instance, pneumonia in certain stage, emphysema, displacement of the heart by serous effusions, or tumors), or local stasis of the part from which blood is taken. Penzoldt J noted that in old hemiplegic cases, the blood from the affected side contained more corpuscles than that from the sound side, and the writer has noticed the same thing in a va- riety of vasomotor affections involving local asphyxia. ANEURISM. As a rule I have found the blood entirely normal, but in the following case it might have thrown light on the diagnosis. A patient was recently admitted to the Massachusetts Hospital with an acute affection of the chest, supposed to be pneumonia in spite of the slightness of the fever and the irregularity of the physical signs. At autopsy a ruptured aortic aneurism was found. The blood count had showed 3,324,000 red cells, 20,800 white, and 33 per cent haemoglobin. The low percentage of haemoglobin and red cells was really inconsistent with an acute pneumonia in a man previously well, and might have hinted strongly toward the correct diagnosis had attention been di- rected more carefully to the blood. DISEASES OF THE KIDNEYS. Many factors other than the disease itself may influence the blood of nephritic cases. For instance, in scarlatinal nephritis the long-standing leucocytosis is probably due largely to the scarlatinal poison, rather than to the nephritis. The occur- rence of large quantities of blood in the urine has the same in- fluence as any other hemorrhage upon the blood. (Edema as such has apparently very little effect upon the 1 Berliner klin. Woch., 1881, p. 457. 300 SPECIAL PATHOLOGY OF THE BLOOD. blood, but the loss of albumin in the urine tells both on the cor- puscles and on the serum, thinning both with consequent lower- ing of the specific gravity of the blood. ACUTE NEPHRITIS. 1. Red Cells and Haemoglobin. AVhether largely from the loss of blood from the kidneys or from other causes, the red cells are often much diminished, but the haemoglobin suffers still more. Laache reports an average loss of nineteen per cent of the red cells and twenty-six per cent of their coloring matter. Hayem found no considerable loss of red cells unless the urine was hemorrhagic. The following cases illustrate his re- sults. CASE I. Acute nephritis, ending in recovery. Red cells. March 17th, 1882 3,069,000 March 31st, 1882 2,759,000 April 7th, 1882 2,821,000 May 1st, 1882, alburainuria ceased. May 17th, 1882 3,038,000 May 31st, 1882 , 3,689,000 CASE II. Acute (puerperal) nephritis; recovery. Red cells. April 6th, 1881 2,945,000 9th, 1881 2,976,000 " 12th, 1881, no albumin in urine. " 13th, 1881 3,137,500 " 20th, 1881 3,310,000 CASE III. Nephritis (chronic ?) with haematuria. Red cells 2,821,000. (It should be noted that Hay em's counts are low on the aver- age, and the instrument used by him not very reliable.) Grawitz in acute nephritis records 3,400,000 red cells at the beginning of the third week, and 3,100,000 ten days later. Koblank 1 counted 5,168,700 in a case of acute nephritis with oedema. 1 Inaug. Dissert., Berlin, 1889. ACUTE NEPHRITIS. 301 Sadler (loc. cit.) in six cases of acute nephritis found in two cases 3,590,000 and 2,262,000 red cells; in the other four practi- cally normal counts. TABLE XXXIV. ACUTE NEPHRITIS. Age. Sex. Red cells. White cells. Per cent haemo- globin. Remarks. 56 F 22 200 Temperature 102.5. 11 45 3 23 94 F. M. F. M. F. 4,068,000 3,532,000 14,000 11,900 12,200 14,000 13,200 12,000 11,700 11,100 '52' 43 50 85 Sixth day. Ninth day, temperature falling. Nineteenth day. 33 22 ?,3 M. M. M. 3,904,000 4,300,000 9,300 8,300 7,600 50 48 60 Purpura also. 44 F 7,500 65 37 20 22 M. M. F. 5', 660! 666 4,944,000 6,800 6,000 5,400 5,100 78 58 Acute parenchymatous. Acute parenchymatous. In none of the few cases examined at the Massachusetts Hospital were the red cells much diminished, but in two cases the haemoglobin was very low, the color index being .62 in one and .61 in the other. The blood plates are much increased (Hayem) and fibrin slightly increased. 2. White Cells. Leucocytosis is usually stated to be the rule, lasting often for weeks at a time and gradually diminishing in convalescence. Hayem gives counts of 14,973, 12,400, 15,000, and 13,000. Koblank (loc. cit.) and Grawitz each in a single case found normal counts (7,300 and 5,600). Sadler found an increase in only one of his six cases, and then the highest point reached was 13,312. Of the thirteen cases of Table XXXIV. leucocytosis was present in six, in one of which it was followed for three weeks and still persisted, but it is my own belief that the leucocytosi& of acute nephritis is due either to loss of blood by the kidney 302 SPECIAL PATHOLOGY OF THE BLOOD. or to uraemia. Where these conditions are absent I have not found any leucocytosis. CHRONIC DIFFUSE AND CHRONIC PARENCHYMATOUS NEPHRITIS. Red Cells. In advanced stages the counts may run very low, but more often it is chiefly the haemoglobin that suffers through the drain of albuminoids from the blood into the urine. Hay em gives the following figures : CASE I. Chronic parenchymatous nephritis. Red cells. Per cent haemoglobin. June 20th 4,309,000 43 July 4th 4,216,000 44 October 18th 2,945,000 34 CASE II. Same diagnosis. Red cells. Per cent haemoglobin. March 6th . . . 2,619,500 36 8th . -. 2,836,500 36 23d 2,464,500 27 ^ Koblank (loc. tit.) in the same disease found 3,291,700 red cells in a single case with much oedema. Keinert found 4,050,000 with 50 per cent of haemoglobin and 3,604,000 with 62 per cent hemoglobin. Sadler: Red cells. Case 1 4,120,000 ( 2,405,000 November 19th. " 2 j 1,100,000 January 14th. ( 1,500,000 January 17th. " 3 4,300,000 " 4 4,300,000 i 3,737,500 June 28th. " 5 ] 3,593,700 July 3d. ( 2,187,500 August 15th. ,3,200,000 July 7th. u 6 ! 3,257,000 July 22d. ( 3,137,000 August 21st. CHRONIC PARENCHYMATOUS NEPHRITIS. 303 Grawitz in an acute exacerbation of a chronic parenchyma- tous nephritis found 1,928,000 red cells. The Massachusetts Hospital cases show a considerable anaemia in nine out of the thirty -five, or one-quarter of the series. Great concentration is probably the cause of the very high counts in certain cases. The majority of cases are not far from normal so far as the number of red cells goes, and the haemo- globin is also very little diminished. CHRONIC NEPHRITIS. Red cells. Between 1,000,000 and 2,000,000 . . . Cases. 1 2,000,000 3,000,000 4,000,000 5,000,000 6,000,000 3,000,000 2 4,000,000 11 5,000,000 12 6,000,000 6 7,000,000 3 Color index averages about .' 35 White Cells. Hayem records 25,000, 19,000, 13,000, 10,000, and 6,000 and concludes that the counts vary much not only in different cases but in the same case at short intervals. Koblank found 14,700 in a single case. Sadler in one case found 6,300 in November and 16,000 in the following January; 12,000 in another case; 8,800, 7,700, and 1,916 in others. TABLE XXXV., A. LEUCOCYTES IN CHRONIC NEPHRITIS WITH UREMIA. Age. Sex. White cells. Remarks. Age. Sex. White cells. Remarks. 32 F. 44,000 Eclampsia. 23 M. 12,500 29 F. 22, 600 58 M. 12,400 29 M. 18,650 Polynuclear cells, 50 F. 12,300 83 per cent. 59 M. 12,100 49 F. 16,800 44 M. 11,300 20 F. 15,800 Differential count 31 M. 11,200 normal. 45 F. 6,600 38 M. 15,000 34 F. 4,600 45 M. 15,000 30 M. 4,200 37 M. 14,200 15 F. 13,800 19 cases average 14,495. 25 M. 13,400 304 SPECIAL PATHOLOGY OF THE BLOOD. TABLE XXXV., B. CHRONIC DIFFUSE AND CHRONIC PARENCHYMATOUS NEPHRITIS. No UREMIA. Age. Sex. White cells. Age. Sex. White cells. Age. Sex. White cells. 41 M. 3,000 50 F. 7,300 43 F. 10,300 30 M. 4,500 34 M. 7,400 56 M. 10,700 58 M. 4,800 55 M. 7.400 47 F. 10,800 30 M. 5,000 17 M,. 7,500 10 F. 11,000 27 M. 5,100 25 M. 7,600 25 M. 11,200 30 F. 5,200 28 M. 7,600 16 F. 12,700 41 M. 5,500 15 F. 7.700 11 M. 13,000 20 F. 6,250 14 M. 7,750 27 M. 13,000 7 M. 6,400 33 M. 7,900 66 M. 14,000 39 M. 6,500 27 F. 8,300 56 F. 14,000 8 M. 6,500 30 M: 8,300 43 M. 14,500 8 M. 6,800 24 F. 9,000 45 M. 16,300 41 M. 6,800 52 F. 9,800 28 M. 7,000 20 F. 10,000 40 cases average 8, 657. The same wide range i& seen in Tables XXXV., A and B, in which I have divided the ursemic and the non-uraemic cases into separate tables. It will be seen from these that fourteen out of nineteen uraemic cases showed leucocytosis, while thirty-one out of forty non-uraemic cases showed no leucocytosis. It is difficult to suppose that this is mere coincidence. CHRONIC INTERSTITIAL NEPHRITIS. Hayem found the fibrin more increased in this form of ne- phritis than in any other, and the anaemia less pronounced. Grawitz distinguishes two stages : I. As long as the heart is strong enough to overcome the in- creased resistance at the pejiphery and the disturbances of cir- culation are not marked, the blood is normal. II. When compensatory hypertrophy is no longer sufficient to do the work of forcing the blood through the system, the usual effects of failing compensation (see Heart Disease, page 296) appear (dilution and subsequent concentration of the blood). The white cells are normal. CHRONIC INTERSTITIAL NEPHRITIS. 305 TABLE XXXVI., A. CHRONIC INTERSTITIAL NEPHRITIS. Age. g 02 Red cells. White cells. Per cent haemo- globin. Remarks. 39 Adult. 46 M. F. M. M. 6,040,000 4,548,000 4,244,000 19,381 15,000 12,000 9,724 80 50 67 Ursemiccoma; moribund. Ursemic ; mitral stenosis. Three and one-half hours after a meal. Ursemic ; moribund 20 34 69 M. F. M 4,088,000 3,536,000 6,000 8,300 8 500 66 52 57 87 March 23d. " 30th. 32 M 6,000 65 TABLE XXXVI., B. PYELONEPHRITIS. 6 ^ Age. $ 02 Red cells. White cells. Per cent haemo- globin. Remarks. 1 2 24 26 F. F 3,056,000 2,976,000 2,696,000 3,272,000 4, 200, 000 21,200 15,200 18,800 25,200 16, 800 41 38 33 33 March 10th. Uraemia. " 13th. " 27th. April 14th. Perinephritic abscess too 3 4 33 26 M. F. 4,536,000 2,356,000 15,550 7,280 36 65 Cystitis also. TABLE XXXVI., C. -CYSTIC KIDNEY. Age. X Red cells. White cells. Per cent haemo- globin. Remarks. 55 M 3 664,000 6,400 Adult cells, 72 per cent. Sup- posed cancer. Enormous firm tumor on each side. Autopsy. The cases recorded in Table XXXVI., A, are probably not inconsistent with these rules. Of the four cases with leuco- cytosis three were ursemic, and in the fourth the influence of digestion is seen. The haemoglobin is lower than we should expect from Grawitz's account. Urcemia, it would appear from these tables, may cause leu- cocytosis or at any rate is not infrequently associated with it. Aside from uraemia and hemorrhage, nephritis probably does not cause leucocytosis. 20 306 SPECIAL PATHOLOGY OF THE BLOOD. PYELO-NEPHRITIS. Table XXX VI., B, speaks for itself. The ancemia is often severe and leucocytosis is the rule. STONE IN THE KIDNEY. (See Table XXXVII., A.) The state of the blood depends on the amount of ulceration caused by the stone ; when this is considerable we have leucocytosis. TABLE XXXVII., A. STONE IN THE KIDNEY. No. Age. Sex. Red cells. White cells. Per cent haemo- globin. Remarks. \ 1 22 800 85 Tender in loin B 19 IVT 16,200 15,200 Much pus in urine 3 4 5 25 48 M. F. IVf 4,350,000 4,160,000 14,750 9,000 8 990 78 65 6 7 8 9 58 52 45 M. M. M 5,680,000 4,340,000 6,100,000 3,048,000 8.000 8,000 16,500 7,500 7,500 30 95 Much pus in urine. Two weeks later. 10 51 M 6 000 95 Uric acid, stone passed 11 30 M. 4,980 85 TABLE XXXVII., B. FLOATING KIDNEY. No. Age. Sex. Red cells. White cells. Per cent haemo- globin. Remarks. 1 2 3 4 5 37 41 23 43 38 F. F. F. F. F 5,056,000 4,684,000 5,400,000 4,700,000 9,200 9,000 6,000 2,400 75 75 69 76 75 Aneurism of arch also. 6 94 F 80 5 6 38 24 F. F. 4,416,000 5,800 7, 600 67 80 A large number of similar counts might be quoted. Diagnostic Value. Cancer would also cause leucocytosis, but would not increase fibrin as a rule, while most cases of stone with ulceration do increase fibrin. ACUTE BRONCHITIS. 307 FLOATING KIDNEY. The blood is normal. This fact has some diagnostic value ; for example, when we confound appendicitis with floating kid- ney, as has been done (see page 230) . The presence of leucocy- tosis excludes the latter and favors the former. Most tumors or abscesses with which a floating kidney might be confused could be distinguished by the same criterion. PYO-NEPHROSIS. CASE I. Female, 36; leucocytes, 16,200, of which 85 per cent are neutrophiles. Half a pint of pus found at operation. CASE II. July 25th red cells, 3,856,000; white cells, 9,800; haemoglobin, 45 per cent. July 29th Bed cells, 3,450,000; white cells, 9,000; haemoglobin, 55 per cent. August 3d White cells, 6,650. August 6th Operation. Pint of foul pus. Death. DISEASES OF THE LUNGS. BRONCHITIS. " Acute catarrhal and chronic purulent bronchitis have rela- tively little leucocytosis in most cases" (v. Limbeck). Except for this and a few other passing references, there is hardly anything in literature on the blood in bronchitis, so that I shall be forced to base my statements chiefly on the few counts recorded at the Massachusetts General Hospital. 1. ACUTE BRONCHITIS. Aside from " capillary bronchitis," cases are not infrequently seen in which the signs are simply those of general bronchitis of the finer tubes, yet the symptoms are much more like pneu- monia. Whatever may be the real conditions in the lungs of such patients, their blood is not infrequently exactly like that of pneumonia and does not help at all in the differential diagnosis between the two diseases (see Cases 1 and 2, Table XXXYIIL, A). 308 SPECIAL PATHOLOGY OF THE BLOOD. TABLE XXXVIII., A. ACUTE BRONCHITIS. Age. Sex. Red cells. White cells. Per cent haemoglobin. Remarks. 70 56 9,9, F. M. M. 4,420,000 4,800,000 41,000 26,000 23 450 70 65 67 Temperature, 103. Temperature 101 41 26 F. M. 4,192,000 15,000 11,300 17, 600 14,200 65 70 November 5th. November 16th. November 25th. Temperature 101 5 28 46 F. M. 6,196,000 12,000 11,800 65 Temperature, 104 9,0 M. 10 600 65 Temperature 104 40 M. 10 300 Temperature 101 49, M. 9 300 50 50 52 F. M. M. 5,260,000 5,952,000 8,000 7,900 7,000 72 50 70 25 M. 7,000 74 Temperature, 103. 59 F. 6,800 36 29 M. M. 4,392,000 6,000 8.600 4,000 72 80 October 31st. November 3d. Temperature, 102. TABLE XXXVIII., B. CHRONIC BRONCHITIS. Age. 1 Red cells. White cells. Per cent haemo- globin. Remarks. Adult. 48 27 61 20 M. M. F. M. F 3,680,000 5,384,000 4,300,000 18,500 15,000 8,800 8,000 7, 925 63 73 63 78 Chronic febrile, with laryngitis. . Recovery. Constipation ; neurasthenia ; two weeks afebrile. Five months. 18 26 29 20 M. F. F. M. 4,700,000 4,100,000 7,792 6,700 5,500 5,062 70 61 Keratitis, conjunctivitis. No symptoms. Asthma. Empyema of the antrum. One month. In the majority of acute cases, however, the blood shows no changes unless concentration due to cyanosis be present (see Cases 4 and 7, Table XXXVIII., A). In chronic cases (Table XXXVIII. , B) leucocytosis is very uncommon, more so, I think, than the table represents. If more counts were added, nearly all, I think, would bo normal. EMPHYSEMA AND ASTHMA. 309 The red cells and hemoglobin show no changes to speak of in either acute or chronic cases. The blood has no diagnostic value so far as I know except that when pneumonia is in question a normal count of white cells speaks against it and in favor of bronchitis. If emphy- sema is also present it sometimes produces a different blood condition from that of simple bronchitis. EMPHYSEMA AND ASTHMA. Grawitz reports an increase in the number of red cells in emphysema, which he believes to be due to cyanosis and to cover up the really anaemic condition of the blood of many patients. Practically the same conditions are present as in the cyanosis of heart disease (see page 296) and the concentration of the blood is brought about in the same way. Leichtenstern 1 noticed a di- minution in haemoglobin at the time when the heart first fails, due probably to the diminished blood pressure which allows the lymph from neighboring tissues to flow into the vessels and dilute the blood. In both asthma and emphysema it has been noted by Miiller, 2 Gollasch, 3 Gabritschewsky 4 and others that eosinophiles are very numerous in the sputum, and Fink 2 also noted an in- crease of the same cells in the blood, running as high as 14.6 per cent instead of the normal one to two per cent. This increase is present only at the time of the paroxysm and for a short time before and after it. Billings 5 reports the following counts : January 26th. February 4th. February llth. Bed cells 3 911 000 4 221 000 4 630 000 White cells 8 300 7 500 7 600 Haemoglobin 68 per cent 75 per cent Poly morp honuclear cells 36 " Lymphocytes (small) . . Lymphocytes (large) . . Eosinophiles. . . 5 5.2 53 6 1 " 38 2 per cent 33 9 per cent Few normoblasts No nucleated red cells. " Ueber das Hb-Gehalt des Blutes, " etc. , Leipzig, 1878. 2 Ref. in Fink, "Beitrage z. Kennt. des Eiters," Dissert., Bonn, 1890. 3 Fortschrittder Med., 1889. 4 Arch. f. exp. Path, und Pharm., 1890, p. 83. 6 New York Med. Journal, May 22d, 1897. 310 SPECIAL PATHOLOGY OF THE BLOOD. Tlieir presence in increased numbers before a paroxysm makes it possible to predict its coming (v. Noorden, Schwerskewski). As this applies only to pure bronchial asthma and not to cases secondary to disease of the heart or kidney, Schreiber states that we are enabled to distinguish bronchial from cardiac or renal asthma by the increase of eosinophiles in the blood and sputa in bronchial cases, which does not occur in asthma due to cardiac and renal trouble. ASTHMA. Age. Sex. Red cells. White cells. Per cent hffimo- globin. Remarks. 26 M 32,500 Fifth. Temperature 100. Bron- 50 F 19,200 19 800 50 chitis and emphysema. Seventh. Temperature normal. Typical bronchial asthma during 70 29 M. M 5,500,000 13,000 9,750 paroxysm." Chronic asthma and emphysema in paroxysm ; polynuclear cells, 79 per cent. For Pneumonia, see page 184. For Phthisis, see page 255. For Abscess of Lung, see page 235. SYPHILIS OF THE LUNG. In a case of syphilitic infiltration of the lung (autopsy Drs. Councilman and Wright) recently observed at the Massachusetts Hospital the leucocytes rose rapidly from. 8,700 to 27,400 as death approached. PART V . DISEASES OF THE NERVOUS SYSTEM, CON- STITUTIONAL DISEASES, AND HEMOR- RHAGIC DISEASES. CHAPTEE YHI. DISEASES OF THE NERVOUS SYSTEM. NEURITIS. IN a single case of multiple neuritis, febrile and apparently of an infectious nature, the following counts are found in the rec- ords of the Massachusetts General Hospital : Date. Temperature. Red cells. White cells. Per cent haemoglobin. July 10th 101. 4,816,000 25,000 42 " 1 3th 24,800 " 16th 18, 700 " 20th .. 21,000 " 25th 4,320 000 16,000 60 " 31st (No fever ) . ... 28, 700 AuEjust 7th u u 19,500 " 20th u 23,200 The patient, a boy of eleven, recovered and left the hospital well. But these changes occur also in alcoholic (afebrile) neuritis, as the following counts show. Case. Red cells. White cells. Per cent haemoglobin. 1 3 608 000 15, 000 75 2 3,260,000 14,000 64 3 . .. . 13, 700 60 4 11,200 68 5 7,700 80 6 6,700 82 312 SPECIAL PATHOLOGY OF THE BLOOD. In all cases the counts were made just at mealtime, so that the leucocytosis is not due to digestion. Gastritis was not present in either case. One case of post-diphtheritic neuritis in a child of eight showed the presence of anaemia only: Eed cells, 3,850,000; white cells, 7, 393 ; haemoglobin, 70 per cent. Neuritis in lead poisoning does not affect the count of leuco- cytes, as twenty -five cases studied at the Massachusetts Hospital have shown. Neuralgia, whether facial, intercostal, sciatic, or ovarian, showed normal blood in numerous cases examined at the Massa- chusetts Hospital. DISEASES OF THE BRAIN. Meningitis (see Inflammation of Serous Membranes, page 248). Zappert in one case of brain abscess found only 4,000 white cells. In pachy meningitis hcemorrJwgica and cerebral syphilis (one case of each) v. Jaksch found leucocy tosis. My own experience has been the same. Cerebral and cerebellar tumors have no effect on the blood as far as could be judged from nine counts in the former and three in the latter disease. Von Jaksch found slight leucocytosis in two cases of brain tumor and one of cysticercosis. Zappert found normal blood in one case of cerebral tumor. Fresh cerebral hemorrhage usually causes leucocytosis, as the following table shows : Age. (Sex. Red cells. White cells. Per cent haemo- globin. Remarks. 42 53 47 M. M. M 5, 51 2,0(30 31,000 30,000 25,000 19 400 95 '85* Autopsy. Polynuclear cells, 92 per cent. Autopsy. Autopsy 70 F 16 800 68 Hemorrhage four days before count 87 M 5,560,000 15,600 12 300 90 70 Autopsy. 38 M 10 400 58 Conscious ; recovered. 87 M 10 300 90 Autopsy <65 M 10 200 60 GENERAL PARALYSIS OF THE INSANE. 313 CHOREA AND TETANY. Chorea showed in twelve cases normal blood except for in- creased percentages of eosinophiles, as in Zappert's two cases, which Louga confirms. Burr has made a careful study of the blood in thirty-six cases and arrived at the following conclusions : There is usually a slight diminution in red cells and a moderate diminution in haemoglobin. Any severe grade of anaemia is due to some com- plication. He did not record the leucocytes. Tetany shows no blood changes. DISEASES OF THE SPINAL CORD. Chronic diseases of the spinal cord, such as tabes dorsalis, syringomyelia, spastic paraplegia, diffuse myelitis, paralysis agitans, and progressive muscular atrophy, are found to pro- duce no changes in the blood. For Spinal Meningitis, see page 249. GENERAL PARALYSIS OF THE INSANE. Capps ' has made a careful study of the blood in nineteen cases and comes to the following conclusions : 1. Eed corpuscles and haemoglobin are always slightly dim- inished, the averages being 4,789,900 and 85 per cent. 2. Most cases show a slight leucocytosis 22 per cent above the normal on the average. Early cases may have no leucocy- tosis. 3. The differential counts show that the blood is slightly older than that of normal adults. The polymorphonuclear leu- cocytes average nearly 74 per cent and the smaller forms of lym- phocytes only 14.2 per cent, while the larger forms of lympho- cytes are relatively numerous, averaging 7.8 per cent. In a few cases the eosinophiles were very numerous 2 (8.7 and 6.4 per cent) . 4. At the time of convulsions the red cells and haemoglobin are apparently increased (due no doubt to the violent muscular 1 American Journal of the Medical Sciences, July, 1896. 2 Roncoroni (Archiv. di Psichiat. Scien. 1894, p. 293) finds eosiDO- philes increased even to twenty -five per cent in the agitated and violent cases. 314 SPECIAL PATHOLOGY OF THE BLOOD. contractions which raise blood pressure and concentrate the blood, or to cyanosis). There is a sudden and pronounced increase in the leucocytes during and after convulsions or apoplectiform attacks. That this is not due to concentration of the blood or to stasis Capps thinks is shown by the fact that not only the number but the differential count of white cells show changes, the " large mono- nuclear" cells being relatively increased, sometimes as high as 25 per cent. Myelocytes were seen in one case after the convulsions, and especially just before death when in a leucocytosis of 18,250 11 per cent were myelocytes. 1 HYSTERIA AND NEURASTHENIA; HYPOCHONDRIASIS. A large number of cases have been counted at the Massa- chusetts General Hospital, with a view to excluding other diseases. The blood count is always normal except that in a certain number of the hysterical cases eosinophiles are rela- tively increased, and that many of the neurasthenics show the increased percentage of lymphocytes which I have alluded to above (page 97) as characteristic of a variety of debilitated conditions. Marked ancemia is seldom present, although the haemoglobin is not infrequently as low as 65 per cent. Beinert 2 found the haemoglobin under 60 per cent in only 4 out of 48 cases of hys- teria, and in none of 36 neurasthenics. The value of the blood examination in such cases, like that of the urine or the lungs in hysteria, is as negative evidence, and in this respect it is important. When the discrepancy between complaints and signs is great, we want to be doubly sure that nothing hidden escapes our notice, and the blood examination is one of the most valuable adjuvants we have in the discovery of deep-seated inflammation or malignant disease, as well as in giving us a general measure of the patient's degree of bodily health as distinguished from nervous force. The former may be high when the latter is low, or both may be low, and the dis- 1 Leucocytosis has been repeatedly noticed in convulsions from various causes. Probably the irritant which causes the motor discharge also acts on the leucocytes by chemotaxis. 2 Munch, med. Woch., 1805, No. 14. CONSTITUTIONAL DISEASES. 315 tinction marks out two classes of cases in which somewhat dif- ferent treatment is appropriate. There is no use in undertak- ing to make " blood and fat" when the patient has already plenty of each, though it may be well to carry out the same regime as a matter of suggestion. MENTAL DISEASES. The association of anaemia with insanity is too frequent to be a mere coincidence, though it is hard to make either serve as a cause for the other. Very possibly they should both be looked upon as symptoms of a common underlying (unknown) cause. This form of anaemia has been noticed by Houston 1 in mel- ancholia and general paralysis, and by Smith 2 in various forms of insanity. - Krypiakiewicz 8 noticed an increase of eosinophiles in acute forms of insanity but not in the chronic forms. The leucocyto- sis of acute delirium* is exemplified by the following case from the Massachusetts Hospital records : A girl of fifteen; acute delirium; leucocytes, 12,750; no food for eight hours; red cells, 4,510,000; haemoglobin, 63 per cent: Puerperal mania is to be distinguished from the delirium of puerperal sepsis by the fact that the latter shows leucocytosis with increased percentage of polymorphouuclear cells, while the former has no leucocytosis (if uncomplicated) and the eosinophiles are apt to be increased 6 (diminished in sepsis). A case of puerperal mania seen by the writer showed : Bed cells, 5,210,000; white cells, 6,500; haemoglobin, 84 per cent; eosinophiles, 8 per cent. CONSTITUTIONAL DISEASES. OBESITY. Oertel distinguishes a plethoric and an anaemic form of obesity not merely clinically but by the evidence of post-mortem 1 Houston : Boston Med. and Surg. Journal, January llth, 1894. 1 Smith : Jour, of Ment. Sc , October, 1890. ^Krypiakiewicz: Wien. med.Woch., 1892, No. 25. 4 Ref . in Klein-Volkmann's "Sammlung klin. Vortrage," December 1893. Neusser . Loc. cit. 310 SPECIAL PATHOLOGY OF THE BLOOD. examinations. He believes that there is a real over-filling of the vessels in the first. The second form occurs most often in women. Kisch examined (with v. Fleischl's instrument) the haemo- globin of 100 obese patients; 79 showed over 100 per cent of haemoglobin, 1 reaching 120 per cent, while the other 21 were anssniic. DIABETES. There is nothing characteristic about the blood except the increased amount of sugar to be detected (.57 per cent as against .1 normally); but this is not a clinically applicable test. Two simple tests for diabetic blood have recently attracted attention : 1. Bremer's test: Heat thick-spread blood films to 135 C. ; cool and stain with one-per-cent aqueous solution of Congo red for two minutes. The blood if diabetic looks yellow (to the naked eye). Normal blood similarly treated looks red. Staining with methyl blue also shows a difference between normal blood and diabetic blood. The normal is blue, the diabetic yellowish green. 2. Williamson's test: Make a mixture of Blood, . . . . ... . 20 c. ram. (2 drops). Aqueous methyl blue (1 : 6, 000), . . . 1 c.c. Liquor potassse, 60 per cent (sp. gr. , 1.058), . 40 c.c. Water, 40 c. c. Xiet the mixture stand three to four minutes in boiling water. With diabetic blood the mixture turns yellow, with normal blood it does not. Williamson has found this test positive in eleven diabetics and negative in one hundred cases of other diseases. Bremer claims that by his method cases of diabetes can be recognized before sugar appears in the urine or after it has (temporarily) disappeared. Le Goif confirms the value of the test. Eichner and Folkel find Bremer's reaction to be as stated, but find similar color changes in leukaemia, Hodgkiu's disease, and Graves' disease, and changes something like it in a variety of cachectic conditions. Badger has studied the blood of dia- betics, leukaemics, cases of Graves' disease, and other cases at the Massachusetts Hospital. Only in Graves' disease did he find reactions like those of diabetic blood. GOUT. 317 The alkalinity has been said to be greatly diminished, espe- cially in the fatal coma, but v. Noorden thinks the tests are unreliable. Fat is often increased in the blood, up to about twelve times the normal, so that the serum is milky, and glycogen has been demonstrated microchemically in the corpuscles. Red Cells. Sugar in the blood draws water from the tissue into the vessels, thereby diluting the blood; but in a short time the blood frees itself of the excess of sugar and fluid through increased diuresis so as to concentrate the blood. These two alternating influences serve to explain the widely different counts of different observers. Toward the end of the disease a decided cachexia often de- velops, the anaemia of which may be temporarily covered up by the concentration above noted, or accentuated by the dilution which sometimes occurs. Accordingly we may find the corpus- cles increased, normal, or diminished in different cases or at different times with the same case. Grawitz counted 4,900,000 red cells in a patient in compara- tively good health, and three weeks later, when the patient had just been seized with the fatal coma, the count showed 6,400,000 per cubic millimetre. The white cells show no constant changes, except that.v. Limbeck has noted in several cases that the digestion-leucocy- tosis is unusually large even without previous fasting. Von Jaksch found leucocytosis in one of his eight cases, but on this point as on many others his results are almost unique. The only similar observation is that of Habershon, 1 who reports moderate leucocyosis, decreased by strict diet. In thirteen cases I have never seen leucocytosis. A single case of diabetic coma showed 4,200 leucocytes per cubic millimetre. GOUT. A few cubic centimetres of serum from gouty blood made acid with acetic acid (six drops of a twenty-eight-per-cent solution to every drachm of serum) deposit crystals of uric acid on a 1 St. Bartholomew Hosp. Rep. , 1890, p. 153. 318 SPECIAL PATHOLOGY OF THE BLOOD. thread in from eighteen to forty-eight hours ; but this is not al- ways to be found, and is by no means peculiar to gout. ' Uric acid is to be found in the blood in pneumonia, cirrhotic liver, nephritis, grave anaemia, leukaemia, and gravel; also in health and after a meal of calf's thymus or any food containing much nuclein. The red corpuscles show no special changes except in severe chronic cases which are sometimes anaemic. The white cells are increased according to Neusser, while v. Limbeck and Grawitz found the blood wholly normal. It is particularly in this disease that Neusser supposed the " perinuclear basophilic granulations" to exist in the white cells, which he believes to be characteristic of any " uric-acid diath- esis." Futcher has conclusively disproved this. Fibrin is in- creased in acute cases. MYXCEDEMA. Le Breton 2 examined the blood in one case before and after thyroid treatment and found after forty days' treatment that the red cells had risen from 1,750,000 to 2,450,000, the white from 4,500 to 9,600, and the haemoglobin from 65 to 68 per cent. The remarkably high color index in this case before treat- ment (nearly 2. !) corresponds with the observations of Le Breton in the dried specimen, which showed a decided increase in the size of the red corpuscles. He also noticed before instituting the thyroid treatment the presence of nucleated red cells and an excess of the polymorphonuclear form of leucocytes. Under treatment the nucleated red cells disappeared and the lympho- cytes rose to their normal per cent. Putnam 3 has watched a similar case in which the red cells rose from 3,120,000 to 5,700,000 under thyroid treatment. Murray 4 has collected 23 cases with blood examinations. Of these 7 showed normal blood, 10 were anaemic, 4 had leucocyte- sis, and 2 had both anaemia and leucocytosis. 1 It is important to evaporate the serum at a temperature not above 70 F. , otherwise crystals will not form. 2 Le Breton : Ref. in Wien. med. Blatter, 1895, p. 49. 3 Putnam : Eef. in Murray's article in "Twentieth Century Practice of Medicine," vol. iv. 4 Murray : "Twentieth Century Practice of Medicine," vol. iv., p. 710. CRETINISM. 319 Kraepelin : noticed (like Le Breton) a marked increase in the average diameter of the corpuscles in three cases, even when the count and the haemoglobin were normal. I have had an opportunity to examine the blood in three cases of this disease, but did not find anything remarkable in any of them. Case. Red cells. White cells. Per cent Haemoglobin. 1 . 4,670,000 6,000 87 2 4,460,000 8,800 3 4 856 000 5 200 80 Differential counts were made in three cases and no increase in the size of the corpuscles, such as Le Breton and Kraepelin saw, was present in these cases. The count showed: Case. Polymorphonuclear cells. Lymphocytes. Eosinophiles. 1 . 67 28 5 2 67 27.8 4.4 3 74 26 The increase of eosinophiles in two of these cases may per- haps be due to the skin troubles present in the disease. J. J. Thomas found a few myelocytes in a case of Putnam's. CRETINISM. Koplik 2 records the following in two cases of sporadic cretinism : CASE I. Fifteen months old; advanced stage of disease. Hemoglobin, 18 per cent. CASE II. Eed cells, 3,026,000; white cells, 13,500; hemo- globin, 105 per cent. This high haemoglobin corresponds to normal foetal blood. The child was nine weeks old, but its back- ward development is mirrored in the blood. As the case im- proved under thyroids the haemoglobin came down. 'Kraepelin : Deut. Arch. f. klin. Med., vol. xlix., p. 587. 2 New York Medical Record, October 2d, 1897. 320 SPECIAL PATHOLOGY OF THE BLOOD. GRAVES' DISEASE (BASEDOW'S DISEASE; EXOPHTHALMIC GOITRE). The blood is normal, except for an occasional associated chlorosis and sometimes a marked lymphocytosis. In one case I found 51.3 per cent of lymphocytes and 1 per cent of mye- locytes in 1,000 leucocytes, the polymorphonuclear cells being only 48 per cent ; but in fourteen other cases I have never found this again. The same fact has been noticed by Neusser (cited in Klein, loc. cit.}. Oppenheimer 1 found the red cells and haemoglobin normal in two cases. Yon Jaksch 2 in one case " complicated with myx- oedema" found 3,818,000 red and 8,000 white cells. The association of Graves' disease with chlorosis is illus- trated by two cases from Zappert : 3 Case. Red cells. White cells. Per cent haemoglobin. 1 2,858,000 3,800 32 2 2,738,000 3,800 30 The same writer found eosinophiles much increased (8.5 per cent) in one out of four cases. ADDISON'S DISEASE. Some, but not all, cases are accompanied by marked anaemia. Neumann 4 observed a case in which the symptoms came on acutely and the red cells sank to 1,120,000 x^er cubic millimetre. During the convalescence which followed the cells ran up above normal, reaching 7,700,000. Tschirkoff 6 reports two cases in which the red cells were re- spectively 3,280,000 and 2,933,000 at the lowest, but whose haemoglobin was extraordinarily high, over 100 per cent in one case. This he found on spectroscopic examination to be due to a great increase of reduced haemoglobin in the corpuscles. Methaemoglobin was also noted. The white corpuscles showed no changes, quantitative or 1 Deut. med. Woch., 1889, p. 861. * Zeit. f. klin. Med. , 1893, p. 187. 3 Zeit. f. kliD. Med., 1893, p. 266. 4 Neumann : Deut. med. Woch., 1894, p. 105. 'Zeit. f. klin. Med., 1891, vol. xix., Suppl. Heft 37. OSTEOMALACIA. 321 qualitative, except that they contained black pigment granules. Three cases have been examined at the Massachusetts Hospital. The first, a woman of thirty, showed 6,240,000 red cells with 14,000 white, and 90 per cent of haemoglobin. The differential count of 900 leucocytes showed the following figures : Polymor- phonuclear cells, 53.4 per cent; lymphocytes, 41 per cent; eosino- philes, 4.5 per cent; myelocytes, .9 per cent. The eosinophiles were very large, some of them eosinophilic myelocytes. The second, a man of forty-two, was very anaemic and weak at entrance and showed: Bed cells, 2,196,000; white cells, 7,500; haemoglobin, 20 per cent. Differential count of 200 leucoctyes showed: Polymorphonuclear cells, 65 percent; lym- phocytes, 31.5 per cent; eosinophiles, 3.5 per cent; five normo- blasts; marked poikylocytosis. Under suprarenal extract his blood improved in a month till his red cells numbered 4,700,000; white cells, 9,000; haemo- globin, 65 per cent. The third a man of fifty-two, showed : October 20th Ked cells, 2,848,000; white cells, 4,800; haemoglobin, 45 per cent. December 10th Eed cells, 2,624,000; white cells, 7,100; haemo- globin, 45 per cent. Differential count: Polynuclear, 74 per cent; small lymphocytes, 22 per cent; large lymphocytes, 4 per cent; eosinophiles, .4 per cent. No nucleated red cells. A fourth patient, kindly sent me by Dr. Kogers, of Dor- chester, showed: Eed cells, 2,864,000; white cells, 2,000; haemoglobin, 51 per cent. Differential count of 300 cells showed: Polymorphouuclear cells, 63.3 percent; lymphocytes, 33.3 per cent; eosinophiles, 2.3 per cent; basophiles, .3 per cent. I have never seen melanin in the leucocytes as Tschirkoff did in his two cases. OSTEOMALACIA. The blood has for a long time been supposed, on the author- ity of v. Jaksch (Zeit. f. Jclin. Med., Vol. 13, page 360), to exhibit a diminished alkalinity, the bones being supposed to be eaten away by acids in the blood. Von Limbeck and many other ob- servers have lately shown that the blood is normal in alka- lescence. 21 322 SPECIAL PATHOLOGY OF THE BLOOD. Corpuscles and haemoglobin are usually within normal limits quantitatively, but Neusser reports an increase of eosinophiles and the presence of myelocytes in the blood. Ritchie 1 confirms Neusser and found also the young leu- cocytes more numerous than normal. Fehling, 2 Sternberg, 2 Chrobak 2 found no increase of eosino- philes. Eieder's case was normal in all respects: Bed cells, 4,892,000; white cells, 5,600; eosinophiles, 3.6 per cent; poly- morphonuclear cells, 61 per cent. EICKETS. 1. Anaemia is always present in severe cases and often in moderate ones. This, together with the fact that many cases of rickets are associated with an enlargement of the spleen, has led to the use of the misleading term " splenic anaemia." There is no form of anaemia found in rickets that may not be found in other conditions (Morse). Hock and Schlesinger found an average of 2,500,000 red cells in a considerable number of cases with and without en- larged spleen. Von Jaksch describes a case in which the red cells sank from 1,600,000 to 750,000 within three months, and Luzet saw a similarly rapid process, the cells falling from 2,110,000 to 1,596,000 within three weeks. On the other hand, in Morse's admirable study of twenty well-marked cases the red cells aver- aged over 4,500,000 and not a case fell below 3,500,000. 2. The haemoglobin is always relatively low; it averaged 63 per cent in Morse's cases, a color index of about .7. Felsenthal got similar results. White Corpuscles. It is often difficult to say whether or not the leucocytes are increased, owing to the occurrence of most cases in infants at an age when leucocytes are always higher than in adults how much higher at any given age depends largely upon the degree of vigor and forwardness of development of the individual child. In Morse's series, for example, the average age of the infants 1 Edin. Med. Journal, June, 1896. 2 Cited by Ritchie (loc. cit.). RICKETS. 323 is twelve months. And for this age none of the counts in his series seem to me necessarily abnormal. They are all under 16,000 except three, these three being 17,900, 18,800, and 22,000 respectively, the latter in a nine months' infant. Many of the counts seem to me subnormal for infancy (5,500, 7,200). Most observers find leucocytosis present in many cases, but not in all. QUALITATIVE CHANGES. Bed Cells. As in all anaemias of infants, the " degenerative" and " re- generative" changes are relatively common. Polychromatophilic forms and nucleated corpuscles are fre- quently to be found, the latter often in great numbers but with a majority of the normoblast type. White Cells. Lymphocytosis is said to be marked, but, as with the ques- tion of leucocytosis, we are never quite sure whether the numbers are abnormal for that age., for lymphocytosis is the normal con- dition in infants' blood. When, however, as in a case mentioned by Eieder, we find 75 lymphocytes in every 100 leucocytes, the child being four years old, we are surely dealing with a pathological condition. Another of his cases, a seven-months' child, rachitic, with 57 per cent of lymphocytes, seems to fall within normal limits. Not so with Morse's cases. The highest percentage of lympho- cytes in his series was 69, in an infant of two months. I have similar counts in health at that age. The average of his twenty cases is 43 per cent, which is, if anything, rather low for that age. The same difficulty arises with regard to the reports of eosinophilia in rickets, since eosinophiles are always rela- tively numerous in infancy. Morse's highest figure was 7 per cent, his average 3 per cent. They were highest in cases with splenic tumor. In Rieder's four cases and in the three seen at the Massachusetts Hospital, no eosinophilia was present. Myelocytes in small number (.5-2. per cent) are not uncommon, and may be considerably more numerous. CHAPTEE IX. BLOOD DESTRUCTION AND HEMORRHAGIC DIS- EASES. 1. PURPURA HJEMORRHAGICA. THE blood is practically that of anaemia from hemorrhage (red cells and haemoglobin reduced, white cells increased, occa- sional nucleated red corpuscles or polychromatophilic forms). Agello ' has found methsemoglobin in the blood, and hence con- cludes that the disease is a poisoning of the corpuscles by ptomains absorbed from the intestine. The blood plates are much diminished and may be entirely absent in the worst stages. Bacteria of various kinds have been reported in the disease, but negative results are also common, and their presence is probably not significant. The red cells may fall as low as 2,500,000, but are much oftener slightly or not at all diminished. In many mild cases there are no demonstrable blood changes. On the other hand, Osier mentions a case which sank to 1,800,000, and the loss of blood may give rise to a fatal anaemia of the microcyte type (see page 158). Bensaude 2 has observed that in 16 cases characterized by large hemorrhages (2 = acute "infectious," 2 with tuberculosis, 2 chronic, 10= Werlhof 's disease) the clot shows no retraction and no transudation of the serum. Cases with small hemorrhages (toxic, rheumatic, cachectic, and nervous) do not show any such abnormal characteristics. Hence he concludes that at the outset of a case of purpura, observation of the clotting process may enable us to foretell whether or not the case is to be of a severe or of a mild type. He found the blood lesion above described to 'RiformaMed., Napoli, 1894, p. 103. 2 LaSemaineMed., 1897, p. 21. HEMOPHILIA. 325 be greatest during the hemorrhagic crises, slowly disappearing between them. Hayem has confirmed these observations. He finds the fibrin network almost invisible. Despite this and despite the absence of contraction in the clot, the actual rate of clotting is normal. SCURVY. There are no characteristic blood changes known. When hemorrhage is severe the red cells may sink very low, to 557,875 in a case of Bouchut's; Ouskow and Hayem saw counts of 3,500,000 and 4,700,000. The usual qualitative changes of secondary anaemia are present in severe cases ; haemoglobin suf- fers as usual more than the count of red cells. Leucocytes are generally increased, whether from hemor- rhage or from some complicating inflammatory process. Barlow's disease may lower the red cells as far as 976,000 as in a case of Eeinert's the haemoglobin being seventeen per cent and the white cells 12,000. This was the day before death. In any form of scurvy the blood plates may be much dimin- ished. HEMOPHILIA. The blood changes are practically those just described and show nothing characteristic of the disease. Coagulation is slower than normal and blood plates are sometimes very scanty. The white cells are sometimes persistently diminished as in the following cases : I. Sept. llth. Sept. 14th. Sept. 17th. Sept. 20th. Sept. 23d. Sept. 24th. Red cells 2,960,000 3 800 000 White cells Haemoglobin 3,400 42 per cent. 3,400 3,800 3,900 3,700 64 per cent. 3,300 49 per cent. II. February 8th. February 28th. 4,400,000 5,000 30 per cent. 3,600,000 5,000 28 per cent. 1- Daily nosebleed. 326 SPECIAL PATHOLOGY OF THE BLOOD. BLOOD DESTRUCTION (HJEMOCYTOLY8IS). I. Besides the slow destruction of corpuscles which takes place in any ordinary anaemia, we have a group of conditions under which a large number of red cells are suddenly destroyed in the circulation itself. This may take place by 1. Separation of the haemoglobin from the corpuscles so that it colors the serum. 2. Actual breaking to pieces of the red cells without separa- tion of the haemoglobin. If normal blood is drawn and left to stand, the serum which separates from the corpuscles is not red-tinged or but very slightly so, provided all shaking and jarring are avoided. A very slight reddish tinge may appear in the serum even with most careful technique. In some condition* the haemoglobin, while not actually separated from the corpuscles within the vessels, is so loosely connected to them that a considerable quantity sepa- rates post mortem and colors the serum in spite of the avoid- ance of any jar. This condition is to be distinguished from true haemoglobin- aemia, in which the serum is actually colored before leaving the vessels, although the two conditions really represent only dif- ferent degrees of vulnerability of the red cells. We are surer of a diagnosis of haemoglobinaemia when we find bits of broken-down cells in the fresh blood and the additional evidence of haemoglobinuria or jaundice. 1. Severe forms of malaria, yellow fever, typhus fever, severe forms of septicaemia, and rarely scarlet fever may cause haemoglobinaemia. 2. Paroxysmal hcemoglobincemia, so-called, is a variety whose cause is unknown and which does not seem secondary to any other disease, unless a certain relationship to syphilis be es- tablished, and to malaria. The attacks are brought on by a great variety of causes (cold, muscular or mental strain, etc.). Some persons can always bring on an attack by putting the hand or foot into cold water. Blood Examination. Coagulation is very rapid, but the clot soon dissolves again (Hay em). The fresh blood occasionally shows deformities BLOOD DESTRUCTION (H^EMOCYTOLYSIS). 327 in the corpuscles or bits of broken cells, and lack of rouleaux if examined during a paroxysm. As a rule the corpuscles of the peripheral blood look normal. Frazer has recently reported a case in which he excited a paroxysm by a cold bath and studied the blood with great care. Time. Red cells. White cells. Per cent haemoglobin. Blood plates. 10A.M. Before bath 11.05 A.M. Twenty-five minutes after bath; urine pale. 11.45 (urine dark) 1 15 p M 4,075,000 3,633,300 3,760,000 4, 200 000 15,000 21,800 21.300 21,500 50 50 60 50 450,000 696,000 525,000 4,250,000(1) 3 45 P M 3,800,000 17,700 50 1,600,000 Next day, 1 P.M 4,100,000 18,700 50 500,000 The enormous increase of " blood plates" is striking. It is difficult to resist the conclusion that these blood plates were bits of broken red corpuscles. The serum was currant-jelly colored. The appearance of the corpuscles was quite normal. All that is known of the disease is expressed by saying that for some reason the red cells are abnormally sensitive, so that any one of a variety of slight disturbances is sufficient to separate their haemoglobin and set it loose in the plasma. 3. Extensive burns have been reported to cause hsernoglo- binsemia with breaking up of the red cells, presumably through changes in the serum similar to those which make duodenal ulcer so common a sequel to bad burns. 4. Snake poison and scorpion poison may have similar effects. II. Another group of corpuscle destroyers is that which works by changing the hcemoglobin to methcemoglobin. The most important of these is 1. Chlorate of Potash. This destroys the corpuscles and pro- duces hsemoglobinaemia and the usual train of symptoms (jaun- dice, dark urine, etc.) due to this. Brandenburg 1 examined the blood of a woman who had taken two and one-half ounces of chlorate of potash in water the night before. The blood showed marked leucocytosis, broken and 1 Berliner klin. Woch., 1895, No. 27. 328 SPECIAL PATHOLOGY OF THE BLOOD. distorted red cells. In gross it was chocolate-colored and the serum after separation of the clot was brown. The red cells progressively decreased as follows : Red cells. White cells. First day 4,300,000 20,000 Second day 2,500,000 Fourth day 2,300,000 Fifth day 2,100,000 Sixth day 1,900,000 Seventh day 1.600,000 15, 000 (death). 2. Ehrlich and Lindenthal 1 report the case of a patient who was poisoned with nitrobenzol. Ten hours after the blood was chocolate-colored and showed methaemoglobin bands. Un- der the microscope there were no changes till the third day, when poikylocytosis appeared. * Red cells. White cells. Per cent haemo- globin. - Nucleated red cells per cubic millimetre. Fifth day 2 275 000 Much increased 55 2,070 Seventh day 1,845,000 it ( 50 7 900 Eleventh day Fifteenth day Seventeenth day Nineteenth day " death. 1,600,000 905,000 1,102,000 900,200 a 44 40 24,700(!) 12,000 1,300 540 The nucleated red cells were at first mostly normoblasts; later mostly megaloblasts. 3. Antipyrin and antifebrin in doses of thirty to forty-five grains may cause great cyanosis and dangerous prostration through transformation of the haemoglobin and methaemoglobin. In certain persons much smaller doses produce the same effect. 4. Phenacetin poisoning (Kronig: Berl. Idin. Woch., 1895) may cause actual blood destruction with anaemia in case the patient survives the immediate effects of the deprivation of oxygen. A fatal case of chloral poisoning at the Massachusetts Hospital showed 14,400 leucocytes with 54 per cent of haemo- globin. 5. Phosphorus poisoning (see Liver, page 290). 6. Workers in aniline dyes and nitroglycerine factories may 1 Zeit. f . klin. Med. , 1896, p. 427. ILLUMINATING GAS POISONING. 329 be severely poisoned by nitrobenzol compounds inhaled and pro- ducing methaemoglobinaemia. 7. Pyrogallic acid and pyrogallol as used in treatment of skin diseases may lead to death through destruction of the red cells. Chromic acid (for instance, as applied through the vagina) may have a similar effect. Many other less common substances work the same ill-effects on the blood. III. A third group of substances, of which carbonic oxide gas is the type, poison by combining chemically with the haemo- globin and preventing its combination with the oxygen of the air. 1. Illuminating gas is for our purposes the most important of this group. The appearance of individual blood cells is not altered nor do they break up, but the corpuscles are useless to breathe with, as they cannot take up oxygen. The color of the blood is very bright red, much brighter than normal. Eed Cells. Yon Limbeck 1 found in two cases 6,630,000 and 5,700,000 respectively. The volume of these corpuscles (estimated by Bleibtreu's method) was greatly increased, amounting to 70.7 per cent (normal 41-48 per cent) , so that apparently the size of the individual cells is increased. Miinzer and Palina 2 found 5,700,000 red cells in one case. Leucocytes. Eaton 3 reported four cases, in all of which the white cells were increased, the counts ranging between 15,000 and 22,000 per cubic millimetre. Miinzer and Palma (loc. cit.) found 13,300 in their case. Twelve such cases have been examined at the Massachusetts Hos- pital with the following results : 1 Loc. cit. , p. 234. 2 Zeit. f. Heilk., vol. xv., p. 1. 3 Boston Medical and Surgical Journal, March 14th, 1895. 330 SPECIAL PATHOLOGY OF THE BLOOD. ILLUMINATING GAS POISONING. Age. Sex. Red cells. White cells. Per cent haemo- globin. Remarks. 41 M 31 200 Coma ; recovery. 49 M 27, 100 September 12th ; coma 21 M 19,900 26,000 70 September 13th, entirely well. Coma ; recovery. 19 M 25,470 97 40 M 22 900 75 60 M 21,200 15,500 20,400 75 November 27th ; coma. November 29 , convalescent. Coma ; recovery. 95 M 20,360 110 Death. 45 M 20,100 Coma ; death. 16 F 18,500 84 Coma ; recovery. 19 M 17 000 December 22d 4,930,000 17,500 17,000 December 23d. War-then ' reports the same condition in a single case. Here the specific gravity was also very high (v. Limbeck finds that this is to be explained by the increase in the actual size of the corpuscles) . When there is any doubt as to diagnosis, the following test will settle it : Shake a small quantity of fresh-drawn blood into three times its volume of subacetate of lead. If the blood con- tains CO the mixture becomes of a fine red color; otherwise it turns chocolate-colored." 2. Da Costa (Med. News, March 2, 1895) reported a consider- able diminution in haemoglobin of patients during etherization, especially anaemic patients, but the investigations of Lerber 3 do not confirm this. Tansy Poisoning. A single case examined at the Massa- chusetts General Hospital showed: Red cells, 4,600,000; white cells, 21,000; haemoglobin, 70 per cent. Corrosive Poisoning (Ammonia Fumes). A patient whose throat was covered with a fibrinous pseudo-membrane in conse- quence of inhaling ammonia fumes showed a leucocytosis of 25,800. Red cells and haemoglobin normal. Opium Poisoning (Chronic). The majority of cases of the ' Virchow's Archiv, vol. cxxxvi. ' 2 Rubner . Z<-it. f. anal. Chemie, xxx. , p. 112. 3 Inaug. Dissert., Basel, 1896 (seep. 10d). ACUTE ALCOHOLISM. 331 morphine habit show normal blood, but in October, 1897, a man of twenty-six entered the Massachusetts Hospital for the morphine habit who showed at entrance 36,000 leucocytes per cubic millimetre. Five days later the count was 21,200. A differential count of 500 leucocytes made on this day showed : Polymorphonuclear neutrophiles, 71 per cent; small lympho- cytes, 12 ; large lymphocytes, 10 ; eosinophiles, 6 ; myelocytes, 1. At the time of leaving the hospital he still showed a leuco- cytosis of 16,400. He had no fever and physical examination was entirely negative. Ptomain Poisoning (Kotten Fish). A mother and her four children were brought to the Massachusetts Hospital suffering from the effects of decayed fish eaten that day. The blood showed the following : (1), mother: leucoctyes, 21,600, of which 95.3 per cent were polymorphonuclear; (2), boy of seven years: leucocytes, 19,900; (3), boy of three years: leucocytes, 56,800, of which 92 per cent were polymorphonuclear ; (4) , girl of five years : leucocytes, 32,600; (5), girl of thirteen months: leucocytes. 55,400. The red cells and haemoglobin were normal. All the patients made prompt recoveries. ACUTE ALCOHOLISM. It has been shown experimentally that in animals made drunk with alcohol, there is an invasion of the blood and tissues by micro-organisms from the intestine. It may be that some of the counts here recorded are thus to be explained. ACUTE ALCOHOLISM. Age. Sex. Red cells. White cells. Per cent haemo- globin. Remarks. 36 F 15 900 Two weeks drinking hard. 88 M 14 200 74 Temperature 102 ; died ; D.T. Temperature 101. 42 M 12 000 62 Temperature 101 ; D. T. 32 44 M. M 3,946,000 10,200 9 600 30(?) 80 29 87 F. M 4,288,000 8,000 7 800 55 62 D. T. 60 F 7 450 65 8?, M 5 700 88 Autopsy 88 M 5 600 D T PART VI. MALIGNANT DISEASE, BLOOD PARASITES, AND INTESTINAL PARASITES. CHAPTEE X. MALIGNANT DISEASE. THE BLOOD AS A WHOLE. 1. The specific gravity is reduced in most cases, running roughly parallel with the haemoglobin. 2. Coagulation is normal or slower than normal in uncom- plicated cases. When sloughing and inflammation are present it may be rapid. 3. Fibrin is usually normal; an increase means inflammation in or around the tumor or an inflammatory complication. CANCER. Red Corpuscles. As in tuberculosis, we are frequently surprised to find but little diminution in the number of red cells. In all but very ad- vanced cases this is the rule. It is a change of the individual red cells (pallor, loss of size, of weight, degenerative changes), rather than a reduction of numbers. Nevertheless in the later cachectic stages of most cases of malignant disease, we do find a quantitative anaemia, the counts often running as low as 2,500,000 and occasionally sinking as low as in pernicious anaemia. Thus v. Limbeck records a case (complicated by repeated hemorrhages) with only 950,000 red cells per cubic millimetre. The lowest of my own cases was 1,457,000 per cubic millimetre. CANCER. 333 There seems to be no considerable difference between cancer and sarcoma as regards their effect on the red cells. The fol- lowing table summarizes our cases : TABLE XXXIX., A. GASTRIC CANCER. Rbd Cells. Between 1, 000, 000 and 2, 000, 000 4 cases. 2,000,000 " 3,000,000 11 " 3, 000, 000 " 4, 000, 000 23 " 4,000,000 " 5,000,000 15 " 5,000,000 " 6,000,000 16 " Over 6,000,000 3 " Average, 4,090,000+ 72 Nucleated red cells present in eleven cases out of fourteen examined. Normoblasts always in majority. A few megaloblasts in three cases. TABLE XXXIX., B. GASTRIC CANCER. Leucocytes per Cubic Millimetre. Between 3, 000 and 4, 000 1 case. 4,000 " 5,000 :.. ... 4 cases. 5,000 " 6,000 17 " 6,000 " 7,000 9 " 7,000 " 8,000 8 " 8,000 " 9,000 8 " 9,000 k < 10,000 9 " 10,000 " 12,000 6 " 12,000 " 15,000 4 " 15,000 " 20,000 12 " 20,000 " 30,000 .... 5 " 30,000 " 40,000 3 " Total, 167 counts in 86 cases. Average, 10,600 + As will be seen by consulting Table XXXIX., A, the count of red cells is sometimes above normal, doubtless due to concentration of the blood from some cause. Probably the same influence is at work in other cases, and many of those showing normal counts have really fewer red cells than they should. Such abnormally high counts are not rare, as the following examples show: 334 SPECIAL PATHOLOGY OF THE BLOOD. Author. Case. Affection. Red cells. Per cent haemoglobin. Osterspey *. 1 Cancer of the stomach 5 040 000 80 Osterspey .... Osterspey. . .. 2 3 Caiicer of the liver and stomach Cancer of the gullet 6,184,000 8,280 000 87 48 Neubert 2 1 Cancer of the stomach. 5,085,000 73 Neubert 2 Cancer of the liver 4 918 000 70 Reinert 3 Cancer of the stomach . 6 200 000 77 I wish to lay some stress upon this point, because it has been stated by some recent writers (e.g., Grawitz: "Pathologie des Blutes," Berlin, 1896) that the red cells are almost always dim- inished in malignant disease. The high counts in cancer of the gullet are obviously to be explained by the lack of liquid taken, the blood being greatly concentrated as in any other form of starvation. That this increase is not invariably present (see Table XL., page 345) is doubtless because some cesophageal tumors permit the ingestion of liquid in normal amounts and of a certain amount of solids. The highest counts in the Massachusetts Hospital series are in simple gastric cancer without any stenosis at either end of the organs (see Cases L, III., XVI., and XXL), and the lowest count (1,632,000) was in a similar case just before death. Tak- ing all the cases of cancer in this series together, the average of the seventy-five cases at the time when treatment began was 4,140,- 000 red cells per cubic millimetre. Haemoglobin. Bierfreund, 4 who has examined seventy-two cases with re- gard to their percentage of coloring matter, found that in rela- tively slow and long-standing cases it averaged 68.5 per cent, and in the worst cases 57.5 per cent. In cases of mammary cancer after operation the haemoglobin is of course lower owing to hemorrhage, and Bierfreund noticed that as a rule the haemo- globin began to rise toward normal much later than after opera- 1 Dissert., Berlin, 1892. 2 lnaug. Dissert., Dorpat, 1889. 3 " Zahlung d. Blutkorp. , " Leipzig, 1891, 4 Langenbeck's Archiv, vol. xli. CANCER. 335 tions for non-malignant conditions a week later on the average and that it never reached the point at ivhich it ivas before the operation. ' The following table from Bierfreund is of interest as illustrating these points. Cases were examined before and after operation, and the examinations were continued daily after the operation until the haemoglobin began to rise again. This occurred very late as compared with other operations. Diagnosis. Per cent haemoglobin before, operation. Per cent haemoglobin after operation. Per cent loss. Regeneration time. Malignant tumor without complication. Very large or rapidly grow- ing tumors. Tumors with "softening" or disturbances of function. 68.5 56.6 57.5 53 38.4 39.7 15.5 18.2 17.8 23 days. 27.8 days. 27 days. Total, 72 cases. Av. , 60 Av.,42.8 17.2 Av., 25.9 days. By "regeneration time" is meant the number of days elapsed after operation before the haemoglobin begins to rise. After operations for other causes (non-malignant) the average regeneration time is fourteen to twenty days. It is very important that these results of Bierfreund 's should be tested. In Mikulicz's surgical clinic at Breslau all patients have their haemoglobin tested regularly. In this country the surgical portion of the profession have not as yet taken hold of blood examination, and many questions about the blood in sur- gical affections remain unanswered. Eeinbach 2 examined 16 cases and found the haemoglobin range between 18 to 70 per cent, with an average of 50 per cent. Eieder's 3 cases average 53 per cent (sarcoma much lower see below) . 1 This is all the more extraordinary because Bierfreund specially noted that even in patients who gained weight notably after the operation the haemoglobin did not rise so high as it had been before operation ; he watched them for months after it. Apparently the actual presence of the tumors is not the only cause of the lack of corpuscle substance. 2 Langenbeck's Archiv, 1893, p. 486. 3 "Beitragez. Kenntniss cl. Leucocytes is," Leipzig, 1892 (Vogel). 336 SPECIAL PATHOLOGY OF THE BLOOD. Laker ' noticed the low haemoglobin percentage in malignant tumors and thought it a help in excluding benign tumors or tuberculosis, in which the haemoglobin is much less diminished. In the 87 cases of malignant tumors in which I have notes of the haemoglobin (see tables) the average is 58 per cent. Com- paring this with the average count of red cells (4,140,000), we get a color index of .65, distinctly higher than the average of chlorotic cases, of which, however, the figures distinctly remind us. The highest cases of this series had 100 per cent and 90 per cent of haemoglobin respectively, and the lowest 20 per cent and 22 per cent ; in these last two cases the color indexes were .36 and .58 respectively, not excessively low. As pointed out by Taylor (loc. cit.) cases of malignant disease can be divided into three groups with reference to their blood : 1. Those with approximately normal blood. 2. Those with a low haemoglobin but a nearly normal num- ber of cells. 3. Those with great diminution both in cells and coloring matter. Among our own cases at the Massachusetts Hospital about one-half fall under the second group, one-quarter under the first, and one-quarter under the third. As the disease progresses, the red cells and haemoglobin steadily go down (except in cancer of the gullet), and at the time of death 1,000,000 cells per cubic millimetre is not rare. The color index usually remains below 1. Compared to most other varieties of secondary anaemia (e.g., those in tuberculosis or nephritis) a quantitative anaemia that is, a loss of red cells as well as of haemoglobin is relatively more frequent. In gen- eral the degree of anaemia is parallel to the amount of cachexia, except when hemorrhage increases it (as in tumors of the stom- ach or uterus). How far the anaemia may be due to actual destruction of cells by toxic (?) products of the tumors is doubtful. Grawitz found that the injection of extracts of cancerous tissues caused in rabbits a temporary dilution of the blood, so that the cells per cubic millimetre were diminished, and it may be that this plays, some part in the causation of the low blood counts. 1 Wien. med. Woch., 1886, Nos. 18 and 19. CANCER. 387 Qualitative Changes. (a) The average diameter of the red cells is often diminished either (as in chlorosis) by a diminution of the size of nearly every corpuscle, or by a less general shrinkage, many cells being of normal size. The very large forms seen in pernicious anaemia are rare in the anaemia of malignant disease, and never, I think, reach the size of the giant forms seen in the former condition. Very small cells, on the other hand, are as common in ad- vanced cases as in any other form of anaemia, except chlorosis. Deformities and degenerative changes are very common in well- marked cases, often as great as in pernicious anaemia, though they may be slight or absent. According to Strauer, the deformities found in malignant dis- ease are greater than those found in any form of tuberculosis, and this fact he thinks of value in diagnosis. This observation has been confirmed by Taylor. Degenerative changes are sometimes well marked, but seldom, if ever, reach so extreme a condition as occurs in many cases of pernicious anaemia. (b) Nucleated red corpuscles are the rule in all advanced cases, and in some others. Taylor found them in one-half of the twenty-two cases examined by him. Malignant disease dif- fers in this respect from tuberculosis and most other conditions involving secondary anaemia, in that the nucleated red cells are much more common in cancer and may appear even when there is no considerable loss of red cells (numerically) or even when the haemoglobin is also normal (Schreiber). I have found them in four-fifths of all severe cases examined. As a rule the nucleated corpuscles are of the norrnoblast types (including small forms with dividing nuclei) , but in very cachectic cases we may find megaloblasts as well always, so far as I know, fewer in number than in the normoblasts. This constitutes one of the points of distinction between pernicious anaemia and the severest types of secondary anaemia, such as occur in malignant disease. The megaloblasts, when present, are in the minority as compared with the normoblasts. For example : 22 338 SPECIAL PATHOLOGY OP THE BLOOD. r (Five normoblasts. / ~ Case I. < ,, > Seen while counting 400 leucocytes. ( Three megaloblasts. ) _ ( Two normoblasts. ) _ Case II. < ^ > Seen while counting 500 cells. ( No megaloblasts. J _ ( Five normoblasts. ) _, Case III. < XT } Seen while counting 200 cells. / No megaloblasts. f Cases could easily be multiplied. The characteristics of the blood changes in malignant disease, then, so far as concerns the red cells, are those of secondary anaemia, which at times attains the severest type but only when cachexia is marked, or when hemorrhage complicates the disease. The specific gravity follows in a general way the haemoglobin percentage. On the white corpuscles in malignant disease a great deal of interest has centred, and very conflicting reports have been published. As the effects of cancer and sarcoma seem to be somewhat different we will consider them separately. 1. THE LEUCOCYTES IN CANCER. (a) Quantitative Changes. We should expect great differences in the blood of different cases if we consider what a wide range is included between the small, hard, slow-growing, curable cancer of the lip which may produce little or no impairment of the general health, and the "fulminating," rapidly growing cases with numerous metastases and profound prostration. The former class of cases may show a blood normal in all respects, including a normal leucocyte count; while in the latter the blood may be so profoundly altered as to be confused with that of pernicious anaemia on the one hand, or with that of leu- kaemia on the other. In a general way it may be said that the more " malignant" the cases the greater the changes in the blood. The effect upon the leucocytes depends upon the following conditions : 1. The position of the tumor. 2. Its size, rapidity of growth, and the number, size, and position of its metastases. CANCER. 339 3. The resisting power of the individual. 1. Position. (a) Tumors of the gullet involving stricture but not extending to other tissues are often accompanied by a dim- inution of the leucocyte count, owing to the starvation which they produce. This is not true of all cases, as is shown in the accompanying tables, but when the leucocytes are increased there is usually an involvement of other organs as well. (b) Cancers of the uterus and some of those of the stomach, by reason of the hemorrhage which they produce, are apt to be associated with a very high leucocyte count. (c) Tumors of the thyroid and of the pancreas are said by some writers to cause a specially great leucocytosis. In my own experience, tumors of the kidney have shown very marked in- crease of white cells. 2. Size. Other things being equal, the larger and more rapidly growing tumors show in most cases a greater leucocy- tosis than small, slow-growing ones. Thus the cancers of the lip and of the pylorus, scirrhus of the breast or of the penis, show smaller counts than tumors of the liver, omentum, and kidney, which are apt to grow more rapidly. Metastases in the bone marrow are thought by some observers to give peculiar qualitative blood changes (see below) . In general, metastases, being a method of rapid growth, simply add to the leucocyte count. These distinctions eliminate some of the apparent contradic- tions between the findings of different individuals who were simply describing cancers of different types. But even within a single type, there are very marked differences in different cases. For instance, Alexander 1 found the leucocyte count in cases of scirrhus of the breast to vary between 2,360 and 21,700. Similar differences have been reported in cancers of the stomach (e.g., Schneider 2 finding leucocytosis in all of twelve cases, while Osterspey 3 in another series of twelve cases found leucocytosis in only two). 3. Resisting Power. Possibly a part of these differences is to be explained by differences in the resisting power of the in- dividual. But if this is so, we cannot measure the endurance of a given patient by his general health. As in the Civil War the 1 Alexander : These de Paris, 1887. -Inaug. Dissert., Berlin, 1888. 3 Iuaug. Dissert., Berlin, 1892. 340 SPECIAL PATHOLOGY OF THE BLOOD. pale, city -bred men outlasted the healthy farmers, so here the tumor's rapidity of growth seems often to be greatest in the most vigorous young individuals, while dried-up old women will resist its advance for a longer period. We come now to the conditions to be found in particular types of cancerous growth. Surprisingly little work has been done on the blood in malig- nant disease, such cases usually being under the charge of sur- geons who rarely value such investigations. Except for scattered counts here and there, all our knowledge of the corpuscles rests on the work of Hayem and Alexander in France, and Kieder, v. Limbeck, Pee, Sadler, Reinbach, Osterspey, Grawitz, Strauer, Schneyer, and Schneider in Germany. CANCER OF THE BREAST. Most of our data come from Hayem 1 and his pupil Alexan- der. 2 1. Scirrhus Groivths. Number of cases, 14. Average leu- cocyte count, 11,400. Highest count, 21,700 ; lowest, 2,360 the last is somewhat doubtful as to diagnosis ; except for this case, which was in a very old, dried-up woman, the lowest count was 7,400. In 10 out of the 14 cases, the count was over 10,000. IB the 3 cases seen by the writer 2 showed no leucocytosis, 1 a consid- erable leucocytosis. 2. Medullary (Encephaloid) GrowiJis. Three cases, all over 10,000 average 11,300. Effects of Operation. The following figures from Hayem are also of interest : CASE I. Scirrhus of the Breast. Before operation 21, 700 Five weeks after operation (wound not quite healed) 10,000 Wound completely healed 6,200 Seven months after operation 8,990 (beginning to rise again) The growth recurred some months later and leucocytosis was again present. 1 Hayem : "Du Sang," Paris, 1889, p. 947. 2 G. Alexander: "De la Leucocytosis dans les Cancers," Paris Thesis, 1887. CANCER OF THE STOMACH. 341 CASE II. Scirrlms of the Breast. First Second count. count. Before operation 11, 500 11, 450 After operation , 8,500 6,200 CASE III. Scirrhus of the Breast. First Second count. count. Before operation 11,000 12,400 After operation 8, 400 CASE IV. Scirrhus of the Breast. Before operation 7,400 After operation 1, 300 CASE Y. Medullary Cancer of the Breast. * Before operation 10, 000 After operation 9, 000 Hay em considers' that by watching the leucocyte count we can predict the coming of a recurrence before any physical signs are present. This he did in Case I. of the series just given. I have seen no confirmation or refutation of this statement. It is one of the many points to which the attention of surgeons should be directed. CANCER OF THE STOMACH. Here we have a much larger body of data to judge from. Thus : Hayem 1 in 12 cases found leucocytosis present in 5, absent in 7. Schneider 2 in 12 cases found leucocytosis in 12 (all). Schneyer* in 18 cases found leucocytosis in 4, and these 4 all under 11,000. Osterspey 4 in 12 cases found leucocytosis in 5. Eieder 4 in 6 cases found leucocytosis in 3. Sadler 5 in 13 cases found leucocytosis in 2, and in both there were complications (abscess of liver, perforation of gullet with gangrene) to which the leucocytosis might be due. 1 "Du Sang," Pa,ris, 1889, p. 948. 'Inaug. Dissert., Berlin, 1888. 3 Inaug. Dissert., Berlin, 1892. 4 Loc. cit. 5 Original-Mittheilungen aus der Klinik v. Jaksch, 1891. 342 SPECIAL PATHOLOGY OF THE BLOOD. Reinbach 1 iu 4 cases found leucocytosis in 2. Eeinert 2 in 2 cases found leucocytosis in 2. Laache 3 in 5 cases found leucocytosis in none. 4 Despite these facts we have the record of a certain number of single cases in which the leucocytosis has been enormous. For instance, Welch in "Pepper's System of Medicine" men- tioned a case in which the ratio of white to red cells was 1 : 25 (normally 1: 750 ). Eisenlohr's 5 case showed 1 white to 50 red, and Potain's 6 case showed 1 white to 48 red cells. The Massachusetts Hospital series of 86 cases showed leuco- cytosis in 30 cases and none in 56 (see Table XXXIX., B). Out of those showing leucocytosis 10 were under 12,500, that is the leucocytes were but slightly increased, leaving only 20 out of 86 (or twenty-three per cent) in which the leucocytosis was very marked. Among these 20, the highest counts were 40,000 and 39,000, and the highest ratio 1 : 62. In this series I have excluded all cases in which there was evidence of metastasis in other organs ; this means excluding 7 cases, 6 of which showed leucocytosis, and helps to account for the low average leucocyte count in the other 86 cases. In over three-fourths of these cases the diagnosis was made certain either by operation or by autopsy ; all the others showed either a palpable tumor in old cachectic patients with pain and vomiting, or other equally clear evidence for the diagnosis. Doubtful cases have been excluded. As will be seen by the table, in some of the cases the counts were verified by repeated examinations, while in others only a single count that made when the patient entered the hospital was recorded. As a rule, the high leucocyte counts were in the more cachec- tic cases; but this does not always hold. Cases 10, 11, and 28 in Table XXXIX., A, were very cachectic but showed no leucocytosis. The position of the tumor in one or another part of the 1 Langenbeck's Archiv, 1893, p. 486. * LOG. cit. 3 "Die Anamie," Christiania, 1883. 4 Apparently, since he draws attention to the fact that there is leuco- cytosis in a case of cancer of the uterus. s Deut. Arch. f. klin. Med., 1877, vol. xx. " Gaz. des Hop., 1888, No. 57. DIGESTION LEUCOCYTOSIS IN CANCER OF THE STOMACH. 343" stomach seemed to have no connection with the number of leu- cocytes. On the whole, leucocytosis is relatively infrequent in cancer of the stomach, occurring in only about one-third of the early cases. As the disease progresses we may get a leucocytosis, par- ticularly in case its growth is rapid and metastases are frequent and numerous ; but some cases, particularly those in which the tumor is small and grows slowly, may run their entire course without any leucocytosis being present. In this respect they are like the majority of small, slow-growing cancers in other parts of the body (see below). Hemorrhage or perforation is of course accompanied by an increase in the number of white cells in fact the highest count in the present series (105,600) occurred in a case in which a can- cer of the stomach with metastases in the liver perforated into the peritoneal cavity and started a virulent, quickly fatal peri- tonitis. DIGESTION LEUCOCYTOSIS IN CANCER OF THE STOMACH. A considerable body of statistics has accumulated to show that in the great majority of cases of gastric cancer the leucocy- tosis of digestion (see above, page 98) does not occur. B. Muller ' noticed this fact in 5 cases of cancer of the stomach. Schneyer 2 in 18 cases found it invariably absent, while in 3 cases of benign stenosis of the pylorus a considerable digestion leuco- cytosis appeared, as was also the case in 7 out of 8 cases of ulcer of the stomach, the exception being a fatal case. He found both incipient and advanced cases to be similarly affected. In 5 of his cases and in some of Muller 's HC1 was present in the gastric contents, so that the absence of digestion leucocytosis was not due to absence of HC1. Hartung (Wiener klin. Woch., p. 697, 1895) in a series of 10 cases (mostly advanced) found no digestion leucocytosis, where- as a marked increase occurred in cases of malignant disease of other organs. Capps b in 17 cases examined at the Massachusetts General 'Prag. med. Woch., 1890, No. 17. 2 Zeit. f. klin. Med., 1895, p. 475. 3 Boston Med. and Surg. Journal, November 4th, 1897. 34:4 SPECIAL PATHOLOGY OF THE BLOOD. Hospital found a digestion leucocytosis in 2, the increase being respectively 3,270 and 3,850 cells over the count before the be- ginning of digestion. In the other 15 cases there was no in- crease after a large proteid meal. Since Dr. Capps' article 20 more cases have been investigated at the hospital, in 19 of which the digestion leucocytosis was absent. Thus in a total of 37 cases only 3, or eight per cent, shoived any digestion leucocytosis. In 5 out of 10 cases of chronic gastric catarrh the digestion leucocytosis was present ; it was also present in a case of benign stricture of the pylorus in a man of forty-nine on whom an operation was successfully performed later. Three cases of ulcer of the stomach showed marked increase as did several cases of hyperacidity and other gastric disorders (see Diseases of the Stomach, page 280). CANCER OF THE STOMACH WITH METASTASES. Most writers have not separated the cases with metastasis from those without it. A glance at the seven cases of Table XXXIX., C, shows that with one exception leucocytosis was present throughout most of the disease. TABLE XXXIX., C. CANCER OF THE STOMACH WITH METASTASES. Age. i Bed cells. White cells. Per cent haemo- globin. Remarks. 48 M. 4,228,000 5,000 6,200 70 January 23d. Stomach and liver. January 28th, mealtime. 7 300 January 28th three hours later 41 38 M. M. 4,272,009 5,432,000 10,000 10,190 13 653 57 52 Stomach, liver, and glands. January 6th. Stomach and liver. January 12th. 66 M. 5, 168', 030 7,000 14,400 19,600 70 62 January 22d, died. February 14th, no cachexia. March 6th, liver involved. March 12th. 21,640 March 17th, cachectic. Adult M 3 352 000 16 000 Stomach liver and spleen. 54 47 M. M 4,160,000 24,000 24,200 22,500 34,350 60 Stomach and liver. November 7th, cancer of stomach 30,600 and liver. November llth. 105,600 ! November 14th, perforation peri- tonitis. CANCEK OF THE LIVER. 345 CANCER OF THE GULLET. Most authors are agreed that no increase in fact usually a decrease of white cells is the rule in this disease. Thus Rie- der found 6,900 in one case; Osterspey's two cases showed no leucocytosis, and Escherich and Pee found similar results. 1 This is probably due to the fact that the position of the tumor, by causing starvation, tends to lower the leucocytes, while it be- longs to the class of small, slow-growing cancers which do not as a rule tend to produce leucocytosis. Nevertheless, two of the five cases in the Massachusetts Hos- pital series (see Table XL. ) did have leucocytosis, perhaps owing to some metastasis or complication. There was no autopsy in either. TABLE XL. CANCER OF THE GULLET. No. Age. Sex. Red cells. White cells. Per cent haemo- globin. Remarks. 1 9, 58 46 M. F 5,488,000 6,800 6,800 7 000 100 May llth. May 18th. October 18th 3 4 5 51 56 86 M. M. M 2,824,000 4,920,000 10,600 5,400 6,600 9,860 7,600 11,500 8,725 11 800 50 72 30 October 19th, before food. October 19th, after food. October 20th. Before food. Four hours later. Hsernaturia also fi 65 M 11 100 68 7 47 M 15 400 80 8 9 10 38 67 38 M M. F. 4,604,000 4,560,000 15,600 16,400 20,600 60 60 50 During digestion. CANCER OF THE LIVER. (See Table XLI.) Ou,t of fourteen cases, leucocytosis was present in eight a larger proportion than in gastric cancer. The cases were not all primary in the liver or bile ducts, but none originated in the stomach, and in all the greater part of the growth was in the liver itself. 1 Reinbach's two cases showed a diminution in the polymorphonuclear cells, which in all probability means a normal or diminished leucocyte count. 346 SPECIAL PATHOLOGY OF THE BLOOD. The comparatively great diminution in the red corpuscles will be noted in the Table XLI. The condition both of red and white cells is doubtless due to the rapid growth of tumors of the liver as compared, e.g., with those of the stomach or lip (see below). TABLE XLI. CANCER OP THE LIVER. Age. j jj ! OQ Red cells. White cells. Per cent haemo- globin. Remarks. 55 M. 4,170,000 5,000 Bile ducts = starting-point. Autopsy. 61 M. 3,824,000 5,200 52 59 M. 4,570,000 8,000 Operated. 72 M. 4,100,000 9,000 44 F. 4,953,000 3,784,000 7,800 19,700 69 68 January 4th, 1896. Autopsy. February 12th, 1896. 31 F. 4,572,000 8,000 62 65 M. 10,300 58 57 M. 11,150 50 54 50 M. M. 4,072,666 ,200,000 9,300 10,800 ' Differential count of 1,000 cells: Poly., 82.4 per cent: small lymphocytes, 8.5; large lympho- cytes, 8.1; old lymphocytes, 1. Primary in bile ducts. Autopsy. y "M. 4,108,000 9,970 45 January 1st, 1896. 43 M. 4,160,000 11,200 14,100 9 January 3d, 1896. Autopsy July 17th. July 19th . Autopsy. 64 M. 2,768,000 15,800 45 May 6th . 2,880,000 21,900 May 24th 1,530 45 May 28th 2,928,000 11,700 June 8th 35 M. 3,800,000 9,800 No vein be r3d. 22,000 November 5th. Novembe r 6th. Differential count of 500 cells: 48 F. 2,900,000 17,500 48 Poly. , 92 per cent ; lymphocytes, 8. Autopsy. Differential count of 500 cells: Poly.. 92 per 30 F. 3,660,000 17,200 82 cent; lymphocytes, 5.8: eosinophiles, .2; myelocytes, 2. Autopsy. Polynuclear cells, 83 per cent ; Myelocytes, 1 per 16,800 B ( cent. 31 M. 3,120,000 18,700 52 December 20th. 15,600 Decembe r 30th ; before food. 14,000 Four hours later. Adult. M. 4,408,000 25,500 70 50 M. 4,544,000 35.600 November 29th, 1895. 36,400 Decembe r 10th, 1896. 3,136,000 23,000 January 15th, 1896. 4,056,000 28,800 February 16th, 1896. Autopsy. Jan. Feb. 28, 000 24 28 30 4 5 8 9 10 11 12 13 14 26,000 A 24 000 > / \ 7 20 000 ; ' _L A / 18 000 1 V-7- v -LO, \J\J\J 16 000 I rr 14,000 1 \ j / \l . f> f\f\(\ / JL ' X" V / Died o 1 15th 10,000 / FIG. 36. Chart of Leucocytes in a Case of Cancer of the Liver. CANCER OF THE OMENTUM, ETC. 347 CANCER OF THE INTESTINE. Here the counts range both high and low. Hay em : found cancer of the rectum to show only 9,500 leu- socytes. Eeinbach 2 found in three cases of cancer of the rectum moderate leucocytosis. 3 Only four of the ten cases in our series (see Table XLII.) showed leucocytosis, and in one of these there was a complicating pylephlebitis which probably raised the count. The red cells show little change. TABLE XLII. CANCER OF THE INTESTINE. Age. M i Red cells. White cells. Per cent haemo- globin. Remarks. 56 41 31 33 59 66 50 58 34 52 M. F. M. M. F. M. F. M. M. M 4,408,000 5,560,000 4,921,000 4,368,000 4,800,000 4,268,000 5,416,000 4,160,000 12,700 5,800 8,800 5,800 5,500 7,150 12,000 15,200 15,500 7 400 60 45 83 33 78 50 40 55 Cancer of duodenal papilla with pylephlebitis. Autopsy. Cancer of caecum . Operated successfully. Cancer of hepatic flexure. Operated. Cancer of colon. Operated. Cancer of caecum. Autopsy. Cancer of intestine (where ?J. Cancer of rectum. Cancer of rectum (operation). Metastases, pi imary in sigmoid. 47 F 9 300 63 28 52 M. M. 2,424,000 2,440,COO 5,300 7,800 6,800 72 Cancer of caecum. Operation. Cancer of caecum. No digestion leucocytosis. CANCER OF OMENTUM AND ABDOMINAL ORGANS GENERALLY. The nine cases seen at the Massachusetts General Hospital in which cancerous tissue was pretty generally distributed through the abdominal organs, all showed leucocytosis with two exceptions (see Table XLIIL, A). 1 Loc. cit. 2 Loc. cit. 3 Apparently that is, the percentage of adult cells was increased, did not count the leucocytes as a whole. He 348 SPECIAL PATHOLOGY OF THE BLOOD. TABLE XLIIL, A. CANCER OF OMENTUM AND ABDOMINAL ORGANS GENERALLY. Age. M i Red cells. White cells. Per cent haemo- globin. Remarks. 50 F Greatly Markedly 48 M increased. 7 250 diminished. of 400 cells: Poly. ,'84.5 per cent; small lymphocytes, 8; large lymphocytes, 5; eosinophiles, 2.5. Autopsy. May 13th 6,500| May 20th No digestion leucocytosis 42 26 M. M 4,560,000 7,300 f 7,800 10,600 9 000 60 October 13th. Poly., 88 per cent; lym- phocytes, 10; eosinophiles, 2. October 16th. 65 Adult M. M 3,772 000 11,700 13700 Adult. F 5,500,000 26,200 Autopsy 45 F 27400 Adult M Greatly Markedly Differential count of 500 cells* P'oly 80 increased. diminished. percent; lymphocytes, 20. CANCER OF THE KIDNEY. Of five cases which I have examined (see Table XLIIL, B) all showed very large leucocyte counts viz., 25,000, 27,000, 28,500, 43,100, 82,000, and 91,000, an average of 54,000. In three of these cases, however, the tumors may have been sarcomata, as no microscopic examination was made. Most of the cases had fever, chills, and signs of inflammation, which may account for part of the leucocytosis. TABLE XLIIL, B. CANCER (OR SARCOMA) OF KIDNEY. Age. y, 36 F. F, 4,500,000 3.248.000 25.000 32.800 62 Operation. Operation. CANCER OF THE PROSTATE. 1| 45 JM. | | 10,200 | | CANCER OF THE LIP. 1 I 51 | M. | 7,000,000 | 6,300 | | CANCER OF THE BREAST. 31 f F. F. 6,000,000 8,000 Not Differential count of 600 cells: Poly., 72.4 increased per cent; lymphocytes, 25.4; eosinophiles. 2.2. ? F. Marked Differential count of 400 cells: Poly., 89 per increase. cent; lymphocytes, 11. CANCER OF THE NECK. 42 M. Marked Poly., 88.5 per cent. increase. Loc. cit. 2 Loc. cit. 3 Loc. cit. 350 SPECIAL PATHOLOGY OF THE BLOOD. TABLE XLIV. , B (Continued). CANCER OF THE PANCREAS. Age. X Red cells, White cells. Per cent haemo- globin. Remarks. J 54 M 18300 70 Metastases 2 56 M 17,600 Liver and spleen also 8 64 M. 15,900 General peritonitis. CANCER OF VERTEBRJE. 62 |F. | | 13,200 | | ADENOMA OF THE SUPRARENAL BODY. 59 | M | 24,200 Autopsy. Cancer of the Up has apparently been neglected so far as blood examination is concerned. Hayem, Kieder, and Keinbach give but one case each, the counts being respectively 7,000, 11,- 600, and "not increased." In a single case at the Massachu- setts Hospital I found 6,300. The following scattered counts may be added: Cancer of tongue, 7,000 (Hayem); cancer of scrotum, 6,700 (Hayem); cancer of navel, 7,100 (Hayem); cancer of larynx, 7,200 (Hayem), 16,000 (Keinbach); cancer of ovary, 25,000 (Massa- chusetts Hospital) and "no increase" (Eeinbach); cancer of neck, 20,000 (Massachusetts Hospital) and "no increase" (Eein- bach); cancer of pancreas: Hayem, 2 cases 9,400 and 9,900; Schneider, 1 case 12,000; cancer of vagina, 9,800 (Eieder) ; cancer of penis, 7,000 (Hayem); cancer of thyroid, 70,000 (Hayem) (a very rapidly growing tumor) ; cancer of media- stinum, "marked increase" (Eeinbach); cancer of prostate, 1 10,200. Qualitative Changes in the Leucocytes. 1. The percentage of polymorphonuclear neutrophiles is usually high in cases with leucocytosis and normal in those without it. This rule holds for perhaps three-fourths of the cases, but there are many exceptions to it. For instance, Taylor 2 reports 27,840 leucocytes with 65.6 per cent polymorphonuclear cells, 14,800 leucocytes with 66.2 per cent polymorphonuclear 1 Braun (Wien. med. Woch., 1896, p. 582) mentions a cancer of the prostate in which the leucocytosis instead of being made up mostly by the adult leucocytes, was associated with a large increase of the small lympho- cytes together with numerous eosinophilic myelocytes. 2 Taylor: Internat. Med. Mag., July, 1897. THE BLOOD IN CANCER. 351 cells, 25,000 leucocytes with 58.2 per cent polymorphonu- clear cells, 45,000 leucocytes with 43.7 per cent polymorpho- nuclear cells, the last a marked lymphocytosis. On the other hand, he found 88.7 per cent of the polymorphonuclear cells in a total leucocyte count of 3,000. My own experience is similar i.e., 88 per cent of polymorphonuclear cells with a total count of 7,800 leucocytes, though I have never seen so marked a lymphocytosis as was present in 'Taylor's cases. He also noted a relative increase in the large lymphocytes which my counts have not shown. Keinbach found in 8 cases with leucocytosis 89 per cent in 2 cases and 87, 86, 83, 81, 80, and 77 per cent in others. In the Massachusetts General Hospital series the following per- centages occurred: When no leucocytosis was present 88.7, 88, 86, 79, 66, 62.5, 62, 60, 57 per cent, etc. With leucocytosis, 96, 98, 92, 90, 90, 88, 87, 86, 84, 83, 74 per cent, etc. (See Tables XXXIX. , XLL, XLIIL, XLIV.). 2. Eosinop Idles are not always notably decreased (as they are in many other leucocytoses) nor are they increased except when bone metastasis occurs (see below). In Keinbach 's 16 cases the average percentage was 2 + per cent. In the Massachusetts Hos- pital cases the average was 1.2 per cent, but in 7 of the 38 cases in which differential counts were made, no eosinophiles were seen. 3. Myelocytes. Perhaps more commonly than in other con- ditions except leukaemia and pernicious anaemia, we find in ma- lignant disease small percentages of myelocytes, as the following cases show : CASE I. Extensive abdominal cancer; great cachexia. Six hundred cells showed: Polynuclear neutrophiles. .. 89.4 per cent Lymphocytes 10. Eosinopbiles 1 " Myelocytes (3 in 600 cells) 5 " CASE II. Cancer of uterus; marked cachexia and leucocy- tosis. One thousand cells showed : Poly nuclear neutrophiles 82.3 per cent. Lymphocytes 17.3 " Myelocytes (4 in 1,000 cells) 4 CASE III. Cancer of uterus; died two days later. Eed cor- 352 SPECIAL PATHOLOGY OF THE BLOOD. puscles, 7,000,000; white, 62,000. Considerable stasis helps to explain the count. Differential count of 500 cells showed : Poly nuclear neutrophiles 93 per cent. Lymphocytes 6 " Eosinophiles " Myelocytes (5 out of 500) 1 " CASE IV. Cancer of liver, jaundice and cachexia; died soon after. Differential count of 500 cells showed : Poly nuclear neutrophiles 92. per cent. Lymphocytes. 6. " Small myelocytes. 1.2 " Large myelocytes (4 in 500) 8 " CASEY. Cancer of abdomen ; cachectic. Differential count of 1,000 cells showed: Polynuclear neutrophiles 82. per cent. Lymphocytes 16.6 " Eosinophiles 1. " Myelocytes (4 in 1,000) 4. " CASE VI. Cancer of stomach, liver, etc., with perforated stomach; cachexia. Leucocytes, 105,000. Fifteen hundred cells showed : Polynuclear neutrophiles 90. 7 per cent. Lymphocytes , 4. 8 " Eosinophiles .2 " Myelocytes (68 in 1,500) 4.3 CASE VII. Cancer of uterus ; cachexia. In 1,000 cells there were: Polynuclear neutrophiles 88. per cent. Lymphocytes 11,7 " Eosinophiles .2 " Myelocytes 1 CASE VIII. Cancer of kidney; great cachexia. In 1,000 cells there were : Polynuclear neutrophiles 92.9 per cent. Lymphocytes 6. 2 " Myelocytes , 9 SARCOMA. 353 CASE IX. Cancer of kidney ; great cacliexia. Leucocytes, 27,000. Five hundred cells showed: Polymiclear neutrophiles 66. per cent. Lymphocytes 29.5 " Eosinophiles 2. " Myelocytes 2.5 " CASE X. Cancer of liver. Five hundred cells showed : Polynuclear rieutropniles 92. per cent. Lymphocytes 5. 8 " Eosinophiles 2 " Myelocytes 2. " About one-half of all the cases of cancer examined by me have shown myelocytes. Epstein (Wiener med. Presse, December, 1894) in a case of cancer with metastatic bone nodules noticed large numbers of nucleated corpuscles (normoblasts and megaloblasts) and mye- locytes, but I think the association was a mere coincidence, since I find that myelocytes and erythroblasts are very commonly present in cacliexia from any cause. SARCOMA. In general the effects of sarcoma are like those of cancer, but worse. Great anaemia and higher leucocyte counts are the rule. The literature of the subject is rather scanty. Red Cells. Hay em in a case of osteosarcoma counted the r^d cells at 663,400 per cubic millimetre. Laker 1 describes an " abdominal cystosarcoma" in which two counts of red cells showed 2,800,000 and 2,500,000. Yon Limbeck 2 in 1 case found 1,118,000, and in another 2,- 240,000. Both were osteosarcomata. Sadler 3 in 3 cases found 2,710,000, 3,637,000, 4,500,000. Riecler* in 3 cases (all osteosarcomata) found 1,846,160, 3,- 770,000, and 3,995,000. The Massachusetts Hospital blood counts include 15 cases in which the red cells were counted (see Table XLY., A and B), the average being 4,400,000, not nearly so low as that recorded. 1 Wien. med. Woch., 1886, p. 926. 2 Loc. cit., p. 343. 3 Loc. cit. , pp. 88, 3D. 4 Loc. cit., pp. 98, 100. 23 354 SPECIAL PATHOLOGY OF THE BLOOD. by other observers; still low counts occurred (2,706,000, 2,637,- 000, 3,842,000). The qualitative changes in the red cells consist (as in cancer) of the "degenerative" changes (deformities in size and shape, englobular changes) present in marked cases, and the presence of nucleated corpuscles, when cachexia is marked. TABLE XLV., A. SARCOMA WITH LEUCOCYTOSIS. Age. M Red cells. White cells. Per cent haemo- globin. Remarks. 4,188,000 98000 Polynuclear cells, 90. 2 per cent 4312000 25 000 32 Polynuclear cells, 80 9 per cent 4,000,000 44,600 42 Polynuclear cells, 70 per cent (infant of twenty 21 F 2 706 000 56 000 months). Sarcoma of kidney Autopsy 35 F. 4,560,000 17,000 23900 65 Melanotic sarcoma all abdominal organs (bone metastasis ?). November 30th, 1895. Differ- ential count of 600 cells: Poly., 71 per cent; small lymphocytes, 11 ; large lymphocytes, 5.2; eosinophiles, 12.4(1); myelocytes, 4. December 7th 33,400 37,900 41,200 33 000 December 13th. December 19th. December 22d 36,000 December 26th 40200 January 14th 46 M 4700000 55,400 16000 January 28th. Sarcoma of abdominal organs 19 000 Three days later Autopsy 32 24 M. M 2,630,000 2,900,000 4,352,000 24,000 21,000 13600 50 General sarcomatosis. One week later. Autopsy. Sarcoma of kidney. 41 68 M. F 3,842,000 6,200,000 61,100 16,000 55 Sarcoma of lung, etc. Autopsy. Sarcomatosis. 48 M Marked Differential count of 700 cells Poly 70 per cent 57 M. 4,180,000 in- crease. 13,000 16250 47 lymphocytes, 22 ; eosiuophiles, 1 ; myelocytes, 7. Sarcomatosis. Melanotic sarcoma of abdominal organs. One week later Adult. M 15,180 18,000 Sarcoma of abdominal organs. Great Osteosarcoma (thigh) Differential count of 500 1 in- crease. Great cells: Poly., 74 per cent; small lymphocytes, 19; large lymphocytes, 6; eosinophiles, 1. Sarcoma of abdominal organs Differential in- crease. 13 200 count of 800 cells : Poly. , 84 per cent ; lympho- cytes, 15.5; eosinophiles, 5. 36 M. 6,700 TABLE XLV., B. SARCOMA WITHOUT LEUCOCYTOSIS. Per Age. Red cells. White cells. cent haemo- Remarks. 1 globin. 29 M. 5,280,000 8,200 Sarcoma of testicle. 37 ? F. M. 4,980,000 4,946,000 9,000 9,000 78 Sarcoma of ovary. Osteosarcoma of shoulder. 24 M 4,952,000 6,000 Small recurrent sarcoma of groin. Small tumors are often without any effect on the blood (see SARCOMA. 355 Table XLV., B). According to v. Limbeck 1 this is oftener true than in cancer. Hemoglobin. KeinbachV 20 cases ranged between 23 and 75 per cent, averaging' 52 per cent. Bierfreund 3 in 29 cases found variations between 40 and 75 per cent. Von Limbeck's 2 cases had 28 and 48 per cent respectively. Kieder's*4 cases showed at the beginning of treatment 29, 56, 57, and 65 per cent respectively, but in 1 case the haemoglobin went down gradually while under observation until it reached 6 per cent (!), the lowest point, Eieder says, that he has ever seen in any disease. Sadler's 5 cases showed 33, 45, and 78 per cent. In the 5 cases of Table XLV. in which this point was noted, the average is 58 per cent. On the whole, the coloring matter seems to be more dimin- ished than in most cases of cancer. Leucocytes. The following tables, slightly modified from v. Limbeck, show the important points. No. Observer. Diagnosis. Count. 1 . Hay em. Osteosarcoma. 11 250 2 Alexander. 52 700 3 16 430 4 16 275 5 17,050 15 900 6 15,570 13 020 7 10 950 8 .. 12,090 11 248 9 Rieder. 12 700 10 10,900 9 100 11 8 000 12 v Limbeck 32 000 13 .. u Reinhach. 26,800 20, 000 14 u 13 000 15 Massachusetts Hospital 21 000 16 u 9 000 Average, 17,000 1 Loc. cit. 4 Loc. cit. 2 Loc. cit. 5 Loc. cit. 3 Loc. cit. 356 SPECIAL PATHOLOGY OF THE BLOOD. No. Observer. Diagnosis. Count. 1 Hay em. Lymphosarcoma. 11,700 2 Alexander 19,910 3 19,530 4 11,696 5 11,470 6 10 540 7 . . v. Limbeck 55 100 8 38 000 9 10 800 10 Sadler. 33, 248 11 19,299 12 9,044 Average, 20,000-f No. Observer. Diagnosis. Count. 1 ., Rieder. Melanosarcoma. 41,600 2 28 500 3 u 22 300 4 ... . Reinbach. 25 000 5 u 8 000 6 Massachusetts Hospital 37 900 7 u u 13,000 Average, 25, 100 -f For other sarcomata, see Table XLY., A and B. On the whole, leucocytosis appears to be more constant and of greater extent in sarcoma than in cancer. Qualitative Changes. 1. The increase of polymorphonuclear leucocytes which we find in most forms of leucocytosis is not always present in sar- coma ' and seems to be less frequent than in cancer (see Cases 5, 11, 14, Table XLY.) As in cancer, it may be present when no increase in the total leucocyte is to be found, and may be the only indication of any disease in the organism. 2. A few cases are on record in which a large percentage of eosinophiles has been present. Reinbach found 48 per cent of eosinophiles in a case of sar- coma of the neck with sloughing and ulcerative endocarditis, the 1 Palma (Deut. Med. Woch. , 1892) reports lymphocytosis in sarcoma. SARCOMA. 357 percentage continuing over 40 for several weeks. 1 Autopsy showed sarcomatous nodules in the bone marrow. In another case, a tumor of the abdomen, the eosinophiles were 10.5 per cent, and in two others 8 per cent. A case of apparent sarcoma of the abdominal organs (no autopsy) at the Massachusetts General Hospital in January, 1896, had 12.4 per cent of eosinophiles. Such cases should certainly make us think of bone metas- tases, and Neusser speaks of osteosarcomata as being accom- panied by eosinophilia, but the evidence is as yet fragment- ary. 3. Myelocytes. Keinbach's case just described had a low per- centage of rnyelocytes. The following cases illustrate the same point : CASE I. is a case of sarcomatosis in a man in whom sarcoma- tous nodules were distributed all over the internal organs and in the skin. A differential count of 700 white cells showed in his case: Typical myelocytes (over 15^) 2 per cent. Small myelocytes (under 15/*) 5 " Lymphocytes 22 " " Polynuclear neutrophiles" 70 " Eosinophiles 1 " The autopsy showed no special lesions in the spleen, glands, or bone marrow, except those due to the sarcomatous nodules. 1 The full counts are as follows : April 4th, 1892. May 20th, 18 tt, Red cells 5 396 000 Red cells 4 512 000 White cells 120 000 (') White cells 52 000 Haemoglobin. 60 per cent Haemoglobin 55 per cent. DIFFERENTIAL COUNTS. April 4th. May 1st. May 20th. May 26th. Poly, neut Per cent. 48 Per cent. 51 Per cent. 664- Per cent. 51 4- Eosinophiles . . 48 46 42 444- Lymphocytes .... 2 7 2 32 1 5 3.2 Myelocytes 1 .68 .64 .8 358 SPECIAL PATHOLOGY OF THE BLOOD. CASE II. Sarcoma of abdominal wall. Differential count of 800 cells showed : Poly nuclear neutrophiles 84. per cent. Ly mphocy tes 10. 5 " Large lymphocytes 5. " Eosinophiles 2 " Myelocytes 3 CASE III. (No. 2, Table XLV., A). Six hundred cells con- tained : Polynuclear neutrophiles 71. per cent. Lymphocytes 16.2 " Eosinophiles 12.4 " Myelocytes 4 " Summary of Blood Changes in Malignant Disease. 1. Small, slow-growing tumors and the early stages of all tumors may have no effect on the blood appreciable by our present methods of examination. 2. In advanced cases the red corpuscles often become thin, light, and pale, and finally their number may be greatly de- creased, the counts running sometimes as low as in pernicious anaemia. In this respect, as in others, sarcomata seem to in- jure the blood more than cancers. 3. The color index is always below 1, but is rarely as low as we find it in severe chlorosis. 4. Normoblasts and megaloblasts (the latter being in the minority) may occur, the former even in the absence of severe anaemia. Deformities in size and shape are common. 5. Leucocytosis is present in the cachectic end-stages of many cases, but is frequently absent in small tumors of slow growth and without metastases. The polymorphonuclear cells are often relatively increased. 6. Fibrin is not increased. Diagnostic Value. 1. When we are dealing with an obscure, deep-seated dis- ease, if hemorrhage is excluded, the presence of persistent leuco- cytosis suggests suppuration or malignant disease (rather than tuberculosis or syphilis, for example), and excludes any simply MALIGNANT DISEASE. 359 functional or hysterical affection. The absence of leucocytosis, however, does not exclude malignant disease, though it makes suppuration very unlikely. 2. Between malignant disease and suppuration if the other signs and symptoms do not decide there may be nothing in the blood to decide. In decided pyaemia we may get pyogenic cocci from the blood by culture, but a negative result would not ex- clude the suppurating focus. The absence of any increase of fibrin in the blood speaks against suppuration, and therefore in favor of malignant disease ; but the presence of increased fibrin network is not decisive either way, as it may be met with in connection with neoplasms, though more common in suppuration. 3. Between malignant disease and hemorrhage a marked anaemia favors the latter, provided the case is a recent one ; for the anaemia of malignant disease is comparatively slow to develop. The leucocytes give no help. 4. Between cancer and ulcer of the stomach, if there has been no recent hemorrhage, leucocytosis favors cancer; but its absence is of no weight either way. The haemoglobin is said to decrease steadily in cancer, while in ulcer it tends to return toward normal after the cessation of hemorrhage. The presence and persistence of digestion leucocytosis speak against cancer, and its absence in favor of cancer. It must be remembered, however, that any variety of catarrh or dilatation (should such be present) can also prevent digestion leucocytosis, and that the latter is not invariably present even in health. 5. Between cancer of the liver or bile ducts on the one hand and simple gall-stone colic or gall-stone obstruction, the presence of leucocytosis favors cancer. As usual, however, its absence does not exclude cancer, and we must bear in mind that gall stones with cholangitis may raise the leucocyte count as much as cancer. Simple cysts or echinococcus cysts cause no leucocy- tosis, nor does syphilis of the liver. 6. The appearance in the blood of large numbers of eosino- philes, myelocytes, and nucleated corpuscles during the course of a malignant disease points to a bone metastasis. 7. When a leucocytosis which has disappeared after removal 360 SPECIAL PATHOLOGY OF THE BLOOD. of a neoplasm reappears, we may expect recurrence of the growth shortly. 8. A steadily increasing leucocytosis in a case of malignant disease points to a rapidly growing tumor or to the occurrence of metastasis. 9. Between malignant disease and pernicious anaemia the diagnosis rests on the following points : I. Color index low in malignant, apt to be high in per- nicious anaemia. II. Leucocytes often increased in malignant, diminished in pernicious anaemia. III. Lymphocytes often decreased in malignant, increased in pernicious anaemia. IV. Average size of red cells often decreased in malignant, and often increased in pernicious anaemia. Y. If nucleated red corpuscles are present the normoblasts are in a majority in malignant disease, and in a minority in per- nicious anaemia. 10. The presence of leucocytosis is against the benignness of any tumor. 11. When no actual increase of leucocytes is present, an in- creased percentage of the polymorphonuclear variety among those present may have the same significance as a leucocytosis. CHAPTEE XI. BLOOD PARASITES AND INTESTINAL PARASITES. EXAMINATION FOR THE PLASMODIUM MALARLE AND ITS PRODUCTS. I. Time for Examination. It is often stated that the organism is most easily found during the chill. But this is not the writer's experience. During a chill it is often difficult and sometimes impossible to find the organisms. Eight hours before or after a chill is the most favorable time (Thayer), although parasites have been found as late as forty -eight hours after the last chill. During the chill many organisms retire to the internal organs. The number of organisms varies a great deal. In some cases they are present in every field of a one-twelfth immersion lens, while in others we may find only one after an hour or more of patient search. In the majority of the cases occurring near Boston, it needs but a few minutes' search to find them if the blood be taken within twelve hours before or after a chill, and provided no quinine has been lately given. Occasionally in mild cases the organisms are very scanty ; and it may be almost impossible to find any. The quartan and sestivo-autumnal forms of malaria are so rare in New England that I shall not attempt to describe in detail the parasites found in them, but shall con- fine myself mostly to the parasites of common tertian and double tertian fevers with which I am personally familiar. II. Method of Examination. A slide of fresh blood is pre- pared as above described (pages 6-8) and examined with a one- twelfth immersion lens. 1 Lower powers should not be used, although in skilful hands they are often sufficient. Portions of the slide in which the corpuscles do not overlie each other should be chosen for examination. As we pass the slide along beneath the lens it is well to be on the lookout for any specially large or specially pale corpuscle. Such a one will catch the eye if we 1 In cold weather both slide and cover should be warmed before using. Indeed this is always well, as it makes the corpuscles spread better. 362 SPECIAL PATHOLOGY OF THE BLOOD. are on the watch for it, even though the slide is being passed along very rapidly, and all such should be carefully examined. Another thing to watch for is anything black or dark brown. If the slide is not perfectly clean, or if the cover-glass has touched the skin in collecting the blood, there will often be black spots which make us pull up short and examine, only to find that they are bits of dirt. This loses time, and hence, as above noted, the importance of care and cleanliness in the earlier stages of the process. Besides any strikingly pale or swollen corpuscle or any black dots, we should be on the lookout for any movements in the field. The movements of Miiller's "blood-dust" (see page 59) are often mistaken by beginners for those of the malarial organ- ism. Their greatly smaller size and extracorpuscular position serve to distinguish them in most cases. I have sometimes thought I saw pigment in these bodies. If, as Stokes believes, the " blood-dust" is derived from the leucocytes, it is possible that they might carry out with them some pigment ingested by the leucocyte. III. The Malarial Organism. (a) "Hyaline Forms." In the earlier stages of its growth, i.e., during and soon after the chill, the organism is not pigmented, but appears only as a light spot in the pale greenish-yellow of the corpuscle. It practically is never to be seen outside the corpuscle. Most malarial organ- isms are to be found within the corpuscle, and only there. l For those who have not examined many specimens of ma- larial blood it is a very difficult thing to find the organism at this stage of its growth, and the number of mistakes in diag- nosis is very large. I always look with great suspicion on any report of malarial blood as containing only "hyaline forms." In the later stages, when the organism has become well pig- mented, there is nothing that at all resembles it, and those who have seen and watched it a few times can hardly mistake any- thing else for it. Not so with the so-called " hyaline" or young- est form of the organism. Personally I think the name " hyaline bodies" is responsible for a part of the mistakes. We are led to expect something more shiny and refractile than the organism 1 Except degenerate forms, free flagellse, and spores at the moment of segmentation (rarely to be seen). Crescents and ovoid bodies are inter- cellular. BLOOD PARASITES AND INTESTINAL PARASITES. 363 really is, and so are misled by the brilliant white circles to be found at the centre of many normal corpuscles under certain conditions of light and partial drying up. Time and again I have been asked to look at malarial organisms (always the "hyaline" forms), and found nothing more than one of these effects of light which can be found in any normal blood, if the conditions are right. There are certain marks by which we can exclude these artifacts from consideration : I. They are generally far too numerous to be malarial or- ganisms. One usually finds a dozen or more in a field which would be almost unheard of with the plasmodium malarise. II. They are generally in the centre of the corpuscle, while the young malarial organism is almost never at the centre. III. They are almost invariably round, the malarial organism being generally more irregular and branching. IV. They seem to increase and diminish in size as we focus up and down upon them, while the malarial organism only grows dimmer or clearer. V. They are, as before mentioned, more brilliantly white and shiny than the malarial organism, which has often a faint tinge of yellow, although much paler than the surrounding corpuscle substance. VI. Their edges are sharper, the malarial organism often fading off very gradually into the corpuscle color. VII. Their movement is different. The malarial organism is not at all the only thing to be seen moving in the blood, as has sometimes been stated. The red corpuscles have the Brown- ian motion, and as they begin to crenate often move very actively. But their motion is very different from that of the hyaline malarial organism, for the latter changes both its shape and its position in the corpuscle quite rapidly, while the motion of the light space in an ordinary red cell is a wavy undulation of the outlines back and forth without any considerable change of shape. (6) As soon as the organism gets any pigment (and there are very few times in the cycle of a malarial case when there are not some pigmented organisms present), the active rapid motion of the black pigment dots is unlike anything else seen in the blood, and when once recognized can never be forgotten or mistaken. It is only when the pigment has ceased moving (owing to the 364 SPECIAL PATHOLOGY OF THE BLOOD. death of the organism) that the differentiation between dirt and malarial pigment becomes difficult. Sometimes it is really difficult to distinguish motionless pig- ment in a malarial organism from dirt even on careful scrutiny. The best way is to get a fresh slide when the pigment is in motion. To any one fairly familiar with the appearance of pigmented forms of malarial organisms, failure to find them in a case of malaria is due generally (1) to too thickly spread a layer of 1 ^ 34- 10 FIG. 37. The Parasite of Tertian Malaria (after Thayer). 1, Normal red cell ; 2 and 3, hyaline parasites; 4, 5, 6, 7, pigmented forms; 8, 9, 10, 11, segmentation. blood, the corpuscles overlying each other; (2) to not looking long enough (Figs. 37 and 38) ; (3) to lack of proper light. I have not attempted to go into the marks by which we can differentiate the tertian, quartan, and eestivo-autumnal forms of the organism for clinical evidence usually suffices to determine this point. For information on this and all the finer points in regard to the life history and habits of the organism W. S. Thayer 's admirable monograph should be consulted. Here it is sufficient to say that as the paroxysm draws near, the pig- ment granules begin to work in toward the centre in radiating lines until they are all collected in a solid black mass. While this is going on, the pigment granules not infrequently gather into short rod-like masses not at all unlike bacilli. BLOOD PARASITES AND INTESTINAL PARASITES. 365 Round the central mass of pigment, indistinct radiating divisions may sometimes be seen just before the organism breaks up. These divisions have been compared to the petals of a flower, but it is very difficult to see more than the faintest indi- 10 13 15* 11 FIG. 38. Parasite of Quartan Malaria (after Thayer). 1, Normal red cell; 2, hyaline parasite; 3 to 11, pigmented forms; 13 to 15, segmenting forms. o j 11 14 13 FIG. 39. Parasite of JEstivo-Autumnal Malaria (after Thayer). 1 to 6, Young forms; 7 to 13, mature forms; 14 to 16, segmenting forms. 366 SPECIAL PATHOLOGY OF THE BLOOD. cations of such an arrangement in most specimens. The cor- puscle itself is by this time wholly lost to sight. (c) The next stage, that of segmentation, is less commonly seen than those just mentioned, and is only to be satisfactorily observed by using a warm stage (vide supra, page 8) and spend- ing considerable time on the watch for it. Around the central pigment mass we may sometimes see in ordinary specimens (without warm stage) the faint outlines of a group of small spherical, colorless bodies (vide Fig. 2, 9, Plate I.) which are the new generation of young organisms. Now we should expect that with the next step in the process we should find these young plasmodia free in the plasma or en- tering a fresh set of red corpuscles. But in the peripheral cir- culation this is rarely if ever observed. Thayer in his immense experience has never seen them. The next evidence we have of the organism is as a "hyaline" body inside the corpuscle again. Almost all stages of the growth of the plasm odium which we can watch in the blood drawn from the peripheral circulation take place within the corpuscle. It is true that as the pigrnented organism gets towards its full growth, and before the granules have begun to gather at the centre, we may find it very difficult to find any trace of corpuscle substance around the margin of the plasmodium. Sometimes we see a ring of non-pigmented glistening white substance outside the moving black dots (see FIG. 40. Flagellate Malarial Organisms (after Thayer). Fig. 2, 7, Plate I.) standing out light against the darker plas- ma. This I suppose to be the remains of the corpuscle. It Is not described or pictured in the standard works on the sub- ject. Occasionally we do find pigmented bodies wholly outside the corpuscle, either partly or fully grown. In the intracorpus- BLOOD PARASITES AND INTESTINAL PARASITES. 367 cular forms the distinction between plasmodium and corpuscle substance is not, I think, so sharp and clear as one would be led to expect from the plates in standard works. With aver- age eyes and lenses the out- line of the organism, as distinguished both from its pigment granules and the surrounding corpuscles, is not easy to see. It is the moving pigment granules that attract our notice. (d) It remains to speak of three comparatively small points : 1. The presence of fla- gella. 2. Pigmented leucocytes. 3. Crescents and o voids. 1. Toward the end of the life history of a malarial par- asite, it sometimes makes its presence very obvious in the microscopic field by knock- ing about the surrounding corpuscles with its arms or "flagetta." Exactly why and under what conditions it shows or fails to show these appendages is not known. 1 They are about two or three times as long as a red cor- puscle and one-sixth or one- eighth as wide. They are usually to be inferred rather than directly seen, as they are nearly transparent. Our 1 McCallum has recently offered interesting evidence that they are sexual organs. FlG " 41 - Fla s ellate Malarial Organisms. (After Manson's photographs.) 368 SPECIAL PATHOLOGY OF THE BLOOD. attention is attracted by an active motion among a group of red cells apparently of spontaneous origin. Gradually we make out a filmy whip-like tail attached to an adjacent malarial parasite. Sometimes there is pigment dotted along the flagellum itself, and then we can make it out more easily. Its distal end is especially apt to be pigmented, and by the help of this pig- ment we make out that it is bulbous, while similar swell- ings can sometimes be seen at other points along the flagel- lum (see Fig. 41). Such a flagellum may break off and dart about free among the corpuscles. As the pigmented end is sometimes all that we can see of it, this gives rise to the appear- ance of a very small, actively locomotive pigmented body free among the corpuscles, and its course may be followed through several fields. When the flagella have ceased moving, their presence is gen- erally detected, if at all, by an irregular line of pigment dots about 20 1*, long, which will be shown by careful focussing to be contained within a nearly transparent membrane. Very often we find a leucocyte in process of closing round the flagellated parasite. Manson has lately succeeded in stain- ing the flagellae, and the accompanying photographs are from his stained specimens. 2. Pigmented leucocytes, containing the whole or part of malarial organisms or simply blocks or granules of black pig- ment, are usually to be found in the blood near the time of the chill. The pigment is to be carefully distinguished from the granules present in most leucocytes, which in certain lights look quite dark even if unstained, dark enough to be mistaken for pig- ment by the untrained eye. Careful focussing and changing the light will easily determine which we are dealing with, provided we are familiar with the appearances of leucocytes in the fresh unstained blood. In certain forms of the disease in which the or- ganisms themselves retire to the internal organs, the presence of pigmented leucocytes may be the only evidence of the disease to be found in the peripheral blood and is therefore of the greatest importance. 3. Crescentic forms are not often seen in New England. They are found only in the sestivo-autumnal forms of malaria which occur chiefly in the South and West and have been seldom reported in any Northeastern State except in patients who have BLOOD PARASITES AND INTESTINAL PARASITES. 369 brought them from the South and West. I have never seen these crescentic forms except in the stained specimens of other obser- vers, and my ideas of them are mostly second-hand (Fig. 42). Full account of them will be found in the monograph of Thayer's above referred to. Hitherto I have spoken wholly of the appearance of the par- asites in the fresh unstained blood, this being by far the simplest, easiest, and surest way of finding them and the only way of studying their development. In cases in which we cannot make a microscopic examination at the bedside, we can sometimes pre- serve the organism alive between slide and cover-glass, until we can get it to the nearest microscope, even if this takes sev- eral hours. I have carried specimens in my handbag a whole morning and yet found the pigment of the malarial parasite in motion at the end of that time. Warm weather favors this. When it is necessary to keep the specimen some time before examination, it is best to paint on the slide a ring of vaseline or any gummy substance, and allow the drop of blood to spread out inside this ring so that the margins of cover glass are glued to the slide by the oily substance and the entrance of air is pre- vented. The cedar oil ordinarily used for immersion lenses answers the purpose very well. Both slide and cover should be gently warmed before spreading the drop of blood. Many physicians who cannot possibly carry a microscope about with them can easily find room for a few slides and cover- glasses and they may be of great service. When specimens have to be sent by mail, or for long dis- tances, or in cold weather we have to fall back on dried speci- mens prepared as described on page 43, provided always that a 24 870 SPECIAL PATHOLOGY OF THE BLOOD. bedside examination is impossible. These can be stained by one of the following methods : Leave the specimen for half an hour or more in equal parts of ether and absolute alcohol, dry them in the air, stain for from one-half to five minutes in a one-half-per-cent solution of eosin in sixty-per-cent alcohol, wash in water, dry and stain one -half to one minute in concentrated watery solution of methylene blue ; wash again in water, dry in filter paper, and mount in Canada balsam. Personally I have found this method rather unsatisfactory on account of the different intensity of different eosin stains and the consequent need of finding out by experiment how long (within the limits of one-half to five minutes) the specimen is to be stained before a distinct yet not violent red color is at- tained in the protoplasm of the corpuscles. The blue stains the plasmodium itself in contrast with the pink corpuscle substance around it; the pigment granules remain, as in the fresh specimen, black or brownish black. (See Fig. 3, Plate I.) Some find the stain of Plehn simpler and more satisfactory as well as quicker. By this method we leave the specimens only three or four minutes in absolute alcohol and then stain five or six minutes in the following mixture : Concentrated watery solution methylene blue 60 One-half-per-cent solution of eosin in seventy-five-per-cent alcohol 20 Distilled water 40 Twenty per cent NaOH 12 gtt. Wash in water and mount in Canada balsam. The trouble of double staining and the uncertainty as to the length of time are avoided by this solution, and the parasites are sometimes beautifully stained. Yet on the whole I have had better success with the eosin and methylene blue de- spite its difficulties. The ordinary Ehrlich-Biondi mixture may also be used to demonstrate pigmented forms. The organ- ism itself does not stain at all with this mixture but stands out light against the yellow of the corpuscle, the pigment looking as it does in the live parasite. The hyaline forms need some other stain for satisfactory recognition, but it is sometimes con- BLOOD PARASITES AND INTESTINAL PARASITES. 371 FIG. 43. venient to use the same stain for the differential count and the malarial organism, as for instance when we have only one cover-glass preparation in a case of doubtful diagnosis. Fixing the specimen in alcohol and ether is here far better than heat; other- wise the technique is as above described under Triple Staining (page 44). The general appear- ance of the organism so stained is shown in Fig. 43. If the organisms are fairly num- erous and the technique is good we can find them by this method even in preparations months old. In general, however, it is very in- ferior to the examination of the live organism in the fresh blood, and gives many more chances for error. So much for technique. We often hear reports of fruitless search for the parasite in the blood of malarial patients, but the regularity with which they are found at all the larger hospitals and by all practised observers in this and other countries leaves no doubt that they are to be found in every case during some portion of the cycle. The practice of taking blood during a chill contributes, I believe, to the number of unsuccessful endeavors to find the organism ; as mentioned above, this is the worst, not the best time to look for them. Too thick a layer of blood between slide and cover accounts for some failures, as I have found in personal experi- ence. No doubt, in many cases in which we fail to find the organ- ism in supposed malaria a faulty diagnosis is the reason. Many of the cases in which latent malaria is supposed to have " come out" after a surgical operation are exploded by the negative examination for parasites and the positive indications of pus- pocketing which are afforded by a marked leucocytosis (never present in simple malaria), and the fact of insufficient wound drainage is often disclosed in this way. Whenever we see the leucocytes increased we begin to doubt the existence of an un- 372 SPECIAL PATHOLOGY OF THE BLOOD. complicated malaria ; if, furthermore, we see no signs of any pal- lor of the corpuscles we doubt the presence of malaria still more, as there is no more rapid deglobularizer than the malarial or- ganism. How long after a chill the organisms may still be found in the peripheral blood is difficult to decide, but certainly they can be found any time within twenty-four hours after the last chill, unless quinine has been given, and sometimes even if it has been given. OTHER CHANGES IN THE BLOOD. Red Corpuscles. The following is from Thayer's remarkable monograph : " A reduction in red corpuscles follows each paroxysm ; these reductions are more marked after the early paroxysms than after those occurring later. When a certain degree of anaemia has been reached the losses per paroxysm are much less. When the number of corpuscles is reduced to 2,000,000 or 1,000,000 there is little tendency toward a further fall ; sometimes there may be slight rises in the curve between the paroxysms ; often, however, the number of corpuscles remains stationary for weeks. " In pernicious cases the number of corpuscles may fall be- tween paroxj^sms." Kelsch has seen the count decrease to as small a number as 500,000 per cubic millimetre. The diminu- tion is greater the longer the disease lasts and the more in tense its manifestations. During the paroxysms, particularly the earlier ones, the red cells tend to increase in number. In tertian and quartan fevers there is a rapid and almost complete restitution of the corpuscles during the afebrile period. In sestivo-autumnal fevers the number of red cells bears a direct relation to the number of organisms. Crescentic bodies seem to have no influence on the number of red cells. When after a paroxysm the number of corpuscles has been greatly diminished the succeeding paroxysm may be followed by a slight reduction only or even by an increase. Bignami and Dionisi distinguish three types of post-malarial anaemia : BLOOD PARASITES AND INTESTINAL PARASITES. 373 1. Ordinary secondary anaemia, but with leucopenia instead of leucocy tosis ; such cases usually recover. 2. Anaemia practically identical with pernicious anaemia, megaloblasts being present, and ending fatally. 3. Anaemias which are progressive, because the bone marrow cannot compensate for the losses of corpuscles. The rapidity of the diminution in red cells may be very great. Kelsch's count of 500,000 cells per cubic millimetre, mentioned above, was after thirty days' illness. Grawitz has seen a loss of 4,000,000 cells in six days. Qualitative changes are those of severe secondary anaemia, deformities in size and shape, normoblasts, occasional megalo- blasts in the worst cases, motility in the " pale, ghostly" cells. Haemoglobin. The loss of haemoglobin bears usually a direct relation to the number of parasites in the blood. As a rule, the corpuscles and haemoglobin are diminished proportionally (color index =1) but sometimes the haemoglobin is reduced dispro- portionately. In convalescence the restitution of haemoglobin is often in- complete; persons living in malarial districts have often a slightly smaller percentage of haemoglobin than those living elsewhere. The rapid diminution in haemoglobin is a valuable point in differential diagnosis between malaria and typhoid or pneu- monia. White Cells. The number of leucocytes is usually subnor- mal, but show a slight increase at the beginning of the par- oxysm. Following this increase there is a rapid decrease con- tinuing throughout the paroxysm. . The small number of leucocytes is to be seen at the end of the paroxysm when the temperature is subnormal. From this time it shows a gradual increase until the beginning of the next attack (Billings) . In a general way the white cells follow the same course as do the red. The differential count shows a lymphocytosis whenever the white cells are subnormal, the larger forms of lymphocytes being especially numerous, while the polymorphonuclear cells and eosinophiles are scanty. In four cases of post-makrial anaemia Billings found quite marked leucocy tosis. 374 SPECIAL PATHOLOGY OF THE BLOOD. The occurrence of pigmented leucocytes has already been mentioned. Grawitz and others have noticed an increase of eosinophiles in post-malarial anaemia. I have frequently found small per- centages of myelocytes, three per cent being the highest in my experience. MALARIAL H^EMOGLOBIN^EMIA. During the paroxysms of this form of the disease, the num- ber of the red cells is much diminished, rouleaux not formed, marked poikilocytosis with nucleated forms. The leucocytes are increased. The regeneration is very swift, twenty-four to forty -eight hours being usually sufficient to re-establish normal conditions. FILARIA SANGUINIS HOMINIS. Although most commonly found in tropical countries, one species of this worm is not very uncommonly found in various parts of the United States. Any case of chylous urine or ele- phantiasis should lead us to make a careful examination of the blood for the filaria. There are at least four species of filaria, one of which is present in the blood chiefly at night, another chiefly during the daytime, and another continuously. Only the filaria nocturna has thus far been seen in America (Fig. 44). In examining for the filaria a slide of the fresh blood is pre- FIG. 44. The Filaria Sanguinis Hominis. The head, curled up, is seen at the right of the cut, the tail at the left. Instantaneous photomicrograph. Four hundred diameters magnification. pared in the usual way, but after 8:30 o'clock in the evening, 1 and examined at once. The embryo of this parasite (which is what 1 In persons who sleep in the daytime and work at night the habits of the filaria are said to become reversed, so that it appears in the peripheral circulation chiefly in the daytime, and is to be looked for then. FILARIA SANGUINIS HOMINIS. 375 we find in the human blood) is from one-ninetieth to one seven- tieth of an inch in length, i.e., about fifty times the diameter of a red cell, and about the width of a red corpuscle. Seen in the blood it retains its vitality and motile power for a considerable time, so that its motions may continue a week or more between slide and cover-glass. Cold has little effect upon it, even freez- ing temperature failing to do more than make the movements slower. A distinction can generally be made out between the embryo proper and its sheath (see Fig. 45). From this sheath the embryo escapes when in the blood of the mosquito, which insect FIG. 45. Tail of Filaria, showing prolongation of the sheath beyond the end of the embryo itself. Magnified 800 diameters. acts not infrequently as intermediary host and conveys the para- site indirectly from man to man through the medium of water. After sucking in the organism with the blood the mosquito lays its eggs and dies in some neighboring pond or stream whence the filaria again gains access to men. It is a long, slender, snake-like, gracefully shaped worm, and when alive its activity is so great that measurements and obser- vations of its structure cannot be made till it is paralyzed by approaching death (Fig. 46). 376 SPECIAL PATHOLOGY OP THE BLOOD. Posteriorly it tapers for one-fifth its length down to a very sharp point. The extreme end of the tail often looks as if ar- FIG. 46. The Movement of a Single Filaria during Four Successive Exposures of one-fifth of a second each, the entire series occupying less than five seconds. Magnified 800 diameters. ticulated, for it does not harmonize with the general curve of the body, but lies bent at an angle. Toward the head it tapers very slightly and when alive a " pouting" movement as if of breath- ing can be seen at its very extremity. About the middle of the body a granular aggregation can be made out along the central axis of the animal. Except for this granular portion the para- site is so translucent that it is not easy to make it out at first. The distinction of body and sheath mentioned above, appears as a "clear space" at each end of the body (vide Fig. 45). After the motions have ceased it becomes darker and traces of transverse striation may be seen (Fig. 47). It has no locomotive power and confines itself to wriggling in the same spot. Saussure 1 says he has watched them "fighting with each other for hours." The head of the filaria is said by some authorities to be sup- 1 Philadelphia Medical News, June 28th, 1890, where he reports twenty cases seen in Charleston, S. C. FILARIA SANGUINIS HOMINIS. 377 FIG. 47. Head of Filaria. Shows structure and beginning granular degeneration. Magni- fied 1,500 diameters. FIG. 48. Head of Filaria Magnified 1,500 Diameters. The blur in front of the head may be due to the motion of flagella. 378 SPECIAL PATHOLOGY OF THE BLOOD. plied with feelers or flagella, and Manson describes what he calls a "cephalic armature" or fang (Fig. 48). The same organism can sometimes be found in the chylous urine, but not every case of chyluria is due to the filaria san- guinis hominis. In a considerable proportion of cases no such organism is to be found. Henry (Med. News, May 2d, 1896) succeeded in staining the parasites intra vitam by giving the patient considerable doses of methylene blue internally for some weeks. Only a faint FIG. 49. Head of Filaria Overlapping a Red Corpuscle. The appearance might be mis- taken for the cephalic end of a sheath. bluish tinge was imparted, however, to the organism by this method. For finding the parasite it is best to use a low power, not an immersion lens, and the whole of several slides should be looked over. Specimens can be dried and preserved for staining provided we do not heat them over a lamp or pass them through a flame. Manson 1 stains with eosin and mounts in " glvcerin jellv" (Fig. 49). Several other species have been observed in England in negroes from the Congo River, but not hitherto in America. But as it frequently is to be found in persons who have no symptoms whatever, it may well be that some of these other species would be found here if one took the trouble to seek out natives of Southern China (one out of every ten of whom carries 1 The * Filaria Sanguinis Hominis," by Patrick Manson, M.D., Amoy, China, 1883. SPIROCH^TE OF RELAPSING FEVER. 379 about the filaria in his blood), or of Central Africa, or other tropical regions. SPIROCH^ETE OF RELAPSING FEVER. During the febrile paroxysms of relapsing fever, and for one or two days before them, Obermeyer and others have found FIG. 50. Spirochaetes of Relapsing Fever in Human Blood. FIG. 51. Spirochsetes of Relapsing Fever in Human Blood. 380 SPECIAL PATHOLOGY OF THE BLOOD. JJ^i: ipv^ FIG. 53. Spirochaetes of Relapsing Fever. SPIROCH^ETE OF RELAPSING FEVER. 381 \ constantly present in the peripheral circulation a parasite whose length averages about six times the diameter of a red corpuscle. Even under high-power lenses it is a mere thread in width, curled upon itself like a corkscrew and actively \ motile, so that in examining the blood with a low power we get " a peculiar impression of disturbance" among the red cells. The number of twists in this spiral-shaped organism varies a good deal, and one of its motions consists in contracting and extend- ing itself like a spiral spring. It can thus multiply its own length three or four times. It has also a delicate, wavy, but rapid motion along its long axis. The whole thread, or a part of it only, may have these motions. Further, the whole parasite has power of loco- motion apparently independent of the currents in the blood plasma of a slide and cover-glass specimen. Its locomotion is slow compared to the movements above described. Particularly in the blood post mor- tem they are apt to wind themselves into each other so as to seem much larger than they actually are, and sometimes a large "nest" of them may look like a leucocyte, except for the fine wavy threads which can be seen in motion at the periphery of the mass. The number present in the blood is very much smaller at the beginning of a paroxysm than after the second day. During the first few hours of a febrile period Mocyntkowsky could find only one spirochsete in ten or twenty microscopic fields, while later on he saw twenty or thirty of FIG. 54. Leucocytes Containing Spirochaetes. 382 SPECIAL PATHOLOGY OF THE BLOOD. them in a single field. There are usually more parasites with each successive paroxysm. Blood taken from different parts of the body often shows a great difference in the number of organisms to be found. The life history of a single parasite seems to be very short, but they multiply with the greatest rapidity. Albrecht has seen them so increase within six hours that whereas at first he saw only a few in the whole slide he later found many in each field. As the spirochsete dies, its movements get languid and finally it breaks up into small granular bits (spores?). Between paroxysms the spirochsetes are not found, but there are to be seen peculiar highly refractile globules compared by v. Jaksch to a diplococcus. The latter author believes that he has seen these develop into the spirochsete at the beginning of a paroxysm and hence believes them to be spores. This spirochaete is found in all cases of relapsing fever and in no other known disease, so that like the plasmodium malariae it is pathognomonic and of the highest importance. Ansemia and leucocytosis (during the paroxysm) are among the secondary results of the presence of this parasite in the blood. A certain resemblance has been noted between the spirochsete and a free flagellum broken off from a malarial parasite, but the clinical history and the presence or absence of other evidence of malaria in the blood would easily decide the question of diag- nosis. Technique of Examination. As in looking for the malarial organism it is best to examine the blood fresh between a slide and cover-glass (vide supra, page 7) and to use an oil immer- sion lens. In dried specimens the organism can be stained with fuchsin, but it is much more difficult to recognize than in the fresh blood. Phagocytosis (see Fig. 54) can easily be watched in the peripheral blood. DISTOMUM H^MATOBIUM. Bilharz found this parasite post mortem in the large internal veins (portal, splenic, mesenteric, etc. ), but as it has never been seen in the peripheral circulation its clinical importance is thus far nil. ANAEMIA DUE TO INTESTINAL PARASITES. 383 BACTERIA IN THE BLOOD. (a) Cover-Glass Specimens. Bacilli of anthrax, tuberculosis, glanders, grippe, typhoid fever, and tetanus have been demon- strated in the blood of human beings as well as have the pyo- genic streptococci and staphylococci, the diplococcus lanceolatus, the gonococcus, and the bacillus coli communis. Nevertheless it is exceedingly difficult and frequently impossible to find them, and no considerable practical use has as yet been made of the -cover-slip examination of blood for micro-organisms. Gunther's method is an excellent one. Cover-glass speci- mens of the blood are prepared as above described (page 43), and left a few seconds in five-per-cent acetic acid to render the red cells invisible ; the acetic acid is then shaken (not tvashed) off and the cover-glass held over the mouth of a bottle of strong ammonia water to neutralize the remaining acid. The covers are then stained with the Ehrlich-Weigert solution, 1 mounted in balsam, and examined with a one-twelfth immersion lens. (b) Cultures (see above, page 47). . ANAEMIA DUE TO INTESTINAL PARASITES. The bothriocephalus latus, ankylostoma duodenale, and a few other parasites are capable of producing by their presence in the intestine a very severe anaemia, which may be indistin- guishable from pernicious anaemia. As yet no such case has been reported in this country, but Askanazy 2 and Schaumann 3 have carefully studied the disease in Germany and found that the blood may correspond exactly with that of pernicious anae- mia, including the presence of high color index and of a major- ity of megaloblasts among the nucleated red cells present. Yet such cases may be rapidly and permanently cured by expelling the parasites from the intestine. No special description of the blood states need be given, as they present nothing which has 1 To 6 c.c. of distilled water add ten drops of aniline oil and filter. To the nitrate add a saturated alcoholic solution of gentian violet till slight (transient) turbidity appears. On the surface of this solution in a watch glass float the cover-glass face downward for twenty-four hours. 2 Vereins-Beilage der Deut. med. Woch., 1895, Bd. 148. 3 " Bothriocephalus-Anaemia, " Berlin, 1894 (Hirschwald). 384 SPECIAL PATHOLOGY OF THE BLOOD. not been already described under pernicious anaemia or severe symptomatic anaemia. ASCARIS LUMBRICOIDES. The reports of Jenner's Hospital at Berne for 1890 include a case in which the blood showed only 2,450,000 red cells before driving out the parasite with santonin, and 4,200,000 two weeks later. CHAPTEK XII. THE BLOOD IN INFANCY. I. All the signs by which sickness is shown in the blood of adults are exaggerated in children. Their blood is apparently more sensitive to the action of any morbid influence. Causes leading to but slight anaemia or leucocytosis in the adult, produce grave anaemia and very marked leucocytosis in children. Into the reasons for this I shall not attempt to enter. The increased toxicity of their serum compared to that of adults, and the rela- tively recent establishment of the functions for producing and destroying blood have been suggested as explanation. Comparatively slight hemorrhages, gastro-intestinal or re- spiratory disorders, which would not impoverish an adult's blood may produce considerable anaemia in a young child. II. All forms of anaemia in infancy are apt to be associated witli enlarged spleen. III. I have already alluded to the polycythaemia and leucocy- tosis of the new-born, and the gradual fading out of these rela- tive abnormalities as the child grows up. In judgments as to the presence or absence of leucocytosis in infancy, these physio- logical variations are too often lost sight of, especially as the proper leucocyte count for any given infant depends not simply on its age but on the backwardness or forwardness of its develop- ment. As with the fontanelles, the growth of the blood toward adult conditions may be retarded by congenital weakness (infan- tile atrophy, marasmus) or inherited disease (tuberculosis, syph- ilis) as well as by acquired sickness (rickets, cholera infantum). Under the influence of any of these drawbacks a sick child's blood may be no further developed at three years than that of a healthy child of eighteen months. IV. When we remember that in early infancy the leucocytes differ from those of adults not only in number but in that the lymphocytes are relatively more numerous (" lymphocy tosis of infancy"), we shall understand that any influence like rickets 25 386 SPECIAL PATHOLOGY OF THE BLOOD. or syphilis which retards development, will show lymphocytosis together with the increased leucocyte count. Qualitatively as well as quantitatively the blood reverts to a more infantile con- dition. V. This shows itself not merely in the leucocytes but in the red corpuscles. During the first days after birth the infant's blood shows greater variations in size and shape than that of adults, as if the type were not yet quite fixed. The majority of authors also find a few normoblasts in the first few days of life. These are not invariably present, doubtless because in some children the blood at the time of birth is more developed than in others. Under pathological conditions the red cells revert to this earlier type and deformed or nucleated corpuscles are plentiful. This is more marked than in anaemia of the same grade oc- curring in adults. An anaemia that shows but thirty nucleated erythrocytes per cubic millimetre in an adult might show ten times that number in a child. VI. As we said before, all blood changes are exaggerated in infancy. This includes such physiological changes as the diges- tion leucocytosis or that following cold bathing as well as patho- logical leucocytosis and anaemia, and changes in the degree of dilution or concentration of the blood seem to be similarly exag- gerated, as is seen, e.g., in the physiological variations in the specific gravity of the serum (Hock and Schlesinger 1 ) . VII. The haemoglobin, though relatively high at birth and for the first few weeks, is lower than that of adults during the rest of childhood. The high percentages of the earliest weeks are not due to a polycythaemia, but to a genuine in- crease of haemoglobin in the individual cells (Schiff 2 ), color indexes being often over 1. It is indispensable, therefore, that we should know the age and degree of development of a child before we can draw accurate inferences from its blood. In many of the cases reported in lit- erature we are unable to judge whether the blood condition is pathological or not, because the age of the child is not given. For example, v. Limbeck 3 quotes a case of acute gastritis re- 1 Hock and Schlesinger Centralb. f. klin. Med., 1891. 2 Schiff : Zeit. f. Heilk., vol. xi., 1890. 3 v. Limbeck: loc. cit., p. 373. THE ANAEMIAS OF INFANCY. 387 ported by Fischl ' as having an unusually high percentage of ij liipnoc.y tes (ijtfA percent). But this is physiological in the first days of life, and may have been so in this case, the age not being given. Observations of this sort should always represent a compari- son between the conditions present before and during the sick- ness in question. Bearing these general considerations in mind, we shall be better able to find our way among the complications and per- plexities of the blood conditions in infancy. THE ANJEMIAS OF INFANCY. As above mentioned, anaemic infants are apt to have enlarged spleens. This may be due either to the anaemia or to some disease accompanying or underlying the anaemia (e.g., rickets, syphilis). It seems more probable that the hypertrophy is not directly or exclusively dependent on the anaemia, inasmuch as similar blood changes are found without splenic enlargement. By far the greater number of reported cases of severe infantile anaemia are accompanied or caused by such diseases as rickets and hereditary syphilis, both of which may cause splenic hy- perplasia even when no anaemia is present. It seems probable that the anaemia and the enlargement of the spleen are alike symptomatic of an underlying disorder. 1. Some writers (e.g., Luzet 2 ) divide the anaemias of infancy into two classes : those with splenic enlargement and those with- out it. Luzet considers that the former class is severer than the latter and more apt to show large numbers of nucleated red corpuscles than those with normal-sized spleens. This classi- fication, however, does not always hold. We may have very severe anaemia without splenic enlargement and splenic en- largement with slight anaemia, and the presence or absence of numerous nucleated red corpuscles is governed by conditions other than the size of the spleen. 2. Another classification of children's anaemias was proposed in 1892 by Monti and Berggriin ("Die chronische Anamie im Kindesalter," Leipzig, 1892). They divided the cases into 1 Fischl : Zeit. f. Heilk., 1892. 2 Luzet : Diss., Paris, 1891. 388 SPECIAL PATHOLOGY OF THE BLOOD. the mild and the grave, each group being subdivided into those with leucocytosis and those without it. C ivriri _ j With leucocytosis. { Without leucocytosis. Secondary anaemia of infancy = < , _. , , n _ j With leucocytosis. ~ ( Without leucocytosis. They rightly discard the term " splenic anaemia," correspond- ing as it does to no single set of blood changes. The above classification puts pernicious anaemia, leukaemia, and anaemia infantum pseudoleukaemica (v. Jaksch) in a different category. (a) Mild cases of secondary anaemia show no deformities in the shape or size of the red cells. The color index may or may not be low. The cases with leucocytosis are much more numer- ous than those without it and more apt to have a low color index ; in other words, the loss of corpuscle substance is greater and the cases are approaching the imaginary boundary between "mild" and "grave." (b) The grave cases have poikilocytosis, and of course a greater reduction of corpuscle substance. " Chlorotic" conditions, and most but not all those with en- larged spleen, come under this heading ; also most of those due to hereditary syphilis, prolonged diarrhoea, and rickets. In 1894 Monti ! gave the following classified lists of the com- monest antecedents of secondary anaemia in infancy : In the mother during pregnancy. due to.... | Malada etc 2. Acquired f From navel. ri TT- J After circumcision. [emorrhage. j Scurvy p ur pura, haemophilia, Werl- [ hnf 's disease, meleena. f Inanition. Bad hygiene (lack of light, air, etc.). Post-febrile. Nephritis, diarrhoea, serous effusions. ^ 2. Other causes. { Syphilis. Rickets. Suppuration. Diseases of liver, spleen, bone, or ^ lymph glands. He points out that cases with leucocytosis are usually graver than those without it and may develop into pernicious anaemia ; 1 Wiener med. Woch., 1894. THE ANAEMIAS OF INFANCY. 389 also that the presence of leucocytosis does not point to malig- nant disease, suppuration, or any of the causes which usually account for it in adults. Grave cases with leucocytosis in infants under twelve months are apt to develop into the anaemia infantum pseudoleukaemica, or into true leukaemia or pernicious anaemia. On the whole, the division of Monti and Berggriin seems much better than that according to the particular causes, e.g., "rachitic anaemia," "syphilitic anaemia," etc., for there is no particular set of blood changes that follows rickets, syphilis, or any other disease. In connection with various diseases of in- fancy, and particularly with those last named, we may have anaemia of any grade of severity from that reducing the red cells to 4,000,000 down to cases with only 500,000 red cells per cubic millimetre or even less. The worse the case is the more likely is it to be accompanied by leucocytosis and the more nu- merous will be the nucleated red corpuscles, always more numer- ous here than in anaemia of adults. In syphilis, hereditary or acquired, the red cells may fall be- low 1,000,000 and the leucocytes may rise as high as 58,000 (Loos). The haemoglobin may be proportionally diminished, or may be even lower than the percentage of red cells, so that a " chlorotic" condition obtains. Such cases have been called chlorosis, but it seems better to confine this term to anaemia of unknown origin and favorable course occurring in women soon after puberty, since obviously secondary cases may have similar blood. Rickets in a case observed by v. Jaksch caused a fall of the red cells to 750,000 and Luzet counted 1,590,000 in a similar case. The haemoglobin is usually low, but Hock and Schles- inger found 60 per cent with 2,300,000 red cell, a color index of 1.2 +. Leucocytosis may occur even when no anaemia is present. Hock and Schlesinger found 45,000 leucocytes in a rachitic child of sixteen months, sound in other respects and not anaemic. . Acute gastritis causes at first only leucocytosis (with increased percentage of lymphocytes). If it becomes chronic the reduction of red cells is severe. Hay em found only 685,000 red cells per cubic millimetre in an infant of two months, though recovery eventually took place. 390 SPECIAL PATHOLOGY OF THE BLOOD. In tuberculosis of lungs and peritoneum in a child of seven, Monti and Berggriin counted 3,230,000 red and 17,200 white cells with 52 per cent of haemoglobin. Qualitative Changes. The exaggeration characteristic of all blood changes in in- fancy extends to the presence of nucleated red corpuscles, which in all forms of severe anaemia are very numerous. What has been described above (page 91) as the typical megaloblast, a large pale-stained nucleus in a very large cell (see Plate IV.), is relatively rare in infancy. The nuclei are almost always deeply stained whatever their size, and apt to be small. Dividing nuclei are very common, both by karyolysis and karyokinesis. These changes are most often found in the anaemias of the severest type and those which resemble leukaemia (see below, page 397), but may occur in any marked secondary anaemia. Polychromatophilic and " degenerative" changes are very com- mon in severe cases. The increased leucocyte count, so frequently fouud, is often made up of a majority of lymphocytes. This change, as above said, is not characteristic of rickets, syphilis, or any other cause of anaemia, but it is to be regarded as a mark of the arrest of development or reversion to an earlier type of tissues brought about by various diseases in early infancy. Sometimes the large lymphocytes and sometimes the small are in excess. A further qualitative change already alluded to (see above, page 121) is the occurrence of myelocytes. We have seen that small percentages of these cells are not uncommonly seen in the anaemias of adults. Now this, like all other blood changes, is exaggerated in infancy. Myelocytes are more apt to appear and in greater numbers. Their presence is not characteristic of any one disease, but they are commonest in the severer types of secondary anaemia, such as those following syphilis and rickets. Their significance is about the same as that of nor- moblasts. At times, however, they are so numerous as to make us hesitate somewhat before we exclude splenic-myelo- genous leukaemia. This brings us naturally to the discussion of the difficulty of distinguishing the different blood diseases in infancy, which natur- "ANEMIA INFANTUM PSEUDOLEUK^EMICA." 391 ally centres in the question of the existence and nature of the so called "ANEMIA INFANTUM PSEUDOLEUK^EMICA." Von Jaksch's 1 decription of this disease (which he was the first to recognize) includes the following elements : 1. Grave anaemia e.g., 820,000 red cells per cubic millimetre in one case. 2. Extensive leucocytosis e.g., 54,660 white cells per cubic millmetre, in the same case. 3. Great variations in the form, size, and staining of the white cells. 4. Deformed, degenerated, and nucleated red cells. Von Jaksch admits that none of these blood changes are characteristic of the disease, but thinks that its title to the posi- tion of a distinct and separate disease rests upon clinical data, the more important of which are : (1) A great enlargement of the spleen without any such accompanying enlargement of the liver as is usually found in leukaemia (the lymph glands are sometimes enlarged). (2) A relatively good prognosis. (3) Post mortem we find no positive evidence of leukaemia. This description was given by v. Jaskch 1 in 1889. He stated the relation of white to red corpuscles as 1 : 12, 1 : 17, and 1 : 20 in the cases seen by him. Later he reported three cases in one of which the white cells numbered 114,150, and the red 1,380,- 000. The differential counts are not carefully given. Almost at the same time Hayem* reported a similar case, and noted the abundance of nucleated red corpuscles many of which were undergoing mitosis. This was verified by Luzet 3 in May, 1891 (Arch. gen. de Med.), who reported two cases. His description of the disease differs considerably from that of v. Jaksch. He finds no greater difference between liver and spleen than often exists in true leukaemia. The course of the disease, though sometimes chronic, usually ends in death. The leucocytosis in Luzet' s cases was less marked than in those of v. Jaksch and not greater than that occurring in many anaemias of children. He dwells particularly on the large number of nu- ' Von Jaksch .- Wien. klin. Wocli., 1889, Nos. 22, 23. 2 Hayem Gaz. des Hopitaux, 1889, No. 30. 3 Luzet : Diss. , Paris, 1891. 092 SPECIAL PATHOLOGY OF THE BLOOD. cleated red cells, and the frequency of mitosis and considers this the most important diagnostic point. Although Luzet's continues to use the name suggested by v. Jaksch, he describes the disease so differently that it is difficult to see why the same title should be given to it. He agrees with v. Jaksch in thinking that it is not simply a severe secondary anae- mia due to syphilis, rickets, tuberculosis, or infectious disease. Somewhat similar cases had already been described by vari- ous Italian writers (e.g., Fede) under the title of "Infective Splenic Anaemia of Infants." Among others who have written on the subject are Baginsky, 1 Senator, 2 Fischl, 3 Andeoud, 4 Monti and Berggriin, 5 Felsenthal, 6 Baudnitz, 7 Epstein, 8 Alt and Weiss, 9 Hock and Schlesinger, 10 Crocq, 11 and Botch. 12 The majority of these writers report very little as to the dif- ferential counts of white corpuscles. An increased percentage of the polymorphonuclear forms is mentioned by many, but Botch in a case with 1,311,250 red cells and 116,500 white cells found only 16 .per cent of the polymorphonuclear variety with 46 per cent of small lymphocytes, 34 per cent of large lympho- cytes, and 4 per cent eosinophiles. A second case had only 14 per cent of polymorphonuclear cells and 84 per cent of lympho- cytes (large and small). Von Jaksch noted the lack of any relative increase of eosino- philes, supposing this to be a means of distinguishing his cases from true leukaemia. Luzet, on the other hand, found eosino- philes numerous. (This of course has no weight for or against leukaemia. ) Klein (loc. cit.) noted the occurrence of myelocytes in small number. I Baginsky: Arch. f. Kinderheilk., 1892, vol. 13. * Senator: Berlin, klin. Woch., 1892. 3 Fischl : loc. cit. 4 Andeoud : Rev. de med. de la Suisse rom., 1894, p. 507. 6 Monti and Berggriin : loc. cit. 6 Felsenthal : loc. cit. 7 Raudnitz: Prag. med. Woch., 1894, p. 6. 8 Epstein: Prag. raed. Woch., 1894, p. 6. 9 Alt and Weiss : Centralb. f . med. Wissenschaft, 1892. 10 Hock and Schlesinger : loc. cit. II Crocq: "Etude sur 1'Adenie," etc., Brussels, 1891 (Lamartin). 12 Rotch : Psediatrics, 1895, p. 361. "ANEMIA INFANTUM PSEUDOLEUKAEMICA." 393 The discrepancy of these different reports is suggestive. The chief importance of the heterogeneous group of cases which have received the name of anaemia infantum pseudoleu- kcemica seems to me to be as a proof of the difficulty of distin- guishing the various blood diseases in infancy. Among the cases reported under this name are some which might be any one of the following list: Pernicious anaemia, secondary anaemia with leucocytosis, Hodgkin's disease, lym- phatic leukaemia, and probably splenic-myelogenous leukaemia. (a) Most of the few reported cases of pernicious anaemia in infancy have shown moderate leucocytosis (as compared with adult blood), a fact which deprives us of one of the means of dis- tinguishing the disease from secondary anaemia. The reports as to nucleated corpuscles very rarely separate normoblasts from megaloblasts, and we have no way, therefore, of being sure on this important point. The high color index and large diameter of the red cells are occasionally seen in other anaemias of infancy and are not always present in pernicious cases. The great fatality of all kinds of anaemia in infancy prevents our calling a case pernicious because of a fatal termination. Enlargements of liver and spleen occur in many cases of each type of infantile anaemia, and occasionally in pernicious anaemia of adults. They do not, therefore, exclude pernicious angcmia in infancy. Bearing these facts in mind, it is evident that some of Luzet's cases of "anaemia infantum pseudoleukaemica" may have been pernicious anaemia. Von Jaksch's own cases may have been either (a) Hodgkin's disease with leucocytosis, (I) grave secon- dary anaemia with leucocytosis (Monti and Berggriin), or (c) leukaemia. (a) Hodgkin's disease, which v. Limbeck finds to be very common in infancy, may affect the liver and spleen and not the external lymph glands, and may be accompanied by anaemia and leucocytosis such as v. Jaksch describes. Epstein con- siders that this is the case, and denies the existence of any such disease as the anaemia infantum pseudoleukaemica. (&) As any anaemia secondary to rickets or syphilis may have enlarged spleen and liver and marked leucocytosis, we cannot tell from v. Jaksch's description that we are not dealing in his cases with secondary anaemia. (c) Since v. Jaksch does not give any accurate differential 394 SPECIAL PATHOLOGY OF THE BLOOD. count of the leucocytes, there may have been large numbers oi myelocytes in his cases for all we know, or an overwhelming percentage of lymphocytes, i.e., either type of leukaemia. One of the cases reported by Rotch as " anaemia infantum pseudoleukaemica" had 80 per cent of lymphocytes in a leucocyte count of 116,500, the ratio of white to red cells being 1:11, and the nucleated corpuscles abundant. The external lymph glands as well as the liver and spleen were enlarged. How such a case is to be distinguished from lymphatic leukaemia without autopsy I cannot see. Large numbers of nucleated corpuscles with mito- ses (present in this case) are to be found in any anaemia of in- fancy where the red cells, as in this case, have sunk as low as 1,311,500, and therefore do not exclude leukaemia. Von Jaksch protests that his cases are not secondary to rickets or any other disease, but Fischl 1 in a careful study of all the published cases finds that out of a total of eighteen cases, sixteen had severe rickets and two hereditary syphilis. The writings of Eaudnitz, Ebstein, Felsenthal, Fischl, and v. Limbeck, which deny the separate existence of the anaemia infantum pseudoleukaemica, are convincing to me, and are rein- forced by the few cases of bad anaemia in children which I have seen. We must distribute the cases of anaemia with leucocy- tosis and large spleen under pernicious anaemia, secondary an- aemia, and leukaemia. But our problem is not yet nearly solved. All we have gained is the belief that v. Jaksch's new disease does not help us to classify these doubtful cases. The difficulty is still very great. The following cases reported by Dr. Vickery in the Medical News for December, 1897, illustrate this : CASE I. A male child of sixteen months with symptoms of grave anaemia, greatly enlarged spleen and slightly enlarged liver, showed the following figures: Eed cells, 2,500,000; white cells, 22,000. Differential count of 500 cells showed: Lym- phocytes, 53.8 per cent (46.2 of the smaller type) ; polymor- phonuclear cells, 29.4 per cent; eosinophiles, 6.2 per cent; myelocytes, 10 per cent. While counting these, 147 nucleated red corpuscles were seen, of which 21 were normoblasts, 50 megaloblasts, and 47 microblasts ; 6 showed mitosis in their nuclei. The child died shortly after without any complication or 1 Fischl Zeit. f. Heilkunde, 1892. 395 intercurreut disease. No autopsy. No evidence of rickets or syphilis or other previous disease. CASE II. Young infant with enlarged external lymph glands and very large spleen. July 14th, 1897 Eed cells, 4,300,000; white cells, 31,000; haemoglobin, 60 per cent; polymorpho- nuclear neutrophiles, 57.5 per cent; small lymphocytes, 26 per cent; large lymphocytes, 15 per cent; eosinophiles, 0.5 per cent; myelocytes, 1 per cent. One or two nucleated red corpuscles in every field. Out of 100 of them 89 were large and 11 small. Many showed mitosis. Polychromatophilic forms numerous. July 19th Seventeen megaloblasts seen while counting 1,000 white cells. Blood is otherwise about the same. The case was lost sight of and not traced. Now I see no reason for supposing these cases to represent a new type of disease, and yet I cannot feel perfectly safe in classifying them as primary anaemia, secondary anaemia, or leu- kaemia. (a) Primary or pernicious anaemia should have a lower count of red cells. The percentage of myelocytes in the first case (ten per cent) is higher than in any other case of pernicious anaemia on record, though in one adult case with autopsy I found 9.2 per cent with a leucocytosis of 12,500, or 1,150 mye- locytes per cubic millimetre, against 2,200 per cubic millimetre in this case. (b) It is hard to call an anaemia secondary which kills with no complications and when there is no evidence of any disease to which it can be secondary. (c) For splenic-myelogenous leukaemia the total leucocyte count and the percentage of myelocytes are very small in either case. Still the leucocyte count may drop very low in leukaemia even without any inflammatory complication. Such a case is reported by Osier, in which the leucocytes fell to 7,500, of which only 300, or four per cent, were myelocytes. Hay em (loc. cit., page 864) in a ten months' child counted 2,712,500 red and 33,000 white cells, almost the same figures as in the case just quoted. [Hay em unfortunately gives no differential count, but apparently considers the case leukaemia because of the enormous number of nucleated red cells, many with mitoses.] Morse's case of leukaemia in infancy had 2,900,000 red and 48.000 white cells. Twenty -one and four-tenths per cent of the 396 SPECIAL PATHOLOGY OF THE BLOOD. leucocytes, or about 10,000, were myelocytes. The same abun- dance of nucleated red cells (some with mitoses) were here pres- ent as in Hay em's case, so that there is evidently nothing pecu- liar in their presence in the disease described by v. Jaksch, as Luzet supposed. These cases show that leukaemia may at certain periods pre- sent just such a blood picture as was present in the above-quoted case and that the number of leucocytes in the leukaemia of in- fants may be no greater than that in any anaemia with the leu- cocytosis so common in children. It seems to me the most natural conclusion to be deduced from these facts is that we meet with cases in infancy which are apparently intermediate betiveen leukaemia and pernicious ancemia. I have pointed out elsewhere that there are many points of re- semblance between the two diseases. The case of leukaemia re- ported by Osier showed at one period the period of remis- sion a fall in the number of leucocytes and in the percentage of myelocytes till the blood was practically that of pernicious anaemia. Dr. Kotch's case (above quoted) is another in which the diag- nosis seems to lie somewhere intermediate between the two diseases, anaemia and leukaemia. The case which I have quoted above seems to me on the whole nearer to the type of pernicious anaemia than of leukaemia, and Dr. Botch's nearer to the latter than to the former; but each is really intermediate, so far as the blood goes, between the two diseases. I have no intention of suggesting that the organic lesions in these cases are intermediate between leukaemia and pernicious anaemia. It is simply the blood that is so. Engel's case, reported in Yirchow's Archiv, Vol. 135, sug- gests the same thing. He calls the case one of "pseudo-perni- cious anwmia." Meylocytes were abundant. PolymorpJious Condition. This illustrates that " polymorphous" condition of the blood which v. Jaksch supposed to be characteristic of the anaemia in- fantum pseudoleukaemica. The same thing was very marked in all the bad cases of anaemia which I have seen, including the case above mentioned, and a case of true leukaemia in a girl of eight. The impression one gets from the field of a LEUKAEMIA. 397 stained specimen is that no two white corpuscles are alike. Every species is subdivided into several sub-varieties and all stages of degeneration are to be seen in each variety. But this is char- acteristic of any very severe infantile anaemia and not of any single type. LEUKEMIA. In Morse's careful article of August, 1894 (Boston Med. and Surg. Journal} , twenty cases of leukaemia in infancy are collected. As he rightly says, probably most of these cases were not gen- uine. Only one of them includes a differential count, and this is in a lymphatic case. Morse's is the only one of the splenic- myelogenous type on record in which the diagnosis is made reasonably certain by a color analysis. Fischl in 1892 said that there was no case on record with a differential count. A case was seen in 1890 by Dr. F. C. Shattuck, which was apparently acute, the symptoms appearing only six weeks be- fore death. Cover-glass preparations examined by W. S. Thayer showed a ratio of about 1 white to 20 red cells. The differ- ential count ' showed: Small lymphocytes, 97.9 per cent; large lymphocytes, .7 percent; poly nuclear cells, 1.4 per cent; eosino- philes, .08 per cent. The other case reported by Morse has been mentioned above. Charon and Giratea " have recently reported a case in a child of eight with 880,000 red cells, 305,000 white cells, and 39 per cent of haemoglobin. It was apparently of the myelocyte type. E. Miiller thinks that there are about five other (German) cases on record, all of acute leukaemia and all with a similar blood count, though in some the large lymphocytes (without neutro- philic granules) have been described as "myelocytes." Miiller 3 has lately reported with great care three cases of leukaemia, all of them in boys four years old all apparently acute, all of the gastro-intestinal type i.e., the glands and fol- licles throughout the whole length of the alimentary tract being 1 Reported by Thayer in the Boston Medical and Surgical Journal, 189P, vol. 128, p. 183. 2 Bull. d. Soc. Roy. d. Sciences Med. , etc., Bruxelles, 1897, No. 7. 3 Jahrbuch fiir Kinderheilk., 1896, vol. 43. 398 SPECIAL PATHOLOGY OF THE BLOOD. the chief seats of infiltration, though the liver and spleen were also enlarged. The counts were as follows : CASE I. CASE II. CASE III. April 30th. *!* May 2d. May 3d. Red cells 1,508,000 109,500 1,684,000 93,800 1,362,000 46,000 1,232,000 6,800 Death. 2,290,000 206,000 1,308,000 420,000 White cells Haemoglobin Polymorphonuc 1 e a r neutrophiles . 85 (8-10 M diameter). 1* Many. Few. Few. Many. 1& .7 2. 97.3 .01 Small lymphocytes 1 . . Large lymphocytes . . Eosinophiles Few. Few. Two nor mo - blasts seen in counting 1,135 1 e u - cocytes. Seven seen in counting 1,118 leucocytes. Megaloblasts All with large pale nuclei. APPENDIX. NEUSSER'S PERINUCLEAR BASOPHILIC GRANULES. Using the following modificatiou of Ehrlich's tricolor mix- ture, Neusser 1 believes that he can bring out certain character- istics in the leucocytes of value in diagnosis and prognosis. i Acid f uchsin 50 c. c. Saturated aqueous solution of ! Orange G 70 " ( Methyl green 80 M Distilled water 150 " Absolute alcohol 80 " Glycerin , 20 u Cover slips stained with this mixture show in certain dis- eases (e.g., gout, leukaemia) a grouping of dark blue-stained granules around the nuclei of the mononuclear leucocytes and over and around the nuclei of polymorphonuclear leucocytes. These granules appear to take up only the basic part of the tri-color mixture. For Neusser's conclusions regarding the meaning of these granules, the reader is referred to pages 258 and 318. The researches of Futcher have in my opinion utterly disproved Neusser's claims. The granules are of no known clinical sig- nificance and certainly have no direct relation to gout or any other alloxuric diathesis. I have a triple stain (not made up for the purpose) which brings out Neusser's granules in every blood, normal or abnormal. 1 Wien. klin. Woch., 1894, No. 39. PART VII. EXAMINATION OF THE SERUM. CHAPTER XIII. THE CLUMP REACTION. GENERAL DESCRIPTION. ALTHOUGH this phenomenon is to be obtained in various in- fections, natural as well as experimental, and with various body fluids, I shall describe as a typical case of it the reaction which takes place when the blood serum of a patient ill with typhoid fever is added in certain proportions (vide infra) to a young bullion culture of well-certified and virulent typhoid bacilli. In a drop of such a mixture, examined between slide and cover- glass 1 with a magnification of 300 diameters or more (an immer- sion lens is not necessary) , we notice, as soon as the serum and culture are mixed, either a marked slowing of the progressive movements of the bacilli or an unequal distribution of them in the different parts of the preparation, some parts showing the bacilli closely crowded, while in others they are more scat- tered. Whichever of these changes occurs first, the slowing of locomotion or the tendency to grouping, the other soon follows, and then both processes go on together, as admirably described by Biggs and Park : 2 " Some of the bacilli soon cease all progressive movement, and it will be seen that they are gathering together in small groups of two or more, the individual bacilli being still some- what separated from each other. Gradually they close up the spaces between them, and clumps are formed. According to the 1 Hanging-drop preparations are often recommended, but a simple slide and cover-glass are as good for the purposes of this reaction. 2 American Journal of the Medical Sciences, March, 1897. THE CLUMP REACTION. 401 completeness of the reaction, either all the bacilli may finally become clumped and immobilized or only a small portion of them, the rest remaining freely motile, and even those clumped may appear to be struggling for freedom. With blood contain- ing a large amount of the agglutinating substances all gradations FIG. 55. Pure Culture. FIG. 56. Partial Reaction. FIG. 57. Typical Clumping. in the intensity of the reaction may be observed, from those shown in a marked and immediate reaction to those appearing in a late and indefinite one, by simply varying the proportion of blood added to the culture fluid" (see Figs. 55, 56, and 57). The process may go on gradually and be much more distinct at the end of half an hour. The groups or clumps above described constitute the impor- tant part of the reaction for diagnostic purposes. Of the loss of motility more will be said later. 26 402 SPECIAL PATHOLOGY OF THE BLOOD. The clumps may hang together for a long time. They have been observed unchanged for one hundred and forty -four hours. On the other hand, they may be dissolved in a few hours and the bacilli regain their motility. In watching the formation of the clumps it is easy to see that the bacilli are positively attracted to each other and do not drift passively into a heap. The loss of motility is not the cause of the clumping, as they often begin to approach each other while in vigorous motion. The power of locomotion is lost much sooner than are the squirming and spinning motions, which often persist among the bacilli in the peripheral parts of the clumps as well as outside them. The clumps tend to adhere to the under side of the cover- glass. Specimens can be fixed and stained with the bacilli in clumps contrasting strongly with the even distribution of the bacilli in ordinary stained preparations. TECHNIQUE OP THE CLUMP EEACTION IN TYPHOID FEVER. Our account of the methods of obtaining the clump reaction may be divided into the following parts : 1. The body fluids to be used and the methods of obtaining them. 2. The cultures. 3. Dilution and the time limit. 1. THE BODY FLUIDS TO BE USED. Experiments have proved that the reaction can be obtained with the following fluids : (a) The whole blood, fluid or dried. (6) The plasma and serum, fluid or dried. (c) The fluid obtained by blistering. (d) The fluid normally present in the pericardium, pleura, peritoneum, and joints; not in rapidly accumulated effusions. (e) The milk and colostrum of women suffering from typhoid during lactation. (/) Pus from persons suffering with typhoid whether the bacilli of Eberth are present in the pus or not. THE CLUMP REACTION. 403 (g) Tears naturally (i.e., gradually) secreted. If secreted in response to the irritation of ammonia fumes, the tears do not produce clumping. (h) Some observers also find it in the fluid of oedema and in the bile. Others do not. (i) The clumping persists in the above-named fluids after death and even in putrefaction. The "juice" of the spleen, kidneys, and rarely of the liver, will give the reaction feebly. (/) Though present in the placental blood of pregnant typhoid patients, it does not usually exist in the foetus. The saliva, gastric juice, and sweat do not produce the reac- tion, so far as known. The aqueous humor sometimes does. The urine and faeces sometimes do and sometimes do not give it, but these excretions in normal persons may also produce the reaction, so that they cannot be made clinically available. Of all these fluids, the blood, the serum, and the fluid of blisters are the only ones used in clinical work, both because of their greater convenience, and because the clumping power is much more marked in the blood and blister fluid than in any of the others. 1. Use of the Wliole Blood Fluid. TJie advantages of this method are (a) its quickness, and (6) the small amount of blood (one drop) sufficient for the test. Its disadvantages are (a) that the corpuscles interfere slightly with the fields in which the reaction is to be watched, and (b) that they sometimes lead to the formatkm of false clumps ("pseudo-amas"), which simulate those present in the real clump reaction, and lead to false inferences. Both these objec- tions are trifling, however, as the corpuscles can be excluded by waiting a minute or two until they settle, leaving a clear liquid above in which the reaction can be observed. The false clumps are rarely seen, and can be differentiated from the true by care- ful technique (see below). I have used this method in many cases and always found it satisfactory and convenient. Widal, McWeeney, Delepine, Coleman, and others have employed it with success. It is most suitable for the " quick method" (see page 406), and is chiefly employed in this way. Procedure. Suck up some water with a medicine-dropper and 404 SPECIAL PATHOLOGY OF THE BLOOD. expel ten drops of it into a watch-glass. Then empty and dry the dropper, draw up from the watch-glass the ten drops just expelled, and mark with a file on the side of the dropper the point up to which the ten-drop column extends. Mark also the point to which one drop (expelled and then sucked up again as before) will rise. Ten drops of the bouillon culture of the bacilli to be used are then expelled into each of several small test-tubes, and one of these tubes is carried to the bedside. After pricking the ear as if for blood examination 1 (see page 6), put the end of the medicine-dropper into the blood drop, and carefully draw back the rubber bulb (which has been previously pushed down over the glass part of the dropper) until the blood rises to the mark for one drop. Wipe from the outside of the dropper any blood that may adhere there and then expel the drop into one of the little test-tubes containing the ten drops of bouillon culture. In this way blood can be taken for examination from a dozen patients in as many minutes. 2. Whole Blood Dried. This method, though previously described and tested by Widal, was first put into effect in large numbers of cases by Wyatt Johnson, of Montreal, for the use of the Board of Health of Quebec, by whom specimens of dried blood sent by mail were examined and diagnoses returned as with diphtheria cultures. It was subsequently employed on a large scale by the Boards of Health of New York and Chicago. The advantages of the method are (a) the ease and quickness with which the blood can be obtained, (b) the convenience for transportation by mail, and (c) that it does not deteriorate or become contaminated by bacterial growth, as specimens of fluid blood or serum are so apt to do. Its clumping power is fully equal to that of the serum in most cases* These advantages are very great and would surely lead to the 1 Squeezing and milking the ear are of no luirm in this procedure and enable us to get on with a trifling and painless puncture. ' 2 Widal and Delepine think the fluid serum is slightly more powerful than the dried blood. Johnson admits that in one-tenth of the cases the serum is the more powerful. I have obtained reactions with the dried blood in only seven-eighths of the cases in which I got them with the fluid serum. THE CLUMP REACTION. 405 immediate and universal adoption of this method were it not for iiie following serious drawbacks : (a) It is difficult to measure the amount of blood to be used in the test. This is important, because, as we shall see later, a positive reaction means not simply a clumping, but a clamping in a 1 : 10 dilution of the blood, to get which we need to know just how much blood we are dealing with. When we take the blood from a patient ourselves we can use the marked medicine drop- per, as above described, but when blood is received through the mails for examination or taken by any one who does not measure it in some way, we cannot accurately gauge the dilu- tion. (b) It is agreed by all who have used the method extensively that the clumping may occur with the blood of healthy people and hence confuse our inferences. Whether these " false clumps" are due, as Widal supposes, to masses of fibrin and debris in which the bacilli become entangled, or whether they are formed in the ordinary way, there can be no doubt that they occur occa- sionally when dried blood is used. Johnson has succeeded in avoiding such errors in his own work by the use of attenuated culture (vide infra). These objections have led most observers to prefer the fluid serum, but when we have not the apparatus necessary for col- lecting and preserving fluid serum, or when such apparatus could not be transported, the method is of great value. Procedure. The blood should be dried either upon a glass slide or on a piece of glazed paper or card. Any absorbent substance is less available. Glass is easier to sterilize than paper. Several large drops should be placed in different parts of the glass or paper and thoroughly dried. If paper has been used, we cut out the dried blood drop with a pair of scissors, keeping close to the blood all round, and drop it into a test-tube containing one or two drops of water, in which with some sharp-pointed instrument we mix the dried blood, freeing it as well as possible from the paper. To the liquid so obtained add eight or nine drops of the bouillon culture of bacilli and proceed in the ordinary way. Or we may drop the fragment of paper holding the blood directly into ten drops of bouillon culture using the bouillon itself to soak off the blood from the paper. 406 SPECIAL PATHOLOGY OF THE BLOOD. When the blood is collected on glass, it may be dissolved by putting water on the glass and rubbing the dried blood in it until a decided red tinge is obtained. A drop of this mixture is then diluted and mixed with the bouillon culture. Johnson does not pay much attention to the dilution of the mixture of dried blood and water, before examination, as he does not find it necessary with his attenuated cultures. If necessary blood can be collected in wire loops of a given size, fairly accurate dilutions can be made. A. The Fluid Serum Quick Method. The ear is pricked in the ordinary way and about twenty drops are forced out by strong squeezing. The blood is received in a small (preferably two-inch) test-tube, with the edge of which each drop is scraped off the ear ; or we may suck the blood into a capillary pipette and expel it again into a test-tube or other receptacle. There is no need of cleansing the skin or sterilizing the test tubes in this method of procedure, as the whole process is finished up so rapidly that there is no time for contaminating organisms to grow. The blood when collected may be at once centrifugalized, and the plasma used for the test, or we may wait till clotting occurs and use the serum. When blood is collected in test- tubes, it is convenient to free the edges of the clot from the tube all round with some sharp instrument, so that the serum may not be pinned down underneath the clot, as it often is. If this is done, a drop of serum can be had within two or three minutes, and is then mixed with ten drops of bouillon culture, as above described, and examined at once between slide and cover-glass. (Dried serum can be used in the same way as dried blood, but has no special advantages and has not been frequently em- ployed by any observer.) B. The Fluid Serum Slow Method. This was the way originally described by Widal, or rather applied by him to the diagnosis of disease. The serum must be collected asepticalty, and many have therefore preferred to take it from a vein of the elbow, which is punctured with a sterile syringe, as described *on page 47. THE CLUMP REACTION. 407 Durham cleans the skin of the ear with a two-per-cent solu- tion of lysol, sucks blood into a sterile pipette, and blows it out again into a sterile test-tube to wait for clotting. Or, if we desire to keep and transport the fluid serum, it is- sucked into the bulb of a modified Pasteur's pipette (sterile), such as is shown in Fig. 58, which is then sealed by heat at the points A and B. In this way the serum will keep for an indefinite period and can be sent across the ocean, as was recently done at the request of the New York Health Department. When we are ready to use the serum, one of the pointed ends of the sealed bulb is broken off and the serum expelled by gently warming the other end. The serum aseptically collected by one of the above-described methods is then added : 1. To ten times its volume of bouillon culture of bacilli, i.e., eight drops to five cubic centimetres of culture, in a test-tube, which is then left from eight to twelve hours in the thermostat at 37 C. ; or 2. The serum may be added to ten times its vol- ume of pure sterile bouillon, and then a trace of the dry agar culture of bacilli added with a platinum loop and thoroughly mixed with the bouillon by rub- bing the loop against the inside of the test-tube, which is then kept twenty-four hours at 37 C. If the first of these ways is used, we get the effect of the serum on the fully grown bacilli; in the second way which usually needs fully twenty-four hours it works on the nascent and immature organisms. Whichever method is used, we find that within from eight to twenty -four hours a remarkable change takes place in the appearance of the culture when serum from a case, e.g., of ty- phoid fever, is added to typhoid bacilli, nascent or full-grown. The uniform turbidity of the bouillon is gone and the liquid is either clear with an abundant flocculent sediment at the bottom of the tube, or is filled with coarse whitish particles separated from each other by clear bouillon. The latter change may take place the instant the serum is added to the culture, but usually needs from six to eight hours, and the full end reaction is often not completed till twenty -four hours elapse. Fraenkel 408 SPECIAL PATHOLOGY OF THE BLOOD. finds the reaction most marked in twelve to fourteen hours less so in twenty -four. A control tube containing the same proportions of the same culture and of a healthy person's serum should always be put into the thermostat along with the serum to be tested. Occa- sionally in such a control tube fine but visible whitish dust forms, but such dust usually disappears later of itself, or can be dissolved and the original diffuse turbidity produced by shaking the tube, while shaking a tube in which the true clump reaction has taken place will not break up the clumps nor restore the original turbidity. As above suggested, the microscopical examination of the " dust" seen in such a test, or of the precipitate formed at the bottom of the tube, shows it to be made up of clumps of bacilli similar to those seen in the quick method, but generally larger. The reaction is considerably less typical when the serum used is dark-colored, but the effects of shaking the tube and the comparison with the control usually enable us to decide. Some precipitate and clumping may occur in cases not typhoid: (a) when the bouillon has not been filtered and con- tains sediment, in which the bacilli may become entangled ; (b) when a large amount of the dry culture is added (in trying the slow method on nascent bacilli) and not thoroughly mixed with the bouillon; (c) in case the platinum loop is not quite cooled before the agar culture is taken upon it ; (d) in case the culture is impure or the serum not aseptic. 3. Blister Fluid. Biggs and Park find the fluid obtained by blistering the most satisfactory. A fly -blister the size of a five-cent piece is applied, and in from six to eighteen hours a blister has formed. The serum from the blister is collected with a capillary tube, the ends of which are then sealed. This serum is admirably clear and free from blood corpuscles and answers the purpose well. This method has never been extensively used by other ob- servers, except Puglieri. Advantages and Disadvantages of the " Quick Method" and of the "Slow Method." In favor of the quick method are: (1) its quickness, (2) the small amount of blood needed, and (3) absence of any need for asepsis and of any danger of contamination. THE CLUMP REACTION. 409 Against it are : (a) the occasional occurrence of pseudo-reac- tions or false clumps, which will be discussed on page 418; (b) that it needs a microscopic examination instead of being evident to the naked eye, as in the slow method ; ' (c) that it needs watching and cannot be left to "go on of itself." In favor of the slow method are : (1) That to some observers it appears more reliable and less apt to give pseudo-reactions. (2) That it can be seen with the naked eye. (3) That we do not need to watch it but simply to note the results at the end of from eight to twenty-four hours. Against it are its slowness, the danger of contamination, the need of a large quantity of blood and of a thermostat." On the whole the great majority of observers prefer the quick method, and it has been used in three-fourths of the reported experiments. My own experience has been exclusively with the quick method. Breuer, Catrin, and Vanlair and Beco are the only ones who distinctly prefer the twenty-four-hour method in all cases. 2. THE CULTURES OF TYPHOID BACILLI TO BE USED. 1. The stock cultures grow best on agar. 2. Ordinary peptone bouillon, free from sediment, is the best medium for the test culture. It should be just on the verge of litmus acidity, giving no blue to the red paper and requiring 3.5 per cent of normal alkali to render it neutral to phenol- phthalein. 3. All observers agree that the cultures should be young that is, that the transplantation to bouillon should have taken place not more than from twelve to twenty-four hours before the culture is used. Many observers find even the twenty-four-hour culture too old and prefer a twelve- to twenty-hours-old culture in all cases. 4. The virulence and motility of the culture are very impor- tant. Most observers agree that the more virulent the culture the more readily and characteristically it is clumped by typhoid serum. Biggs and Park noticed that one culture of peculiarly 1 Greene states that with the quick method a mottling of the specimen can be seen with the naked eye. ? Pick states that no thermostat is needed, and that sedimentation takes place readily at room temperature. 410 SPECIAL PATHOLOGY OF THE BLOOD. great virulence recently received from Pfeiffer of Berlin worked much better in their cases than any other of the cultures used. I have repeatedly noticed that cultures recently taken from autopsies on patients who had died during the acme of the fever were much more easily clumped than those taken in autopsies on patients who had succumbed late, after the tem- perature had been normal for some time. I have also noticed that virulent cultures grown for a long time in the thermostat with weekly transplantations gradually lost a good deal of their susceptibility to the clumping power of typhoid sera. Presumably these changes mean a loss of virulence in the culture, especially as they have always been accompanied by a diminution in the rapidity of motion in the bacilli. Cultures fresh from an autopsy usually show furious motility, the bacilli darting about like a swarm of insects, but after repeated trans- plantations and long sojourn in the thermostat a good deal of this motility is gradually lost. Cultures kept at room tempera- ture preserve their motility for much longer periods. For those who have no opportunity to test the virulence of organisms on animals, the motility is the best guide to virulence, and the rule should be : Among the available cultures select that having the most rapid motility. 4. Certain cultures contain small clumps of bacilli before any serum has been added to them. This is a very important point and has doubtless misled many. In consequence of this possi- bility every culture must be examined each time that a test is made. It is not sufficient to examine each culture once for all, as cultures vary slightly from day to day and also vary in dif- ferent portions of the culture tube. For instance, ten drops taken from the middle of the bouillon may be found free from clumps, while if the next ten drops be taken from the surface or from the bottdm of the liquid, they may contain clumps. This point has been strongly insisted on by Widal, Eenon. and others. 5. It is hardly necessary to say that the cultures used must have been submitted to all the regular tests for the recognition of the typhoid bacillus, and that the greatest care must be used to avoid. their contamination. THE CLUMP REACTION. 411 The Use of Suspensions or Emulsions of the Bacilli instead of Cultures. A few observers particularly Durham and Griiber have preferred to use a mixture of small bits of solid agar culture and bouillon instead of bouillon cultures. The majority of writers prefer cultures. The Use of Attenuated Cultures. Johnson finds that with his methods of technique (dried blood and no definite dilution) pseudo-reactions were not uncom- mon with the blood of healthy people. He avoids this by using attenuated cultures i.e., old stock agar cultures kept at room temperature and not transplanted more than once a month, from which he planted his bouillon cultures. This gives a bacillus of reduced virulence and slow, gliding motion, which is clumped far less readily than the virulent varieties. Bouillon cultures of this kind from twelve to twenty-four hours old he found to react in fifteen minutes with all typhoid sera and not with other sera even after forty- eight hours' waiting. Durham, Biggs and Park, and Delepine, on the contrary, found such cultures unsatisfactory, in that it was not possible to avoid pseudo-reactions with sera of diseases not typhoid. I have been equally unsuccessful with this method, and believe with Biggs and Park that the most virulent cultures are the most reliable, if fluid blood or serum is used. When dried blood must be used, the attenuation of cultures as advised by Johnson is probably the best plan. The Clump Reaction with Dead Bacilli. One of the most remarkable and interesting features of the clump reaction is the possibility of obtaining it with bacilli that have been killed by heat or by formol. Widal observed that bouillon cultures of typhoid bacilli ex- posed to a temperature of 57-60 C. for one-half to three- quarters of an hour lost scarcely any of their susceptibility to the clumping action of typhoid serum, though they are quite dead. Higher temperatures (70-120 C.) take away more and more of the susceptibility to clumping and also cause 412 SPECIAL PATHOLOGY OF THE BLOOD. the formation of false clumps without the addition of any serum whatever. Similarly one drop of ordinary formol mixed with one hun- dred and fifty drops of bouillon culture of Eberth's bacilli kills them, but apparently " embalms" them, so that their suscep- tibility to clumping is scarcely if at all lessened, even after the lapse, of five months. The bacilli gradually sink to the bottom of the tube, but when shaken up distribute themselves evenly throughout the medium and can be used like fresh cultures for diagnostic pur- poses. Bordet has noted the same thing with cultures of the cholera-vibrio killed with chloroform. These facts seem at first sight to conflict with the statement made above, that fresh, motile, and virulent cultures are best, and that old ones are not reliable. But it may be, as Widal sup- poses, that the rapid action of heat or formol on virulent cultures preserves unchanged the power which prolonged growth in old media destroys. If this be true, it will enable us to dispense with our thermostat and careful nursing of cultures, since a single first-rate culture can be thus "em- balmed" and preserved for use at all times and under all cir- cumstances. Widal' s results with this method have not yet been confirmed by others. 3. DILUTION AND THE TIME LIMIT. I. Dilution. We have mentioned without explanation in various parts of this chapter that the blood serum or other fluids used must be diluted with at least ten times their volume of bouillon culture before any observation is made as to their action on the bacilli of typhoid fever. The reasons for this dilution and for the proportions 1 : 10 are the following : It has been found, as mentioned above, that the mere forma- tion of clumps in bouillon cultures of Eberth's bacilli is not a power exclusively possessed by typhoid serum. The serum of persons suffering from other diseases and even of healthy persons will form clumps exactly like those formed by typhoid bacilli, THE CLUMP REACTION. 413 provided it is not diluted. The only known peculiarity of the typhoid serum is that its clumping power is greater than that of other diseases, and persists in spite of dilution, while the sera of diseases other than typhoid lose their power to clump typhoid bacilli when diluted ten times or more. II. Time Limit. But even this statement must be further limited. The sera of various other diseases, and of healthy persons, will sometimes clump typhoid bacilli even in a 1:10 dilution, provided ice give them time enough. We must therefore limit the period within which a serum must " come up to the scratch" and do its work, if it is to bs considered a typhoid serum. Following Griiber and Durham, a time limit of one-half hour has been adopted by Griinbaum, Block, Haedke, Park, and others. All that these more or less arbitrary figures stand for is this : that hitherto no one has reported any considerable number of cases in irliich the serum of any disease or of healthy persons has clumped typhoid bacilli within one-half hour, wlien diluted 1:10 and used with unimpeachable technique. Johnson finds dilution unnecessary with his methods of pre- paring the cultures, and Widal only lately has admitted the ne- cessit} T of a time limit, but the majority of careful and non-par- tisan observers are agreed that these precautions are necessary unless attenuated cultures are used. If at any time cases are re- ported in which, despite these precautions, a clumping of typhoid bacilli has occurred with non-typhoidal sera, it will be necessary to raise the dilution to 1 : 15 or 1 : 20. Indeed there are many who think it should now be placed at one of these two figures or even higher. Many German observers prefer to use a dilu- tion of 1 : 40 or 1 : 50 and a time limit of one or two hours. This amounts to about the same thing as 1 : 10 dilution with a limit of fifteen minutes. . The clump reaction in typhoid fever is to be considered specific and pathognomonic only in the sense that it occurs more readily and in presence of greater dilution in typhoid than in any condition yet reported. (For details and exceptions on these points see page 370.) The serum of most cases of typhoid fever during the second 414 SPECIAL PATHOLOGY OF THE BLOOD. week will clump typhoid bacilli when diluted 1 : 40, and many sera preserve the power even at 1 : 100 or higher. Widal has seen it as high as 1 : 12,000. The following table .from Biggs and Park illustrates these points well : Q History, symptoms, and diagnosis at time of taking blood specimens. Corrected diagnosis on completion of illness. Reaction of bacilli in broth cultures to serum in different dilutions. Reaction. Amount of serum. Amount broth culture. 1 Adult; sick four weeks, con- tinuous high fever; pleurisy ; " tuberculosis " with possi- bility of typhoid. Tuberculosis. 1 1 Not appreciable. 2 Boy; sick two weeks; con- tinued moderate fever, abat- ing when test was made; prostration, constipation ; no typhoid symptoms except fever and prostration; "atypical typhoid fever." Uncertain. 1 1 Not appreciable. 3 Adult; symptoms of acute articular rheumatism only; " acute articular rheuma- tism." Acute rheu- matism. 1 1 1 1 4 9 Delayed moderate. Delayed very slight. Not appreciable. 4 Adult; just convalescent after sickness giving character- istic symptoms and physical signs of pneumonia; " pneu- monia." Pneumonia. 1 1 1 1 1 4 9 19 Immediate marked. Delayed moderate. Delayed slight. Not appreciable. ft Adult; continued high fever; enlarged spleen; typhoid bacilli obtained from spleen; "typhoid fever." Typhoid fever. ] 1 1 1 4 9 Immediate. Delayed incomplete. Delayed very slight. 6 Adult; relapse after four weeks of continuous fever with typhoid symptoms; "relapse after typhoid fever" Typhoid fever. 1 1 1 1 1 1 10 50 100 200 Marked immediate. Marked immediate. Marked immediate. Delayed moderate. Dela}'ed slight. 7 Adult; seven days continued high fever; typhoid symp- toms; two days later an atypical rash ; "typhoid fever." Typhoid fever. 1 1 1 1 1 1 9 49 99 199 Marked immediate. Mai-ked immediate. Marked immediate. Delayed but marked. Delayed moderate. The Microscopic Examination. An artificial light is preferable. The use of hanging-drop preparations is unnecessary, as a simple slide and cover-glass is satisfactory. A hanging-drop cell may be extemporized by cementing with marine glue a small brass curtain ring to a slide, and inverting the cover-glass within it, as advised by Stokes. THE CLUMP REACTION. 415 SEED-DIAGNOSIS or TYPHOID. I have collected over 3,000 cases of supposed typhoid fever in which the clump reaction was tested as above described either with the fluid or dried blood. Of these, 95 per cent showed a serum reaction at some time in their course; 2,500 odd controls showed about 2 per cent of positive results in cases other than typhoid. Altogether then about 5,500 cases have been tested. If we leave out the reports of those whose experience covers less than 100 cases, we have left 4,339 cases observed by 18 physi- cians in which the percentage of error is 2 per cent only. There seems to me no doubt of the fact that the serum reac- tion is present in some part of the course of 95 or more per cent of all cases of typhoid fever, and absent in 95 or more per cent of all other conditions. But it is also true that it is absent in some part of the course of many cases of typhoid usually in the earliest or latest days of the fever and this fact makes it necessary to retest every case in which a negative result has been found, and even in some cases to make a considerable number of tests before a positive result is ob- tained. My own experience covers 202 cases of ty phoid, all but 7 of which were positive, and in 4 of these there was no op- portunity of retesting. In the last 108 cases, all of which were carefully retested, there has not been a single failure, though in some the reaction was very late. Out of 376 controls one re- acted like typhoid, a case of pernicious anaemia in a negro. Hoiv early does the reaction appear ? Few of the many observers who have written on this point have discussed how the beginning of the disease is settled and what they mean, e.g., by the "fifth day of the disease." It might be dated from the first day of malaise arid indisposition, from the nose-bleed or the beginning of headache, or from the time of going to bed. Allowing for such serious uncertainties as this, we find that while the majority of observers record the sixth to eighth day as the earliest on which the reaction appears, there are quite a number of cases mentioned in which it was seen on the fourth or fifth day ; a few record reactions present on the third day, and two or three on the second dav. 416 SPECIAL PATHOLOGY OF THE BLOOD. As above mentioned, we have no way of knowing what the "second day" means in these cases. In my own observations I have called the first day in bed the first day of the disease (though I am aware that many patients are sick some time before taking to bed), because it was the only date that could be definitely fixed in all cases. With this nomenclature I have found the reaction present on the first day in two cases and on the second in three cases. Counting from the first day on which the patient felt sick in any way, the fifth day is the earliest reaction day in my series. In these figures we have always to remember that in no case was the blood tested at all previous to the day on which the positive test occurred, so that their meaning is: In some cases (what proportion of all is unknown) the serum reaction occurs at least as soon as the fifth day of malaise or the first day in bed, andperlui/^ sooner. Not infrequently the reaction antedates the appearance of rose spots, splenic enlargement, or the diazo reaction by several days. On the other hand, I have known the bacilli to be isolated from the stools before the serum reaction ap- peared; but this is rare. Experiments on animals show that the clump reaction ap- pears in the blood on the third to eighth day after inoculation with dead typhoid bacilli. How late in the disease does the reaction last ? The majority of observations agree that in mild cases the reaction may die out even before the end of the fever. On the other hand, the re- action usually lasts several months, and Widal found it still present after one year in 3 out of 22 cases in which he tried it. These 3 subjects had had very severe cases of typhoid three, seven, and nine years previously. It has been reported present twenty and even thirty years after the fever. Biggs and Park found the reaction more constant in the fourth week than at any other time 76 per cent of their cases tested between the thirtieth and sixtieth days still showed the reaction, and 5 of 8 cases still reacted after three to four months. The reaction almost always persists in relapses, even to a second or third relapse, and occasionally it is present only in relapse and not in the original attack at all. Biggs and Park record a case in which the diagnosis was proved during the original attack by puncture of the spleen, which showed a pure THE CLUMP REACTION. 417 culture of Eberth's bacilli, yet no serum reaction was present until the second day of the relapse. I have observed several similar cases, and quite frequently not found the reaction until convalescence. The failure to follow up such cases as these accounts for many negative reports. In one of Elsberg's cases the total duration of the clumping power in the blood was only eight days ; in another only twelve days. The continuance of the reaction after the fall of the tempera- ture is no indication (as some have supposed) that relapse is coming, for in many such cases no relapse follows. Very fre- quently the reaction is absent on a given day though present the day before and after. I The Intensity of the Reaction. Widal and others have studied the intensity of the reaction at different periods of the disease, judging by the amount of dilution which could be practised without destroying the power of a given serum. Examples of this have already been given in the table on page 363. The majority of typhoids in the second and third week yield serum which will clump Eberth's bacilli when diluted 1 : 40 and many cases will do so even at 1 : 100. Strangely enough, some typhoid sera clump better when diluted 1 : 16 or more than when undiluted. This has been repeatedly noted by Grunbaum. Widal and Sicard record clamping with a dilution of 1 : 12,000 and 1:1,800 and consider that in the active stages of the dis- ease a dilution of 1 : 60 or 1 : 80 does not usually present the re- action, while in convalescence the power of the serum falls off gradually and is not always present even at 1 : 10. Biggs and Park find one-half their typhoid cases furnish serum with the power to clump in 1 : 40 dilution by the end of the first week, and have occasionally noted the reaction even with a dilution of 1 : 200. Jemma found the reaction most intense at the acme of the fever and greater during the evening exacerbation of fever than in the morning. 27 418 SPECIAL PATHOLOGY OF THE BLOOD. EFFECTS OF THE SERA OF OTHER DISEASES. Negative results are reported in the following list of diseases experimented on as controls : Pneumonia, typhus, Malta fever, tuberculosis in its various forms, including miliary tuberculo- sis, tubercular meningitis, pneumococcus meningitis (purulent) and epidemic cerebro-spinal meningitis, diphtheria, influenza, ulcerative endocarditis, erysipelas, puerperal septicaemia, gonor- rhceal septicaemia, measles and scarlet fever, tonsilitis, acute articular rheumatism, malaria, leprosy, syphilis, bronchitis, pleurisy with effusion, acute and chronic nephritis, mumps, oti- tis media, catarrhal jaundice, sciatica, acromegalia, hysterical vomiting, Graves' disease, gangrene of the lung, appendicitis, abscess and cirrhosis of the liver, acute febrile gastro-enteritis (" embarras gastrique"), cancer of the various organs, alveolar abscess with fever, osteomyelitis, bubo with fever, arthritis de- formans, chronic laryngitis, intestinal obstruction, general peri- tonitis, leukaemia, Hodgkin's disease, pernicious anaemia, diar- rhoea, chronic gastritis, gallstone colic with fever, dysentery, acute mania, stuporous melancholia, synovitis, neurasthenia, varicose veins, orchitis, suppurative thyroiditis, perinephritis, cystitis, pericarditis, empyema, brain abscess, valvular heart disease, diabetes, gas poisoning, alcoholism, and eclampsia. The important diseases of this list, such as pneumonia, tuberculosis, meningitis, and typhus, have been tried many times. Biggs and Park got a positive result in one case said to be typhus. There is a chance of mistaken diagnosis here. Positive Results of the Sera of Other Diseases ivith Typhoid Bacilli. Many of the supposed contradictions of the law, that the typhoid bacilli are clumped within one-half hour only by typhoid serum when a dilution of one part of serum to ten or more of culture is used, are due to faulty technique. Such are probably the cases reported by Ferrand and Theoari (septicaemia), Villies and Battle (malaria), Gehrmann and Wynkoop (pneumonia, bron- chitis, pleurisy), and Stern (otitis media). On the other hand there are a few cases reported by careful observers in which a genuine clumping of typhoid bacilli has been caused by the sera of other diseases, viz. : Pernicious THE CLUMP REACTION. 419 malaria, comatose, one case (Block) ; diabetic coma, one case (Block) ; jaundice, one case (Catrin) ; tubercular meningitis, one case (Jez), and a few more. A case of malaria, reported by Catrin, with positive reaction of the serum on typhoid bacilli, was in a subject who had had typhoid five years before. In view of Widal's and Fraenkel's results, this cannot be counted an exception to the general law. The same is true of Griinbaum's much-quoted cases, which he reported not as exceptions but to emphasize the necessity of proper dilution. Using a proportion of 1:1 instead of 1 : 10, he got clumping of typhoid bacilli with the sera of jaundice (two cases), meningitis and bronchitis (one case each). The cases reported by Johnson, Brannan, Thomas, Heed, and other observers, in which the dried blood of healthy per- sons and persons with various diseases other than typhoid has clumped typhoid bacilli, are probably owing to the uncertain- ties connected with that method of procedure. In most cases in which the fluid serum was also tried it gave no reaction. The discovery that the bacillus of psittacosis and the " bacillus enteritidis" of Gartner are somewhat sensitive to the action of typhoid serum (see page 421) has led to the fear that infec- tions due to those bacilli might be mistaken for typhoid, but this is wholly an assumption, as in the few cases of these infec- tions which have been studied the serum has not affected typhoid bacilli. Further, it is only by a concentrated artificial typhoid serum that the bacillus of Gartner is clumped, and the clumping of the psittacosis bacillus is quite different from that of the typhoid bacillus, the clumps of the former being very small and few; with the twenty-four-hour method no precipitate forms. Summary of Negative Results. Out of over three thousand cases of various diseases not ty- phoid, not over a dozen have been proved to clump typhoid ba- cilli with proper technique. It is quite possible that further im- provements in technique may enable us to prevent even this very small error. 420 SPECIAL PATHOLOGY OF THE BLOOD. EFFECTS OF TYPHOID SEBUM ON OTHER BACILLI. (a) On the Bacillus Coli Communis. Any blood serum mixed 1 : 10 with a bouillon culture of colon bacilli may cause the formation of small clumps without consid- erable loss of motility. The effect of typhoid serum does not differ from that of other sera, and the clumps which it forms are much smaller and looser than those seen in the typical typhoid clump reaction. Different cultures of colon bacilli differ a good deal in their susceptibility to typhoid serum, and Vanlair and Beco consider that no difference can be made out in certain cases between its effects on typhoid bacilli or on colon bacilli. The majority of observers, however, find a decided difference, espe- cially with the twenty-four-hour method. Undiluted typhoid serum acts more strongly on colon than on typhoid bacilli, ac- cording to Grunbaum. Biggs and Park found that " a number of varieties of motile bacilli other than typhoid bacilli are clumped by the serum of persons suffering from typhoid fever, even when the serum is used in quite high dilutions." Eodet noted that only a very slight effect is produced by ty- phoid serum on colon bacilli until a dilution of one part of serum to two of culture is reached. Fraenkel tested a largo number of colon cultures without getting any decided effect from the addition of typhoid serum. Courmont found that some cultures of the colon bacillus are clumped by typhoid serum. Widal saw no difference between the effect of typhoid serum and that of other sera on colon bacilli, but Vedel thinks that young cultures are better clumped by typhoid serum than by other sera. Johnson, who studied a large number of cases, says : " A complete colon reaction we have found to be exceptional in or- dinary typhoid, and its presence would indicate a condition of coli intoxication, " which may be held to sum up the discussion up to the present time. (b) On the Bacillus Enteritidis (Gartner). Griiber and Durham, using powerful artificial sera from animals immunized against Eberth's bacillus, were able to obtain THE CLUMP REACTION. 421 a clumping of Gartner's organisms. No experiments with hu- man serum are recorded. (c) On the Bacillus of Psittacosis. Psittacosis is a disease affecting parrots and occasionally transferred by them to human beings. A bacillus has been found by Nocard in the marrow of the parrot's wing-bones which is considered the cause. Typhoid serum has an effect on bouillon cultures of this bacillus, which is to be distinguished quantitatively from the clumping of typhoid bacilli by typhoid serum; the heaps of psittacosis bacilli are much fewer and smaller, and in the twenty-four-hour method the turbidity of the cultures does not disappear. (d) The Klebs-Loeffler Bacillus and Pus Cocci. Courmont finds that typhoid serum clumps Klebs-Loeffler bacilli and staphylococci, but is without effect on the strepto- coccus and the bacillus pyocyaneus. Summary of Clinical Evidence on the Sero-Diagnosis of Typhoid Fever. The blood of over ninety-five per cent of all cases of typhoid shows a clumping power in some part of their course, but in at least half the cases this does not appear until the second week of the disease, while in a small number of cases it first appears in relapse or convalescence. The clumping power may disap- pear before the defervescence and may be present only eight days in all ; as a rule it persists from the sixth or eighth day until convalescence is established. In diseases other than typhoid a clump reaction is very rarely to be obtained, provided a dilution of at least 1 : 10 is used with a time limit of one-half hour. There is no one dis- ease in which clumping is especially apt to occur. Clinically the reaction is of considerable value, especially when the diagnosis is in doubt after the first week of the dis- ease. 422 SPECIAL PATHOLOGY OF THE BLOOD. SERO-DIAGNOSIS OF DISEASES OTHER THAN TYPHOID. 1. Cholera. Griiber and Durham first showed that human cholera serum would clump cholera vibrios, following the researches of Pfeiffer in vivo by demonstrating a similar reaction in vitro. Achard and Bensaude have applied this to the actual diag- nosis of cholera in man with considerable success. In fourteen cases, thirteen clumped readily ; two of these were on the first day of the disease. Thirty control cases were negative. The presence of the pellicle renders it unsafe to use bouillon cultures except such as have no pellicle, for bits of it are much like true clumps. Suspensions of twenty-hour gelatin cultures are more convenient. The dilution and time limit are the same as in typhoid. Some cases will react even in 1 : 120 dilution. The reaction can be performed with dried blood and persists into convalescence (seven months or more). 2. Pyocyaneus Infections. The bacillus pyocyaneus has been shown to be in all proba- bility the cause of certain cases of dysentery, broncho-pneu- monia, otitis media, nephritis, pericarditis, cystitis, and of a hemorrhagic septicaemia with enteritis in the new-born. Roger and Charin found in 1889 that the bacillus pyocyaneus is serum of animals immunized against this bacillus. Durham repeated these observations in 1895 and confirmed them. Here we have the clinical infection and laboratory clump- reaction experiments, but so far as I am aware no one has yet brought the two together or tried the serum of patients with pyocyaneus infections on cultures of the bacillus. 3. Diphtheria. Widal reports no success in attempts at the sero-diagnosis of diphtheria, and Fraenkel has not been more successful. Nicolas and Charrin found that, although no true serum reaction could be obtained in diphtheria previous to antitoxin treatment, the injection of antitoxin produces in the patient's serum a de- cided clumping power over the Klebs-Loeffler bacilli within twenty-four hours of the time of injection. This is especially marked in the twenty-four-hour method, using the nascent bacilli, THE CLUMP REACTION. 423 as described on page 407. The serum retains its clumping power for about two weeks after the injection of antitoxin, and then gradually loses it. Outside the body the diphtheria antitoxin easily clumps Klebs-Loeffler bacilli. 4. Pneumococcus Infections. Washburn in 1895 noticed that pneumococci, when mixed with artificial antipneumococcus serum and left twenty-four hours at 37C, were clumped in masses at the bottom of the tube, leaving the upper portions of the liquid clear. In other words he got a typical twenty-four-hour clump reaction, using a powerful arti- ficial serum. The same fact had previously been observed by Metchnikoff in 1891 and by Issaef in 1892, and has been recently confirmed by Mosny. Widal has been entirely unsuccessful in finding any clump- ing with the serum of pneumonia patients, and Block finds the lumping of pneumococci very slow and unsatisfactory. Be- zancon and Griffon get similar results. . Colon-Bacillus Infections. In view of the frequent association of this bacillus with dis- ease, especially with the cystitis of young girls, it is important that the possibility of a sero-diagnosis of colon-bacillus infec- tions should be studied, but as yet very little has been done in this direction. Griiber and Durham showed that serum from animals arti- ficially immunized against the colon bacillus would clump that bacillus strongly, but Widal' s early experiments with supposed cases of colon-bacillus infection did not show any decided reac- tion, nor did the serum of typhoids which showed post-mortem a secondary colon-bacillus infection react during life on cul- tures of this bacillus. Widal has lately claimed that in any case of colon infection only the particular race of bacilli which are actually causing the case in question can be specifically clumped by the serum of that case. In a case in which he first isolated the bacillus and then used it with the patient's serum he got a clumping as high as 1 : 1,000 dilution. Lesage in an epidemic of infants' diarrhoea found that the serum of 40 out of 50 cases clumped the colon 424 SPECIAL PATHOLOGY OF THE BLOOD. bacillus isolated from the stools. Each of the 40 sera clumped each of the 40 cultures isolated from the 40 cases. The colon bacillus of the normal intestine of infants of the same age was not at all affected by the serum of the sick children, nor did the serum of normal infants clump the organisms from the infected children. This tends to confirm Widal's assertion. Appar- ently this epidemic was due to a single bacillus which could be easily isolated and used for experiment. The difficulty of iso- lating the bacillus of every case that is to be tested renders the method of very limited clinical value. 6. Malta Fever. Wright and Smith tested the serum of 15 cases of Malta fever with the micrococcus melitensis of Bruce, and found a strong clump reaction to occur (1:50 in most cases). On the typhoid bacillus the serum of these cases had no action. Sixteen cases of typhoid showed no reaction with Bruce' s organism. The evidence in favor of this organism as the cause of Malta fever is strengthened by these facts. 7. Peripneumonia of Cattle and Hog Cholera. Arloing finds that the serum and other body fluids of cattle suffering from peripneumonia have a marked clumping power on the pneumobacillus bovis. Dawson has had similar positive results working with the bacillus of hog cholera. Hog-cholera serum had no effect on the typhoid or colon bacillus. 8. Proteus Infections. Infections with the proteus vulgaris or proteus mirabilis have been considered causative in cases of mastoid abscess, meningitis, and Potts' disease. When found by culture at autopsies the question often arises whether they have wandered in after or at the time of death, or whether they were really con- cerned in the etiology of the case. The investigations of Achard and Lannelongue appear to give us the means of answering this question. They found that cultures of the two species of pro- teus above mentioned were markedly clumped by the serum of animals rendered immune to them by inoculations. This power THE CLUMP REACTION. 425 persists after death and even in putrefaction, and if present at any given autopsy proves that the infection did not take place during the last two days of life, since it takes at least three days to bring the clumping power into the serum by artificial inocu- lation. 9. Oidium Albicans. Roger showed that the oidium was well clumped by the serum of animals immunized against it, and these observations have been confirmed by Charrin and Ostrowsky. No experi- ments with human thrush have as yet been reported. 10. The Bubonic Plague. Zabolotny 1 studied forty cases at Bombay in April, 1897, and found the reaction absent in the first week, present in 1 : 10 dilu- tion in the second week, and in 1 : 50 dilution in the third or fourth week. He noted that the action of the infected serum seemed to deprive the bacilli of their capsules. In an editorial in the Arch. Busses de Pathologie, May 31st, 1897, it is stated that the reaction increases in intensity until the fourth week of the disease and then declines ; also that it is most marked in the severest cases. Feindel (loc. cit.) states that in the acute pneu- monic cases the reaction is absent. 11. Yellow Fever. Sanarelli 2 states that the organism which bears his name is clumped very strongly and speedily by the serum of dogs im- munized against his bacillus. By normal human serum it is not clumped at all. By patients with yellow fever it clumped very slowly. Post mortem the serum clumps more readily but very variably. Pothier's 3 experience in the recent New Orleans epi- demic has been similar. 12. Eelapsing Fever. (a) Diagnosis. In countries where this disease is common the difficulty of diagnosing cases between attacks (when the. 'Deut. med. Woch., 1897, p. 392. 2 Annales de 1' Institute Pasteur, October 27th, 1897. 3 Personal letter. 426 SPECIAL PATHOLOGY OF THE BLOOD. spirochsetes are absent from the blood) is frequently met with. Lowenthal has perfected a method by which in most cases the diagnosis can be made by means of the effect of the serum of suspected cases on the spirochsetes of other active cases. The organism cannot be cultivated as yet, so that a diagnosis of this kind is possible only during epidemics when fresh blood containing the organism can be obtained. A drop of blood from the suspected case is mixed with a drop from a patient then undergoing a paroxysm, and the two are sealed with wax be- tween slide and cover-glass and left in the thermostat for half an hour together with a mixture of normal blood and blood con- taining spirochsetes as a control. At the end of that time, if the case be one of relapsing fever, the organisms in contact with the blood from that case cease their motion, while those in the control are lively. It is not a clump reaction but a direct bac- tericidal effect which persists in the serum nearly up to the time of the next attack. The diagnosis so made by Lowenthal in forty cases was verified in every case by the course of the dis- ease. In this way mild or abortive cases with few organisms in the blood can also be identified. (6) Prognosis. If the above bactericidal power lasts as late as the seventh day from the last attack, and in sufficient intensity to immobilize the spirochsetes in one hour or less, there will be no relapse. If these conditions are not fulfilled relapse is sure to follow unless prevented by treatment. Lowenthal has veri- fied this prognostic use of the serum in over one hundred cases. 13. Miscellaneous Reports on Other Infections. (a) Griinbaum (Lancet, February 13th, 1897) states that a "non-motile diplococcus" from a case of scarlet fever was clumped by the serum of another case of scarlet fever. (&) Delepine (Medical Chronicle, October, 1896) refers to suc- cessful experiments with the tetanus bacilli its antitoxin hav- ing a decided clumping action upon it. The serum of normal horses clumps the tetanus bacillus to some extent and that of tetanized horses clumps it intensely. In eight cases referred to by Feindel ' the bacillus was clumped by the serum of human tetanus. 1 Arch. Gen. d. Medecine, October, 1897. THE CLUMP REACTION. 427 (c) Durham (Lancet, loc. cit.) speaks of the present antistrep- tococcus serum (Marrnorek's) as having strong clumping power on streptococci, but Masino ' finds very few cultures react to it. (d) Gilbert and Fournier (Compt. rend, de la soc. de biol., December 25th, 1896) mention two cases of human psittacosis whose serum clumped well the bacilli obtained from another human case as well as those taken from parrots. Clumping was present on the fourth and fifteenth days respectively. SEBO-PROGNOSIS. It is agreed by all observers that in a very general way severe cases have more marked reactions than mild ones, but beyond this, in the opinion of the best judges, we cannot yet go. Widal, Fraenkel, Biggs and Park, and Johnson have at- tempted no sero-prognosis, and my own observations are en- tirely in accord with this. The reaction may be strong in mild cases and feeble or absent in fatal ones. Certain writers, however, especially Breuer, Courmont, Catrin, and Ullmann and Wohnerfc, have thought the reaction of prognostic value, an intense and early reaction seeming to them of evil omen. Further evidence on this point is much needed. [For bibliography, see page 429]. 1 La Semaine Medicale, 1897, p. 114. BIBLIOGRAPHY. IT has seemed to me best to give a list only of the books and articles which I have found most useful, since the general bibliography of the subject is now large enough to form a volume by itself. Most of the larger works here described con- tain extensive bibliographies especially that by Grawitz. Text-Books. 1. Hayem: " Du Sang," Paris, 1889, 8vo, 1035 pages (French). This valuable book is the largest that I know of on the subject, and contains a mine of information on the morphology of the blood in health and disease, mostly from the author's own experience, literature being but little re- ferred to. It contains a comparative anatomy of the blood and a long account of blood development. Unfortunately, it is dominated through- out by a theory of blood formation which has never gained acceptance by any other authority. It is very full on the subject of fibrin formation and of chlorosis. The illustrations are excellent. 2. v. Limbeck : " Grundriss ein. klin. Pathologie des Blutes, " Jena, 1896, 8vo, 383 pages (Fischer). The second edition of this book, which appeared in February, 1896, is more than twice the size of the first edition (1892) a fact illustrating the rapidity of the subject's growth. It is on the whole the best general text-book known to me, being equally full on all parts of the subject, including, for example, technique (which Grawitz omits) and of the chemistry of the blood, which is at present the author's special interest and on which Hayem is meagre. The illustrations are poor and the type is trying to the eyes. The writer shows little personal experience with the morphology and micro-chemistry of the blood, and this is the weakest side of the book. A large part of the book is concerned with the physiology of the blood. 3. Grawitz: "Klinische Pathologie des Blutes," Berlin, 1895, 8vo, 333 pages (Enslin). Issued in April, 1896. This book is largely devoted to the matter indicated by the title and contains no account of blood technique, and only thirty pages on the normal anatomy and physiology of the blood, while two hundred and seventy concern the blood in disease. The arrangement of the book is very clear and helpful. The author's main interests are in the estimation of the dried residue of the blood in various diseased condi- tions and in the bacteriology of the blood, so that the book is specially full on these topics. The illustrations are poor. Type and paper are excellent. 4. Ehrlich and Lazarus: "Die Anaemie," Wien, 1898, 8vo, 142 pages 430 BIBLIOGRAPHY. (Holder). An account of the normal and pathological histology of the blood not only in anaemia, but in all diseases of the blood, containing the latest researches and admirably clear as regards the microscopic appear- ances. This is Ehrlich's latest utterance his last book previous to this being issued in 1891 (vide infra) . It contains also a good deal of inter- esting theoretical discussion OD the sources and formation of the elements of the blood. There are but three rather indifferent illustrations (un- colored). The book forms part of vol. iii. in Nothnagel's new "Special Pathology and Therapeutics, " but is issued separately. 5. Coles: "The Blood: How to Examine it, "etc., London, 1898 (J. and A. Churchill) , 8vo. A very clear account of our present knowledge on the subject. Especially full on technique. 6. Schmaltz: "Pathologie des Blutes und die Blutkrankheiten," Leip- zig, 1896, 16mo, 268 pages (Naumann). A much smaller book than any of the others and including the symptoms, pathology, and treatment of blood diseases, as well as a pathology of the blood itself. Specific gravity of the blood is a point of special interest with the author. There are no illustra- tions. The book is excellent as far as it goes, well arranged, and clear. These are the best text-books known to me on the whole subject. None of them have been translated. Text-Book Articles on Blood Diseases. 1. Stengel's article in vol. vii. of the "Twentieth Century Practice of Medicine" is by far the best text-book article in English that I know of. 2. Osier, in the " American Text-book of the Theory and Practice of Medicine," vol. ii. (Philadelphia, 1894, Saunders), writes an excellent fifty -page article on " Diseases of the Blood." It covers, of course, only the blood diseases proper without much account of the blood in other con- ditions. 3. The article "The Blood in Infancy, " in Rotch's Paediatrics, covers this branch of the subject very thoroughly. These are the best articles in English that I know of. 4. The article on " La Pathologie du Sang, " by Gilbert, in the five- volume " Traite de Medecine" edited by Charcot, Bouchard, and Brissaud, Paris, 1892 (Masson), is inferior to those last mentioned and is mostly an echo of Hay em's work above referred to. Theories long exploded (e.g., that eosinophiles are pathognomonic of leukaemia) receive the author's sanction. The article is one hundred large octavo pages long and is in- tended to cover the whole subject. 5. Griffith's eighty-page article in Keating's " Cyclopaedia of the Dis- eases of Children," vol. iii., p. 755 (Philadelphia, 1890, Lippincott), is now a good deal out of date. 6. The articles on blood diseases in the latest editions of the text-books of Osier, Strumpell, Da Costa, Flint, and Fagg, contain relatively little about the blood itself. BIBLIOGRAPHY. 431 Treatises on Special Portions of the Subject. 1. Reinert's "Die Zahlung derBlutkorperchen," Leipzig, 1891 (Vogel), 246 pages, is an admirable account of the avoidable and unavoidable errors in blood examination, and the best methods of reducing error to a minimum. A number of careful examinations of the blood in health and in various diseases are also given ; and an outline of the scope of blood diagnosis closes the book. 2. v. Noorden's "Chlorosis" (Wien, 1897, 8vo, 209 pages) is by far the best piece of work so far published on this subject. The clinical and therapeutic sides of the subject are fully treated, as well as the hsematolo- gical pathology. 3. Turk's monograph on the " Condition of the Blood in A cute Infec- tious Disease" is an admirable resume of German and French literature on the subject, together with a detailed study of fifty-two cases. Published at Wien and Leipzig, 1898 (Braumuller), 347 pages, 8vo. 4. Rieder's "Beitrage zur Kenntniss der Leukocy tosis, " Leipzig, 1892 (Vogel), 220 pages, is an admirable work in all respects, although now considerably out of date. It shows, as very few of the foregoing treatises do, a practical acquaintance, on the author's part, with the details of blood morphology and microchemistry. A very large number of blood counts in many diseases are recorded. 5. Lowitt's "Studien zur Physiol. und Pathol. des Blutes u. der Lymphe" (Jena, 1892 [Fisher], 8vo, 138 pages) is mostly concerned with experiments on animals and intended to throw light on the theory of leu- cocytosis. The conclusions of the book have not been generally adopted, though its facts have been mostly verified. 6. Thayer and Hewetson's book, on the "Malarial Fevers of Baltimore," leaves nothing more to be desired in that direction. It is two hundred and fifteen pages long, published by the Johns Hopkins press of Baltimore in 1895. It contains a summary of the literature of the subject, an analy- sis of six hundred and sixteen new cases, and some admirable colored plates. It is a model of its kind in every respect, and an ideal for others to aim for. 7. Ehrlich's " Farbenanalytische Untersuchungen" (Berlin, 1891 [Hirsch- wald], 137 pages) contains nine short essays by Ehrlich and three by his pupils. Considering the reputation of the writer they are at the present day rather disappointing reading, and contain little that is not better expressed elsewhere. 8. Weiss's " Haematologische Untersuchungen" (Wien, 1896 [Proc- kaska] 112 pages, 8vo) contains many valuable studies on various points. 9. Under a somewhat different heading come the sections on the ex- amination of the blood in v. Jaksch's "Clinical Diagnosis" (English translation, London, 1893, Griffen & Co.), a seventy-five-page article containing many inaccuracies; and Lenharz : "Microscopic und Chemie am Krankenbett" (Berlin, 1896, Springer) , a fifty-page article. 432 BIBLIOGRAPHY. Magazine Articles of Special Value. 1. On Concentration and Dilution of the Blood Oliver : Lancet, June 27, 1896. 2. On Leucocytosis Goldschneider and Jacob: Zeit. fur klin. Med., 1894, vol. 25. Krebs : Inaug. Dissert. , Berlin, 1893. Sadler : Forschr. d. Med., Supplement-Heft, 1892. Also Klein, in Volkmann's Sammlung klinischer Vortrage, December, 1893, and of course Rieder and Turk above referred to. 3. On Anaemia Dunin : Volkmann's Sammlung. klin. Vortrage, 1896, No. 135. Romberg : Berlin, klin. Woch. , June 28, 1897. 4. Parasitic Anaemia Schaumann : Zur Kenntniss der sog. Bothrio- cephalus Anamie, Berlin, 1892, 214 pages ; and Askanazy : Zeitschr. f. klin. Med. , 1895, p. 492. Brown : Journal of Experimental Medicine, May, 1898 (Trichinosis). 5. Leukaemia Fraenkel : Deutsche med. Wochenschrift, 1895, p. 639. Fraenkel: 15th Congress fiir inn. Medicin. Wiesbaden, 1897. Benda : Ibidem. Dock : Moscow Internat. Congress, 1897. 6. Pernicious Anaemia Discussion by Birch-Hirschfeld, Ehrlich, Troje, and others, at the XI. Congress f . inner. Med. (Leipzig, 1892) . 7. Pneumonia Billings : Bulletin of the Johns Hopkins Hospital, No- vember, 1894. Diphtheria Billings : New York Medical Record, April 25, 1896. Typhoid Thayer : Johns Hopkins Hospital Reports, vol. iv., No. 1. Engel: 15th Congress fur inn. Med., Wiesbaden, 1897. Exanthe- mata Felsenthal : Arch, f . Kinderheilk. , 1892, p. 78. Zappert : Zeitschr. f . klin. Med. , 1893, No. 23. Small-pox Pick Arch, f . Dermatol. und Syph., 1893, p. 63. Sepsis Roscher : Inaug. Dissert, Berlin, 1894. Cholera Biernacki : Deutsche med. Wochenschr., 1895, No. 48. Diabetes Bremer : Moscow Internat. Congress, 1897. 8. Syphilis (a) Reiss : Arch. f. Dermat. und Syph., 1895, Hf. 1 and 2. (6) Justus: Virchow's Arch., 1895. 9. Tuberculosis (a) Dane : Boston Medical and Surgical Journal, May 28, 1896. (6) Stein und Erbmann : Deutsche med. Wochenschrift, 1896, No. 56, p. 323. (c) Grawitz : Deutsche med. Wochenschr., 1893, No. 51. 10. Malignant Disease Taylor (International Medical Magazine, July, 1897). (a) Sadler: Loc. cit. (b) Reinbach : Langenbeck's Archiv, 1893, No. 46. (c) Strauer: Dissert., Greifswald, 1893. 11. Bacteriology Sittmann : Deutsches Arch. f. klin. Med., vol. 53. 12. Diseases of the Stomach (especially Cancer) Schneyer : Zeitschrif t f. klin. Med., 1895, p. 475. Osterspey : Inaug. Diss., Berlin, 1892. 13. Eosinophiles Zappert: Zeitschr. f. klin. Med., 1893, vol. 23. 14. Haemoconien Miiller : Wien. med. Presse, 1896, No. 36. INDEX. ABSCESS, 223 iodine reaction in, 223 of liver, 196, 292 of lung, 236 other forms of, 234 pericsecal, 223 perinephritic, 234 subphrenic, 235 Acetic acid, solution for counting white corpuscles, 18 Actinomycosis, 108, 239 Acute anaemia, 190 Acute delirium, 315 yellow atrophy of the liver, 289 Addison's disease, 126, 320 myelocytes in, 121 Adenitis, 175 tubercular, 267 Alkalinity, 48 in cholera, 212 in diabetes, 316 in fever, 183 in gout, 317 in osteomalacia, 321 in rheumatism, 207 Altitude, effect on blood, 79 Amyloid disease, 175 Anaemia, 82 causes of, 94 classification of, in infancy, 388 color of skin and mucous mem- branes in, 82 destruction of corpuscles in, 89 due to intestinal parasites, 383 in infancy, 387 infautum pseudoleukaemica,391 in gastric atrophy, 275 leucooytosis in, 388 myelooytes in, 121 nucleated red cells in, 89 pernicious, 133-151 post-febrile, 191 primary and secondary, 84 red corpuscles in, 85 stages in secondary, 85 splenic hyperplasia in, 387 white cells in (see Leucocytosis) Aneurism, 299 Ankylostoma duodenale, 383 28 Antipyretics, effect on leucocytes, 186 effect on red cells, 183, 327 Antipyrin, eosinophilia from, 118 leucocytosis from, 110 Aoyoma, 238 Appendicitis, 223 Arthritisj gonorrhoeal, 210 septic, 210 Asiatic cholera, 211 Asthma, bronchial, 309 cardio-renal, 310 BACTERIA in blood, 383 in blood of sepsis, 215 Bacteriological examination of the blood, 47 Barlow's disease, 325 Basedow's disease, 320. (See Graves' Disease. ) eosinophilia in, 116 lymphocytosis in, 116 myelocytes in, 121 Basophilic cells, 66, 168 granules, Neusser'sperinuclear, 399 Beri-beri, 241 Blood, bacteriological examination of, 47, 383 counting, 11 crises, 128 destruction, 324 examination of dried and stained, 61 examination of fresh, 5 examination, value of, 3 in infancy, 385 normal appearances of, 9, 50, 61 plates, 53, 59 "Blood-Dust" (Miiller's), 59 Bone diseases, eosinophilia in, 116 tuberculosis, 259, 120 Bothriocephalus latus, 383 Brain diseases, 312 Bronchitis, 189, 194, 307 acute, 307 chronic, 308 Brownian motion, 9, 363 Bubonic plague, 425 434 INDEX. Burns, haemoglobinaemia in, 327 myelocytes in, 120 CACHEXIA, cancerous, red cells in, 88, 332 Cancer, 332 anaemia in, 333 effect of metastases in, 344 lack of anaemia in, 334 leucocytes in, 338 effect of position of, 339 of breast, 340 of gullet, 345 of intestine, 347 of kidney, 348 of liver, 346 of omen turn, 348 of stomach, 341 of uterus, 349 percentage of haemoglobin in, 334 Carbonic acid poisoning, 80, 121, 327 Carcinoma (see Cancer). Castration, effects of, 116-118 Cell death, prevention of, 8 Cerebral disease (see Brain). Cerebral hemorrhage, 128, 251 tumors, 251, 313 Cerebro-spinal meningitis (epidem- ic), 249 Charcot-Leyden crystals, 169 Chicken-pox, 207 Childbirth, influence on red cells, 56 influence on white cells, 102 Chloroform-benzol, use of, in esti- mating specific gravity, 40 Chlorosis, 152 haemoglobin in, 152 infantile, 389 myelocytes in, 121 red cells in, 153 specific gravity in, 156 white cells in, 156, 158 without known blood changes, 159 Cholaemia, 290 Cholangitis, 285, 291 Cholera, Asiatic, 211 Chorea, 313 Cirrhotic liver, 286 changes in red cells in, 78, 89 hyper trophic forms of, 288 Coagulation, 49, 324 Cold, effects of, 58, 102 Colic, hepatic and renal, 229, 290, 305 intestinal, 229 Colitis, 282 Color index, 123 Color of blood in chlorosis, 152 of blood in leukaemia, 160 of blood in pernicious anaemia, 133 of skin in anaemia, 82 Coma, blood in, 251 Concentration of blood, 57, 76, 296 Convulsions, leucocvtes in, 314 Cornil's " Mark cells,"" 71 Counting, difficulties of, 15 of red cells, 11-18 of white cells, 18-22 Cover-glasses, cleaning of, 7, 43 method of holding, 7 necessity of cleanliness, 7 preparations, 43 Crenation, 50, 86 Cretinism, 319 Cyanosis, general, effect on blood, 74, 296 local, 74 Cysticercosis, 312 Cystitis, leucocytes in, 108, 194 myelocytes in, 121 DEBILITY, effect on red cells, 315 effect on white cells, 67, 68, 114 Degenerative changes in red cells, 53, 84 Dermatitis, 109 Diabetes, myelocytes in, 121 Diabetic coma, 316 Diarrhoea, concentration of the blood in, 76 chronic, 282 Differential counting, method of, 46 Digestion leucocytosis, 99 leucocytosis in gastric cancer, 343 leucocytosis in gastric catarrh, 279 leucocytosis in gastric ulcer, 276 Dilatation of the stomach, 280 Dilution of the blood, 12, 57, 78 Diphtheria, 197 effects of antitoxin, 198 leucocytosis in, 199 lymphocytosis in, 200 myelocytes in, 200 red corpuscles in, 197 Distomum haematobium, 382 Dried preparations cf blood, 43 preparations of blood, red cells in, 61 preparations of blood, white cells in, 62 residue of blood, estimation of, 275 Duodenal ulcer, 277, 327 INDEX. 435 Durham's blood counter, 22 Dysentery, 282 Dyspepsia, 278 EAR, puncturing the, 6 Eczema, 109 Effusions, serous, 76, 242 Emphysema, 309 Empyema, 244 Endocarditis, 293 ulcerative, 215, 293 Endoglobular changes in red cells, 86 Enteritis, acute, 279, 282 chronic, 282 Eosinophiles, appearance in fresh blood, 52 appearance in stained blood, 65 in malignant disease, 350, 356 in normal blood, 67 life history of, 68 Eosinophilia, 116-119, 156, 204, 239, 309, 350, 356 Eosinophilic myelocytes, 71, 167 Epidemic dropsy, 240 Epilepsy, coma after, 251 red cells in, 89 white cells in, 119, 314 Erysipelas, 89, 212 Etat cribriforme, 286 Exercise, effect on blood, 56 impaction, 230 Fatigue, effects on blood, 57 Felon, 108, 234 Fever, influence on the blood, 183 Fibrin network, 53, 54 network, general pathology of, 124 network, in cancer, 332 Filaria sanguinis hominis, 374 sanguinis hominis, examina- tion for, 375 Fixation of blood films by alcohol and ether, 43 of blood films by heat, 43, 44 Fleischl's haemometer, 33-37 Floating kidney, 229, 307 Furunculosis, 108 GALL-STONES, 290 Gas, illuminating:, poisoning: by, 80, 109, 121, 327 Gastric cancer, 341 dilatation, 280 ulcer, 275 Gastritis, 278 chronic, 279 corrosive, 281 Gastro- enteritis, 276-278 General paralysis of the insane, 121, 312 Genitals, influence on blood. 116, 117 Glanders, 238 Glands (see Adenitis) Gonorrhoea, 117, 236 Gout, 317 Gowers' solution, 12 Graves' disease, 320 Gravity, specific, method of esti- mating, 40, 125 Grippe, 113, 214 HJEMATOCRIT, accuracy of, 30 advantages of, 30 method of using, 30-33 Haematology, 3 Haemocytolysis, 326 Haemoglobin, estimation of, 33-40 general pathology of, 123 in chlorosis, 152 indirect estimation of, by spe- cific gravity, 41 in malignant disease, 334, 353 in pernicious anaemia, 134 Hsemoglobinaemia, 89, 218, 326 paroxysmal, 326 Haemoglobinometer, Fleischl's, 33 Oliver's, 38 Haemophilia, 6, 325 Hemorrhage, 106, 126 ' hemorrhage, cerebral, 312 Hemorrhagic diseases, 324 Heart, diseases of, 293 diseases, of, aortic, 298 diseases of, concentration of blood in, 296 diseases of, congenital, 298 diseases of, dilution of blood in, 297 diseases of, leucocytes in, 297 diseases of, mitral, 298 Herpes zoster, 109 Hodgkin's disease, 177 disease in infancy, 395 disease, myelocytes in, 181 disease, transition to leukaemia, 177 disease, white cells in, 181 Hydatid, 235, 289 Hydraemia, 82, 95 Hydronephrosis, 175 Hyperacidity and hypersecretion, 280 Hypochondriasis, 314 Hysteria, 117, 251, 314 ICTERUS of the new-born, 285 Illuminating gas, poisoning by, 330 436 INDEX. Infancy, anaemias of, 387 blood in, 385 eosinophilia in, 116, 117 haemoglobin in, 386 leukaemia in, 397 lymphocytosis in, 114 Influenza, 214 Insanity, anaemia in, 315 eosinophilia in acute, 119, 315 Intestine, cancer of, 348 diseases of, 281 obstruction of, 284 Isotonia, isotonic coefficient, 49 JAUNDICE, catarrhal, changes of red cells in, 89 catarrhal, increased resistance of red cells in, 284 catarrhal, volume of red cells in, 284 in pneumonia, 185 KIDNEY, cancer of, 348 diseases of, 299 diseases of, acute, 300 diseases of, chronic, 301 floating, 229, 307 floating, cyst of, 175 stone in the, 305 LACTATION, influence on red cells, 56 Lactic acid, in blood of rheuma- tism, 207 Lead encephalopathy, 251 poisoning, changes of red cells in, 89 Leprosy, 272 Leucocytosis, 96 absence of, 113 after cold baths, exercise, mas- sage, 102 after parturition, 102 atypical leucocytes in, 112 definition of, 96 diagnostic value of, 100 from therapeutic influences, 110 inflammatory, 106-109 in infancy, 101, 388 of digestion, 99, 276, 279, 343 of malignant disease, 338, 354 of pregnancy, 101 of the moribund, 105 of the new-born, 101, 385 pathological, 106-114 physiological, 98 post-hemorrhagic, 106 toxic, 109 Leucopenia, 113 in anaemia, 113 in infectious diseases, IIP Leucopenia in starvation, 113 Leukaemia, 160-176, 397 adult leucocytes in, 166 anaemia in, 162 Charcot-Leyden crystals in, 169 effects of intercurrent infections in, 176 eosinophiles in, 116, 167 haemoglobin in, 161 in infancy, 394 lymphatic, 169 lymphocytes in, 168-170 megaloblasts in, 162-170 myelocytes in, 165 Neusser's granules in, 169 ndrmoblasts in, 162, 170 red corpuscles in, 89, 161, 170 septicaemia in, 171 splenic-myelogencus, 160-167 white cells in, 162, 170 Lipsemia, 125, 315-316 Liver, abscess of, 292 acute yellow atrophy of, 109, 289 cancer, 346 cirrhosis, 286 cholangitis, 286, 291 diseases of the, 284 gumma, 293 hydatid of, 289 Lungs, diseases of, 307 Lymphaemia, 169 Lymphocytes, appearances of, 62 percentage among white cells, 67 Lymphocytosis, 114 in chlorosis, 114, 157 in Graves' disease, 114, 121, 320 in infancy, 114, 385 in leukaemia, 167-169 in malignant disease, 349, 356 in pernicious araemia, 114, 145 in pneumonia, 115, 188 in rickets, 322 in syphilis, 114, 270, 385 in thyroid feeding, 115 in typhoid fever, 195 MALARIA, 361 absence of leucocytes is in, 113 blood destruction in, 89 haemoglobinaemia in, 374 haemoglobin in, 372 myelocytes in, 122 parasite of, 362 parasite of, crescentte forme, 367 parasite of, flagellate forms, 367 parasite of, hyaline forms, 362 parasite of,pigmented forms, 363 parasite of, segmenting forms, 366 INDEX. 437 Malaria, preparation of stained spec- imens, 370 red cells in, 89, 372 white cells in, 373 Malignant disease (see Cancer and Sarcoma), 332 disease, differential diagnosis of, 359 disease, summary of blood- changes in, 358 Malta fever, 238 Marrow, eosinophiles in, 67 rnyelocytes in, 67 Measles, 205 leucocytes in, 113, 205 red cells in, 89, 205 Measuring corpuscles, 55 Megaloblasts, 91 in chlorosis, 155 in malignant disease, 339 , in parasitic anaemia, 384 in pernicious anaemia, 141 in secondary ansemia, 94 Melanaemia, 126, 320 Meningitis, 248 cerebro-spinal, 249, 251 tubercular, 265 Menstruation, effect on red cells, 56 eosinophiles in, 116 Microblasts, 94 Mitosis, 68 Molecular motion in corpuscles, 9, 50, 363 Mononuclear neutrophiles, 65, 72 Movable stage, use of, in differential counting, 46 Mumps, 205 Myelocytaemia, 161 Myelocytes, 69, 72, 120 in cancer, 351 in leukaemia, 165 in other conditions, 121 in pernicious anaemia, 146 in sarcoma, 357 Myocarditis, 295 Myxoeiema and cretinism, 276 NECROBIOSIS of red cells, 85 Nephritis, 300 acute, 300 chronic, 302 chronic diffuse, 301 chronic interstitial, 304 parenchymatous, 301 necrobiotic changes in, 87 red cells in, 88, 300 Nervous system, diseases of, 311 Neuralgia, 312 Neuritis, alcoholic, 311 Neuritis, acute multiple, 311 Neusser's granules, 258, 377 Neutrophilic granules, 64, 69 Newton's rings, 14 Normoblasts, 90, 94 in chlorosis, 155 in infantile blood, 386 in pernicious anaemia, 141 in secondary anaemia, 94 Nuclein, eosinophilia from, 119 leucocytosis from, 110 Nucleated red corpuscles, 89 red corpuscles, atypical forms of, 92 red corpuscles in anaemia of in- fants, 391 red corpuscles in chlorosis, 155 red corpuscles in malignant dis- ease, 338 red corpuscles in pernicious an- aemia, 144 red corpuscles in tuberculosis, 254 Nutrition, effects on blood, 57 OBESITY, 315 Obstruction, intestinal, 284 Omentum, cancer of, 348 Operations, surgical, contraindica- tions in the blood, 130 surgical, effect on leucocytes in cancer, 341 surgical, loss of blood in, 127 Osteomalacia, 321 eosinophiles in, 116 myelocytes, in, 120 Osteomyelitis, 117, 121, 234 tubercular, 259 Osteosarcoma, 355 Otitis media, 194, 233 Oval corpuscles, 87 PACHYMENINGITIS HJEMORRHAGICA, 313 Parasites of the blood, 361 intestinal, influence on the blood, 383 malarial 'see Malaria) Paresis, 313 Pellagra, eosinophilia in, 116 Pelvic abscess, 231 neuralgia, 229 Pemphigus, eosinophilia in, 116 Pericarditis, serous, 247 tubercular, 267 Perinuclear basophilia, 258, 399 Peritonitis, 228, 245 general septic, 245 granular, 246 pelvic, 231 438 INDEX. Peritonitis, tubercular, 264 Pernicious anaemia, 133 anaemia, changes of red cells in, 89, 134, 140 anaemia, diagnosis of, 147 anaemia, eosmophiles in, 146 anaemia, haemoglobin in, 138 anaemia in infancy, 395 anaemia, lymphocytes in, 145 anaemia, megaloblast.s in, 144 anaemia, myelocytes in, 146 anaemia, necrobiotic changes in, 88 anaemia, relapse in, 134, 136 anaemia, white cells in, 136 Perspiration, effect on blood, 57 Phenacetin poisoning, 327 Phlebitis, 194 Phosphorus poisoning, 80, 119, 290 Phthisis, 253 anaemia of, 253 eosinophiles in, 119, 258 fibroid, 25 myelocytes in, 120, 258 white corpuscles in, 255 Physiology of blood, 50-71 Pipette (see Thoma-Zeiss) cleaning of, 18 Pilocarpine, leuc cytosis after, 110 Plague, bubonic, 238 Plethora, 73-80 possibility of a true, 80 Pleurisy, 242 purulent, 244 serous, 242 tubercular, 243, 267 Pneumonia, 184 artificial production of leuco- cytosis in, 185 bacteriology of blood in, 184 coagulation in, 184 leucocytes in, 185 prognosis in, 185 red cells in, 185 Poikilocytosis, 86 Poisoning, alcoholic, 331 corrosive, 330 opium, 330 ptomain, 331 Polychromatophilic changes in pernicious anaemia, 141 Polycythaemia, 73-75, 79 Polymorphonuclear cells, 64-67, 166 Polymorphous blood in leukaemia, 169 Potassium, chlorate of, poisoning by, 327 Pregnancy, 56 extra-uterine, 230 Puerperal m&viia, 315 mania, myelocytes in, 120 sepsis, 218 Puncture of the ear in blood ex- amination, 5 Purgation, concentration of the blood in, 76, 281 Purpura, 324 changes of red cells in, 88 necrobiotic changes in, 88 Pus tube, 231 Pyaemia, 216 changes of red cells in, 88 Pyelo-nephritis, 305 Pyonephrcsis, 307 QUININE, effect on blood in fever, 183 poisoning, leucocytes of, 109 RATIO of red to white cells, estima- tion in fresh blood, 9 of red to white cells, estimation in stained specimei s, 61 Regeneration of blood after hemor- rhage, 127 of blood after operation for cancer, 335 Relapsing fever, 425 Resistance of blood to water, elec- tricity, etc., 48 Rheumatism, acute articular, 207 chronic articular, 210 muscular, 210 red cells in, 208 subacute, 210 white cells in, 209 Rickets, 322 anaemia in, 322, 389 leucocytosis in, 323 lymphocytosis in, 114, 323 myelocytes in, 122, 323 Rotheln, 113, 205 Rouleaux formation, 50 SALINE cathartics, influence of, on the blood, 281 solution, normal, 128 Sarcoma, 353 anaemia in, 353 eosinophilia in, 116, 356 leucocytes in, 354 myelocytts ir, 120, 356 Scarlet fever, 203 changes of red cells in, 89 eosinophilia in, 204 leucocytosis in, 203 necrobiotic changes in, 89 Scurvy, 324 INDEX. 439 Scurvy, lymphocytes is in, 114 Secondary anaemia, 82-95 causes of, 95 grades of, 94 Maragliano's degenerative changes in, 89 nucleated red corpuscles in, 90 Sepsis, 215 bacteria in blood of, 215 in leukaemia, 171 in wounds, 215 leucocytes in, 216 myelocytes in, 120 puerperal, 218 red cells in, 216 streptococcus, 218 Sero-diagnosis, 400 in bubonic plague, 425 in cholera, 422 in colon-bacillus infections, 423 in diphtheria, 422 in hog cholera, 424 in Malta fever, 424 in peripneumonia of cattle, 424 in pneumococcus infections, 422 in proteus infections, 424 in psittacosis, 424 in pyocyaneus infections, 423 in relapsing fever, 425 in scarlet fever, 426 in septicaemia, 426 in tetanus, 426 in thrush, 425 in typhoid, 415 in yellow fever, 425 Sero-prognosis, 427 Serum, examination of, 408 clump reaction, general descrip- tion, 401 clump-reaction technique, 402 dried blood, use of, 405 cultures, how prepared, 409 dilution and the time limit, 412 fluid blood, use of, 404 fluid serum, use of, 407 Skin diseases, eosinophilia in, 116, 117 Small-pox, 206 leucocytes in, 206 pneumonia in, 185 red cells in, 89, 206 Solids, estimation of, in blood, 49 Spinal cord, chronic diseases of, 313 Spirochaete of relapsing fever, 379 Spleen, hypertrophy of, in rickets and all anaemias of infancy, 385 tumors of, 119 Splsnectomy, effects on blood, 183 Staining, Biondi-Heidenhain form- ula, 44 Ehrlich-Biondi formula, 44 fluid, Ehrlich's latest, 44 mast-cells, 66 Neusser's perinuclear basophilic granules, Appendix, of red cells when degenerated, 89 the malarial parasite, 370 Starvation, effects of, on blood, 57, 76 Stomach, 274 cancer of, 341 channels of, influence on blood, 274 dilatation of, 280 inflammation of, 278 ulcer of, 275 Suprarenal extract, concentration of blood by, 77 Surgery, value of blood examina- tions in, 129, 221, 223, 335 Sweating, concentration of blood in, 76 Sympathetic nervous system, in- 'fluenceon blood of, 116, 118 Syphilis, 268 of lung, 310 anaemia in, 268 and 389 seq. cerebral, 313 changes of red cells in, 89 effects of mercury on the blood in, 269 eosinophilia in, 118, 271 haemoglobin in, 270 leucocytosis in, 270 lymphocytosis in, 114 myelocytes in, 120, 271 TETANUS, 241 Tetauy, 313 Thoma-Zeiss blood counter, accuracy of, 15, 26 blood counter, cleaning of, 17 blood counter, disadvantages, 26 blood counter, use of, 11-21 counting slide, 15 Tintometer (Oliver's), 26 Toisson's solution, 12 Tonsillitis, 213 Toxic leucocytosis, 109 Transitional leucocytes, 63 Trichinosis, 239 eosiuophilia in, 119 Tropical anaemia, 82 Tubercular pneumonia, 256 Tuberculin, 119, 256 Tuberculosis, 253 acute miliary, 26i 440 INDEX. Tuberculosis, changes in red cells in, 89, 253, 390 genito-urinary, 268 leucopenia in, 262 leucocytes in, 114, 255-264 of bones, 259 of serous membranes, 264 pulmonary, 255 Tumors (see Cancer and Sarcoma) . Typhoid, 191 anaemia in, 191 bacteriology, 191 changes of red cells in, 89 diagnosis of, 191 effects of complications, 194 effects of intestinal perforation in, 194 leucocyte-sis in, 193 leucopenia in, 113, 193 serum-diagnosis in, 415 Typhus, 237 necrobiotic changes in, 89 UR/EMIA, 303 Uric-acid diathesis, 317 Urine, examination of, compared with blood examination, 3 Uterus, cancer of, 349 VARIOLA, 180 Varicella, 180 Vasomotor influences, effect on blood, 57, 74 Volume of red cells estimated by hsematocrit, 26 Vomiting, concentration of blood in, 76 WARM-STAGE, use of, 9 White corpuscles, amoeboid and non- amoeboid forms, 52 corpuscles, description of, in fresh blood, 52 corpuscles, distinguishing them from red, 18 corpuscles, normal number of, 59 Whooping-cough, 206 Wounds, septic, 219 XANTHIN bases, effect c.', in blood, 120, 259 YELLOW fever, 236, 425 RETURN TO DESK FROM WHICH BORROWED ! IB< 1 This book is due on the last date stamped below, or on the date to which renewed. Renewed books are subject to immediate recall. -~~^tlSSSS~ p> JflWfc '58 frT'tTFi MAR 15 13/3 MAR 1 5 1973 /7