THE LIBRARY OF THE UNIVERSITY OF CALIFORNIA LOS ANGELES GIFT OF Dr. Emil Bogen :^ CLINICAL PATHOLOGY PROPERTY OF WAITER §. ItoUNT. M. fl. APR tso: OF THE BLOOD A TREATISE ON THE GENERAL PRINCIPLES AND SPECIAL APPLICATIONS OF HEMATOLOGY BY JAMES EWING, A.M., M.D. Professor of Pathology in Cornell University JIedical College, New York City. ILLUSTRATED WITH THIRTY ENGRAVINGS, AND FOURTEEN COLORED PLATES DRAWN BY THE AUTHOR LEA BROTHERS & CO. PHILADELPHIA AND NEW YORK Entered according to the Act of Congress in the year 1901, by LEA BROTHERS & CO. In the Office of the Librarian of Congress. All rights reserved. iiemedical Ubsarj 100 TO PROFESSOR T. MITCHELL PRUDDEN, M.D., LL.D., IN APPRECIATION OF HIS CONSTANT AID, FEEELY EENDERED TO THE AUTHOR DURING A TERM OF SERVICE IN THE LABORATORY OF THE NEW YORK COLLEGE OF PHYSICIANS AND SURGEONS, THIS WORK IS 'GRATEFULLY DEDICATED. f^f^'lH^ PREFACE. The rapid advances in the knowledge of the pathology of the blood and the multitudinous applications of this knowledge in clinical diag- nosis which the last decade has witnessed have brought forth several critical treatises on hematology, and still furnish, as the writer believes, abundant reason for the preparation of another such work, in English. Although the clinical bearings of the subject have been partly or fully accessible to English readers in some recent text-books, many later contributions to the pathology of the blood and blood-forming organs have had to be sought elsewhere, often in the original articles. This omission the present work endeavors to supply. Much of the theoretical discussion in the volume, abstracts of special articles, and reports of cases, have been set in fine print, so as not to encumber the main text, which has been constructed for the student and general reader. The wants of the laboratory worker and special student of hematology have been partly considered in the chapter on Technics and Chemistry, and in the limited references to pathological anatomy. There are probably always some deficiencies in a treatise closely re- lating to clinical medicine when that treatise emanates from a patho- logical laboratory. Yet a comparison of the various extant works on the so-called " clinical " pathology of the blood has convinced the writer that clinical pathology is pathology still, and that a wide experi- ence at the autopsy table and in the microscopical examination of dis- eased tissues furnishes an absolutely essential standpoint from which to view pathological changes in the blood. The present volume aims therefore to associate changes in the blood as closely as possible with lesions in the viscera, without which combination the former are very often unintelligible. Vi PREFACE. In the preparation of the work all available sources of information have been freely consulted, and the writer has profited especially by the labors of Ehrlich, Limbeck, Hayem, Lukjanow, Lowit, Grawitz, Stengel, Cabot, and many others. An endeavor has been made to discriminate between authorities, and in all details of important sub- jects the author has invariably consulted the original sources of in- formation. Works without bibliography have been of little value in this task, and the considerable number of references involved have therefore been included as a feature of the present volume. J. E. CONTENTS. INTRODUCTORY. ON THE INTERPRETATION OF ANALYSES OF THE BLOOD. PART I. GENERAL PHYSIOLOGY AND PATHOLOGY. CHAPTER I. TECHNICS. Qualitative tests for blood 23 Volume of red cells and plasma 26 Euumeration of blood cells 30 Hemoglobin, iron 38 Histological examination of blood 46 Specific gravity 49 Alkalescence 51 Osmotic tension 53 Glycogen, Fat 55 Bibliograpby 56 CHAPTER II. CHEMISTRY OF THE BLOOD. Red cells, Leucocytes 59 Serum 62 The Whole Blood .' 64 The Blood-ash 66 Urea, Uric Acid, Glucose, Glycogen 68 Acetonemia, Lipacidemia, Cholemia 71 Specific gravity 73 Osmotic tension 75 Alkalescence, Basic Capacity 77 Bibliography 79 CHAPTER III. MORPHOLOGY AND PHYSIOLOGY OF RED CELLS. Structure and staining reactions 82 Degenerative changes 85 Numbers, Polycythemia, Oligocythemia 89 Bibliography 103 viii CONTENTS. CHAPTER IV. THE LEUCOCYTES AND LEUCOCYTOSIS. Morphology 106 Numbers and proportions 110 Degenerative changes 112 Leucocy tosis, classification 115 Relation to phagocytosis and immunity 119 Clinical types of Leucocytosis 122 Experimental Leucocytosis 131 Bibliography 134 Eosinophilia 137 Lymphocytosis 148 Bibliography 148 CHAPTER V. DEVELOPMENT OF BLOOD CELLS. Erythrocytes ._ 149 Leucocytes 154 Blood plates 159 Bibliography 160 PART II. SPECIAL PATHOLOGY OF THE BLOOD. CHAPTER VI. CHLOROSIS. Etiologj' 168 Changes in the Blood 165 Regeneration of Blood 168 Varieties of Chlorosis, Pseudo- chlorosis 169 Bibliography 171 CHAPTER VII. PROGRESSIVE PERNICIOUS ANEMIA, Historical 172 Etiology ;^74 Pathological changes in Viscera 183 Pathogenesis i og Changes in the Blood Igy Resume ^qo Bibliography '-^gg CONTENTS. IX CHAPTER VIII. LEUKEMIA. Historical 198 Etiology 201 Pathological changes in Viscera 204 Pathogenesis 208 Changes in the Blood 210 Diagnosis of Leukemia 219 Bibliography 220 CHAPTER IX. PSEUDO-LEUKEMIA. Historical 223 Anatomical characters 224 Relation to Leukemia, Pernicious Anemia, Tuberculosis 226 Changes in the Blood 229 Bibliography 231 CHAPTER X. ANEMIA INFANTUM PSEUDO-LEUKEMICA, SPLENECTOMY. Historical 233 Etiology 233 Pathological changes in Viscera 234 Changes in the Blood .^ 234 Significance of v. Jaksch's anemia 235 Bibliography 237 Splenectomy. Effects of Splenectomy in Animals 238 Splenectomy in Man 239 Bibliography 241 PART III. THE ACUTE INFECTIOUS DISEASES. INTRODUCTORY SECTION. THE BLOOD IN FEVER. Chemical changes. Resistance of red cells 243 Febrile Hydremia 244 Action of Bacteria upon the Blood 245 Bibliography 246 X CONTENTS. CHAPTER XI. PNEUMONIA. DIPHTHERIA. Pneumonia. Gross changes, Red cells, Leucocytes 248 Bacteriology ^^'^ Diphtheria. Red cells, Leucocytes 255 Effects of Antitoxine 257 CHAPTER XII. EXANTHEMATA. Variola 259 Vaccinia 260 Varicella 260 Scarlatina 261 Measles 262 CHAPTER XIII. TYPHOID FEVER. Red cells. Leucocytes 264 Bacteriology 267 Bibliography, on Pneumonia, Diphtheria, Exanthems, and Typhoid Fever 268 CHAPTER XIV. WIDAl's TEST. The reaction, etc. , 271 Bibliography 278 CHAPTER XV. MISCELLANEOUS INFECTIOUS DISEASES. Septicemia, Pyemia, Osteomyelitis 280 Appendicitis, Abscess, Erysipelas 283 Acute Rheumatism, Tonsillitis, Whooping-cough 285 Inflammations of Serous Membranes 287 Gonorrhea, Yellow Fever 288 Typhus Fever, Influenza 289 Tetanus, Plague, Malta Fever 290 Actinomycosis, Glanders, Anthrax 291 Bibliography 291 k CONTENTS. XI CHAPTER XVI. SYPHILIS, TUBERCULOSIS, LEPROSY. Syphilis, Grades of Anemia 293 Effects of Mercury 294 Congenital Syphilis 297 Tuberculosis, Grades of Anemia 298 Tuberculosis of Meninges, Bones and Joints 302 Chemistry, Bacteriology 303 Leprosy 304 Bibliography 305 PART IV. CONSTITUTIONAL DISEASES. CHAPTER XVII. HEMORRHAGIC DISEASES AND DIATHESIS. Purpura Hemorrhagica 307 Hemophilia 310 Scurvy 311 Hemocytolysis 314 Methemoglobinemia 316 Paroxysmal Hemoglobinuria 317 Bibliography 320 CHAPTER XVIII. MISCELLANEOUS CONSTITUTIONAL DISEASES. Diabetes 322 Obesity 324 Addison' s Disease 325 Osteomalacia 327 Kachitis 328 Myxedema 330 Bibliography 331 CHAPTER XIX. NERVOUS AND MENTAL DISEASES. Mania, General Paresis, Epilepsy, etc 333 Beri-beri 335 Chorea, Graves' Disease 336 Bibliography 336 xii CONTENTS. PART V. GENERAL DISEASES OF VISCERA. CHAPTER XX. THE HEMOPOIETIC SYSTEM. Liver, general considerations, special diseases 337 Esophagus 342 Stomach, general considerations, special diseases 343 Intestines, absorption and depletion, special disease? 349 Bibliography 352 CHAPTER XXI. LUNGS, HEART, KIDNEYS. Diseases of Lungs, Asphyxia 354 Diseases of Heart, Malignant Endocarditis 357 Diseases of Kidneys, Uremia 361 Bibliography 366 CHAPTER XXII. MALIGNANT TUMORS. Carcinoma 368 Sarcoma 373 Bibliography 375 PAET VI. ANIMAL PARASITES. CHAPTER XXIII. MALARIA. Technics 077 Morphology of Parasites 381 Plurality of species in Estivo-autumnal group 392 Crescents, Flagellate Bodies 396 Development in Mosquito 399 Conjugation '^^^^'^^^^^^^''mO Occurrence of Parasites in the Blood .401 Malarial Anemia AQ'i Leucocytes in Malaria ^q^ Bibliography ^^'^'^^'!^*!^^^^!^^!!!^!"^"!!!^!409 CONTENTS. xiii CHAPTER XXIY. RELAPSING FEVER. Morphology of Spirillum 411 Changes iu the Blood 413 Bibliography 413 CHAPTER XXV. MISCELLANEOUS PARASITIC DISEASES. Trichina Spiralis 415 Distoma Hematobium 416 Ascaris Lumbricoides, Anguilulla Stercoralis 417 Anchylostoma Duodenale, Bothriocephalus Latus 418 Filariasis 421 Bibliography 423 LIST OF PLATES. PLATE. PAGE. I. Normal blood. Triacid stain 82 II. Normal blood. Eosin and Metbylene-blue 106 III. Mild Chlorosis 164 IV. Severe Chlorosis 169 V. Progressive pernicious anemia 188 VI. Secondary pernicious anemia 190 VII. Myelemia 211 VIII. Lymphemia 215 IX. Degenerating leucocytes in leukemia 213 X. Mast-cells. Glycogen in leucocytes 252 XI. Tertian malarial parasite 382 XII. Quartan parasite 388 XIII. Estivo-autumnal parasite 390 XIV. Conjugating cycle of tertian parasite 400 INTRODUCTORY. ON THE INTERPRETATION OF ANALYSES OF THE BLOOD. Plethora. — Although the older physicians regarded the existence of a true plethora as an important and well-established fact in pathology, this view, lacking proof, received a serious blow from the work of Lud wig's pupils, v. Lesser and Worm-Muller, and from the authorita- tive conclusion of Cohnheim. Lesser and Worm-Muller, attempting to produce artificial plethora by transfusion of blood in animals, found that when the red cells were thereby increased 28 percent the num- bers fell to normal on the following day, and when increased 58 per- cent, the blood became normal on the twenty-third day, while the total quantity of blood might be doubled without abnormal symptoms, owing to the rapid return to the normal quantity. Likewise Hamburger was unable to produce a permanent plethora in horses, by the injection of large quantities (7 1.) of 5-percent solution of sodium sulphate, finding that the isotonic power of the blood, thus increased, became normal in one-half to two hours, while nearly all traces of the salt disappeared from the blood in twenty-four hours. Instead of drawing from these data the conclusion that continuous plethora cannot be induced arti- ficially, the unwarranted claim was advanced that a true plethora does not exist (cf. v. Recklinghausen, Lukjanow). Against this conclusion stood the daily observation of clinicians and pathologists, at the bed- side and in the dead-house, that there are extreme variations in the quantity of blood in the vessels of different subjects, in divers states of health and disease. Oertel especially strongly maintained, without being able to ofPer proof, that the volume of blood might be reduced in endocarditis and with good therapeutic effect. Positive data on the question were gathered by Bergmann and Heissler, pupils of Bollinger, and the fact established that there is, in general, a direct ratio between the volume of blood and size of the heart, and the muscular develop- ment of the subject, and an indirect ratio with the subject's fat-de- posits. Though lacking full experimental proof it is now generally accepted that the quantity of blood in the body is variable, may be in- creased by hygienic measures, and is diminished in many unhygienic and diseased conditions. Anhydremia, /. e., a reduction in the volume of blood with concen- tration of solids, must be admitted to result from loss of body fluids, as by sweating and diarrhea, or by diminished ingestion of water. Czerny exposed cats in a warm, dry chamber for 36 hours, finding that they lost 45 percent of their weight, the volume of blood was 2 18 INTRODUCTORY. greatly reduced, its viscosity increased, and the red cells rose, in one case, to ten millions per cubic millimeter. Grawitz found that profuse sweating is followed in the majority of cases by a concentration of the blood, in one instance from s. g. 1040 to 1051. A few subjects showed a contrary reaction and dimi- nution in gravity (1060-1057.5), which Grawitz refers to nervous in- fluences. Limbeck examined the blood of a case of cirrhosis of the liver with extreme ascites, before and after tapping the abdomen. On the day after the removal of 18 1. of fluid the red cells had risen from 3 280,000 to 5,160,000. Stintzing and Gumprecht have made similar observations, before and after the removal of large serous exudates. Grawitz found the specific gravity of the blood at first decreased by the intravenous injection of concentrated salt solutions (absorption of water), while the administration of the salt by mouth concentrated the blood. It therefore follows that the ingestion of water and the loss of body fluids always produces a more or less transitory eifect on the volume of the blood. There can be no doubt also that the prolonged sweats of phthisis and the severe diarrhea of typhoid fever, dysentery, and cholera, lead to a more or less continuous reduction in the volume of the blood and a concentration of blood cells. A very striking illustration of this principle was frequently encountered by the writer among the soldiers at Camp Wickoff (1898). When these pa- tients, suffering from prolonged malaria with severe anemia were attacked by typhoid fever or acute dysentery, the ordinary watery character of the expressed blood drop disappeared, and the blood soon became thick^and deep red. Herz claims to have recognized an " acute swelling " of the red cells in febrile conditions, especially in typhoid fever, and v. Limbeck, finding a considerable average increase in the volume of the red cells in eight febrile cases, believes that relative oligoplasmia with increased volume of red cells is of frequent occurrence in high fevers. The effect of nervous influences in altering the quality of the blood in the Avhole or a part of the circulation has been demonstrated by numerous studies. Cohnstein and Zuntz found that section of the cord above the origin of the splanchnic nerves is followed by general dilatation of blood- vessels, and reduction in the proportion of red cells. Grawitz and Knopfelmacher found, in general, that vaso-motor paral- ysis is followed by local and general increase in the volume of blood with diminished specific gravity and proportion of red cells, while the opposite effects follow vaso-motor constriction of vessels. The nervous stimulus of cold baths has been found by Leichten- stern, Wick, Knopfelmacher, Winternitz, Thayer, and Grawitz, to be followed by contraction of vessels and increase in the proportion of red and white cells, while the hot pack, and the inhalation of amyl nitrite, have an opposite effect. INTERPRETATIOy OF ANALYSES OF THE BLOOD. 19 To a similar origin Diust be attributed the increased gravity of the blood observed by Grawitz after the injection of tuberculin, and cul- tures of the bacillus of cholera and that of diphtheria, and the oppo- site effect produced by streptococcus and staphylococcus pyogenes, and />. anthracis. The influence of psychical emotions on the character of the blood in different regions of the body has been emphasized by Lloyd Jones, and is seen especially in the study of the blood in neurotic women. Jacobi, examining the blood of an hysterical Moman, found on No- vember 12th, 3,892,000 red cells; on December 11th, 8,084,000 red, 102,200 white cells; on December 16th, 3,393,000 red, 22,000 white. Grawitz refers to the remarkable increase seen in the specific gravity and proportion of cells in the blood of rabbits, in experiments conducted without anesthesia. The remarkable variations in the blood taken from the ears of rabbits, depending on the temperature and blood content of the member, are a matter of common observation. In the mechanism by which these changes are brought about, there appear to be many factors. Heidenhain, LoAvit, Gartner and Romer, find that many substances injected into the blood cause an increased flow of lymph. (Extract of cancerous tumors, pepton, tuberculin, toxines of B. pyocyaneus and pneumococcus, hemialbumose, nuclein, uric acid, etc.) Heidenhain and Hamburger believe that some sub- stances excite a secretory action of the capillary endothelium, whereby the fluids of the blood are diminished. Grawitz explains the concentration of the blood under the influence of cold by the escape of fluids into the tissues, but the experiment of Cohnstein and Zuntz offers another explanation, since they found un- der the microscope that capillaries might be so reduced in calibre by irritation of vaso-constrictor nerves that no red cells, but only plasma, could pass in them. It then appears that many red cells may be caught in the contracted capillaries while the plasma passes on into the veins. Grawitz has shown by many comparative tests that the capillary blood is, in all ordinary conditions, richer in cells than that of the veins. It does not appear to have been proven that the vascular dilatation produced by heat is followed, in this particular condition, by the return of tissue fluids into the capillaries. It appears quite as likely that an irregular distribution of cells and plasma is to a larger extent responsible for such local variations in the composition of the blood (cf. Winternitz). The law that increased blood pressure leads to transudation, authoritatively stated by Ludwig and Landois, appears to have a less important application here than in more general and more prolonged processes. It is not likely that a cold bath can be followed by much exudation of serum into the tissues. Massage and electricity have been shown, by Mitchell and Cheron, to cause an immediate increase in the number of red cells in the blood of a part, an effect which may be explained according to the above data. 20 INTRODUCTORY. The importance of the osmotic relations of the blood in control- ling the volume of red cells and plasma, is indicated from the discus- sion in Chapter II. In hydremia and anhydremia Hamburger has shown that the isotonic relations of the blood are maintained by a rapid interchange of salts and albumens between the cells and plasma, with frequent minor changes in their relative volume. From Lim- beck's analyses of venous and arterial blood it appears, also, that the imbibition of CO2 causes swelling of red cells with absorption of water and salts and with relative diminution in the volume of plasma. The foreo-oing considerations are dwelt upon not only because of their theoretical interest, but because they deal with fundamental facts without the knowledge of which it is impossible to properly perform or sensibly interpret the results of an examination of the blood. To summarize the discussion, it has been shown that there are wide variations in the quantity and quality of the blood referable to di- verse conditions other than disease. 1. There are considerable physiological variations in the volume and composition of the blood, according to the constitution of the in- dividual (plethora), and the degree of muscular development. Here may be classed the variations between the sexes and between different periods of life. Such variations are permanent but usually not of extreme grade. 2. There is a great variety of physiological conditions producing marked but transitory changes in the blood, such as active digestion, muscular exertion, the ingestion of fluids, profuse perspiration, tem- porary cyanosis, etc. 3. The nervous system has a very striking temporary influence on the quality of the blood in local or general areas, acting through the cerebral (psychical) or medullary centers, or through local vaso-motor nerves. 4. Various local influences may greatly change the quality of the blood specimen, as seen in the local and transient effects of cold, heat, massage and electricity. 5. Many therapeutic procedures may temporarily alter the blood, as the aspiration of fluids, administration of diaphoretics, purges, vaso-dilators (amyl nitrite), vaso-constrictors, etc. 6. Various pathological conditions may partly or completely ob- scure the real status of the blood, as the sweats of phthisis ; the diar- rhea of typhoid fever, dysentery, and cholera; general cyanosis or local stasis ; the increased arterial tension of uremia ; the polyuria of diabetes and nephritis ; antemortem cardiac failure, etc. Having regard to the possible action of any of the above influences, one may avoid many of the local disturbances by observing special care in the manner of expressing the blood specimen. The blood should be expressed by very slight pressure, exerted at a distance, from a liberal puncture of the finger-tip or ear-lobe. The INTERPRETATION OF ANALYSES OF THE BLOOD. 21 circulation in the part should be as nearly normal as possible and should be uniform. A cold bloodless tissue is not suitable for furnish- ing a blood specimen, and if artificial means are taken to correct the condition a sufficient period must elapse to allow the accelerated circu- lation to subside. Blood should be taken not less than four hours after a hearty meal, and when comparative tests are made, the specimens should be taken at the same hour each day. The examination having been performed, its results are to be inter- j)j'eted only in the light of the fullest possible clinical information. There can be no doubt, as GraAvitz has pointed out, that the numerous con- tradictory results of hematological studies are largely referable to hasty conclusions drawn from figures without regard to the condition of the patient or the stage of the disease under consideration. Like- wise, hematological diagnosis has fallen into much discredit from the tendency to offer opinions from the isolated findings of the blood-test. PART I. GENERAL PHYSIOLOGY AND PATHOLOGY. CHAPTER L TECHNICS. QUALITATIVE TESTS FOR BLOOD. By far the most delicate test for blood is the demonstration of red blood cells under the microscope, but as these cells are not always preserved in demonstrable form, one must frequently resort to various tests for hemoglobin and its derivatives. The Guiacum Test. To a watery solution of the suspected substances is added a few drops of tincture of guiac, producing a milky precipitate. A few drops of hydrogen peroxide added will, in the presence Fig. 1. of blood pigment, produce a distinct blue color. The tincture of guiac should be freshly made, and diluted to the color of pale sherry wine. The guiac must be added before the peroxide, and the blue color must develop immediately. Many substances turn tinc- ture of guiac blue, without the presence of peroxide. This test is very delicate, demonstrating one part of fresh blood in several thou- sand of water. It is most effective with fresh speci- mens ; but after two or three years some specimens fail to react satisfactorily. Brandenburg finds that leucocytes of myelogenous origin, but not the lymphocytes give this same reaction. Hemin crystals. (Riedek. ) 24 GENERAL PHYSIOLOGY AND PATHOLOGY. The Hemin Test (Teichmann's). A drop of blood or portion of suspected detritus is spread out on a glass slide mixed with one drop of salt solution and dried at a low temperature. A cover-glass is then laid over the specimen, filled beneath with glacial acetic acid, and the specimen evaporated with higher heat, but without boiling. When the fluids have entirely evaporated, the specimen may be mounted in distilled water and ex- amined microscopically for the characteristic crystals of hemin (Teichmann's crystals). The hemin test, if successful, is absolutely reliable, but often fails, in unskilled hands, from (1) Alteration of albumins by excessive heat, preventing the formation of crystals ; (2) Excess of salt solution ; (3) Violent boiling which drives off free HCl, thus preventing the formation of hemin crystals (hydrochloride of hematin). SPECTROSCOPIC EXAMINATION OF BLOOD. In all cases where a sufficient quantity of dissolved blood can be obtained for examination the spectroscopic test is the best means of determining not only the presence of blood pigment but also its par- ticular form. Of fresh blood a 1 -percent solution yields very distinct absorption bands. Recently clotted blood dissolves readily in water. Old clots may usually be dissolved by maceration in acetic acid after which the spectrum of acid hematin is obtained. Clots that have been exposed to heat must be macerated in ammonia, when the spectrum of reduced or of alkaline hematin Fig. 2. will result. For ordinary clinical Avork Browning's spectro- scope is very satisfactory, but absorption bands are more accurately located in larger instruments. The small instrument should be supported in a con- venient holder in strong daylight or gaslight. By means of a collar the width of the aperture may be varied according to the strength of the light and opacity of the fluid. By means of the sliding tube Fraunhofer's lines are brought mto accurate focus. The fluids should be examined in small glass vials with flat sides. When dealing with fresh blood and unaltered Hb the spectrum is that of oxyhemogJoh'in, which shows two absorption bands between D and E, one rather thin and sharp near the orange, and the other T.n 'Wiiing's si)ectroscoi)e. SPECTROSCOPIC EXAMINATION OF BLOOD. 25 broader near the green. The indigo and most of the blue is absorbed. In strong solution the two bands of ox_yhemoglobin may be united. If to the solution of oxyhemoglobin is added a little reducing sub- stance, such as ammonium sulphide, the color of the fluid becomes darker and the spectrum changes, becoming that of reduced hemoglobin, giving one broad absorption band between D and E. The transformation of the spectrum of oxyhemoglobin to that of re- duced Hb, by reducing agents, is the positive indication of the presence of blood. Cochineal and ammouiated carmine give spectra very simi- lar to that of oxyhemoglobin, but on the addition of boracic acid the spectra of these substances are displaced into the blue, while that of blood remains unaffected. The spectra of various other vegetable dyes simulate that of blood but these are bleached by sodium bisulphite. Hematin is produced by the addition of acids or strong alkalies to Fig. 3. Red Orange YeUoiv B C D Green Eh Blue Indigo G !! Oxtjlicemoglobin Reduced H(jemoglobin Meduced Hcematin Methcfmnglobiil Acid Co-Haemoglobin Hcematin Alkaline I Spectra of blood pigments. (After Limbeck.) reduced Hb. In acid solution its spectrum is similar to that of acid methemoglobin. In alkaline solution it gives a single rather broad band at jD. Clinically, the most important alteration of oxyhemoglobin is that into methemoglobin, sometimes detected by the chocolate color of the blood. In acid or neutral solution it gives four absorj)tion bands, one quite sharp, between Cand J); a second, faint, in the yellow, immedi- ately to the right of D ; a third, broad, rather distinct, between the yellow and green, and just to the left of E; a fourth, broad, to the left of F, sometimes merged with a complete absorption of the blue end of the spectrum. The demonstration of carbonic-oxide-Hb is of great clinical impor- tance in cases of gas-poisoning. The blood drop has a rosy red tinge, seen alike in both venous and arterial blood. In one-half percent 26 GENERAL PHYSIOLOGY AND PATHOLOGY. dilution a spectrum is obtained which resembles that of oxyhemoglobin, but the bands are broader, and the D band is displaced slightly to the right. On the addition of ammonium sulphide the spectrum of oxy- hemoglobin is replaced by that of reduced Hb, while Co-Hb remains unaltered. In applying this test clinically, since considerable oxyhemo- globin remains in the blood in most cases, the results of spectral analysis are not always clear, and corroborative tests are required. 1. \yarm the specimen with equal parts of 10-percent NaOH ; normal blood becomes dark brownish green, the other becomes cloudy, then clear red, and red flakes gather on the surface. (Hoppe-Seyler.) 2. To a 2-percent solution of blood add a few drops of orange-col- ored ammonium sulphide containing an excess of sulphur, and faintly acidify with a few drops of dilute acetic acid, carefully shaking. Car- bonic-oxide blood then shows a beautiful rose red color, with a floc- calent precipitate, while normal blood becomes greenish or reddish gray. The test may be performed in a porcelain dish, adding a drop of blood to the mixed reagents. (Katayama.) Kunkel and Welzel employ a solution of zinc chloride, or very dilute solution of platinum chloride, which color carbonic-oxide blood bright red, normal blood, black. Rubner recommends that the suspected blood be diluted 4-5 times with acetate of lead, when normal blood becomes chocolate, carbonic- oxide blood red. Watery neutral solutions of Co-Hb, boiled, yield a clear red coagu- lum, while oxyhemoglobin becomes grayish brown. (Hoppe-Seyler.) Estimation of Total Quantity of Blood. The limits of error in the methods of estimating the total quantity of blood, suggested by Valentin, Vierordt, Buntzen, and Thibault, probably exceed the physiological and pathological variations in the bulk of this tissue, and are therefore not available for clinical purposes. ESTIMATION OF THE VOLUME OF RED CELLS AND PLASMA. THE HEMATOCRIT. The idea that the centrifuge might give valuable clinical information concerning the volume of red cells was original with Blix, while his instrument and method have been improved principally by Hedin, Gartner, and Daland. Although hand-centrifuges have been used and recommended, the best results are obtained only with the electric cen- trifuge, which is at present in the market at a reasonable figure. The improved electric centrifuge consists of an iron-clad motor car- rymg a steel shaft and horizontal armature for urine tubes, which may be replaced by the hematocrit. A " speed indicator " may also be attached which strikes a bell with every 100 revolutions. The instrument is practically noiseless even with a very high speed. A rheostat is used to control the current and speed. A speed of 8,000- THE HEMATOCRIT. 27 Fig. 4. 10,000 revolutions may be obtained by a small battery or from the street current. The hand-centrifuge, Fig. 5, may be employed when it is not convenient to use electricity. The hematocrit-attachment consists of two capillary tubes, graduated in 100 degrees, which are held in the armature by springs. (Fig. 5.) Procedure. — With walking patients fresh blood may be used. The capillary tubes may be filled automatically by holding the tube horizontally, or with one end slightly depressed, and touching it to the rather large drop of blood required. The tube is then immedi- ately inserted in the armature, as in Fig. 5, the opposite tube hav- ing previously been filled with water, and the revolutions are begun before the blood coagulates. When the patient is at a distance, the blood must be diluted, prefer- ably in 2.5-percent solution potassium bichromate, as recommended by Daland. The blood should be diluted with equal parts of this solution, which may be accom- plished in the red-cell or white- cell pipette of the hemocy tometer. With the red-cell pipette the cap- illary tube is filled with blood to the mark 1, then a small air bub- ble is drawn in, followed by an- other tube-length of blood. Three or four tube-lengths of blood should be secured in this way and immediately mixed with an equal number of tube-lengths of diluent. With the white-cell pipette a single measure of blood and diluent is sufficient. The blood and diluent should then be mixed by gentle shaking, taking care not to enclose air bubbles. When diluted blood is used both tubes of the hematocrit should be filled with blood, which may be done by allowing the drops to flow in from the point of the pipette. The tubes are now revolved at a speed of eight to ten thou- sand revolutions per minute, for three minutes, after which the volume of the red cells has been found unalterable. The question now arises how many red cells are contained in one degree of the scale. Daland, working extensively with diluted blood, places the number at 99,390, practically 100,000. Inil)r<)vetl olectric hematocrit, with fender, rheo- stat, and speed-indicator. The hematocrit attach- ment replaces the urine tubes seen in the revolving armature. 28 GENERAL PHYSIOLOGY AND PATHOLOGY. Cabot, in a series of 40 comparative tests, using undiluted blood in the hematocrit, found variations between 105,000 and 123,000, with an average of 112,000. Further observations on this point are required, but at present the above figures should be used for diluted and fresh blood respectively, multiplying the result by two Avhen diluted blood has been used. Fig. 5. Daland's hematocrit. Limitations of the Hematoceit. — Since the centrifuge does not necessarily require the use of diluting fluids, a serious cause of error in the method may be removed by the use of fresh blood. It must be admitted, however, that the original volume of the red cells cannot always be exactly determined by this method, as in pathological THE HEMATOCRIT. 29 conditions the compressibility of the corpuscles is not always uniform. Moreover, in altered conditions of the plasma, it is uncertain how much change can be wrought in the natural volume of red cells by violent centrifugal force. Some fragile cells are probably always de- stroyed during the centrifugal process. When diluting fluids are used, the error is doubtless increased by alterations in the density and composition of the plasma and in the volume of the red cells caused by the action of the fluid. Neverthe- less Daland's claim must be admitted that the volume of the red cells, except in leukemia, is determined by this method with accuracy suf- ficient for clinical purposes. The value of such information is, of course, quite evident. The further claim that the hematocrit may give more accurate esti- mates of the number of red cells than does the hematocytometer has not been confirmed. The volume of the red cells differs so markedly in both the chlorotic and the pernicious anemias that one cannot seri- ously consider the project of replacing the hematocytometer by the hematocrit. Only in the moderate secondary anemias, with little change in the size and Hb-content of the cells, can the volume of the red corpuscles yield reliable indications of their number. In cases of leukemia and of extreme leucocytosis so many leucocytes are entangled with the red cells that even the volume of the red cells is not accu- rately told, much less their number. Each of these instruments has its proper field to which it should be restricted, and as the hematocrit is not over-exact in its immediate object, it is unscientific to introduce a second source of error, as is done in attempting to compute the num- ber of red cells from their volume. It may be added that the value of the hematocrit in estimating the character and severity of an anemia has not yet been as fully recog- nized by clinicians as it deserves, possibly because more attention has been paid to the number of red cells than to their functional capacity. The reliability of the centrifuge in determining even the volume of the red cells has been denied principally by the brothers Bleibtreu, and by Bleibtreu and Wendelstadt. These observers devised another method of determining the volume of the red cells, which they claim gives more trustworthy results than are obtained by the hematocrit. They employed .6-percent salt solution to prevent coagulation and allowed the blood to settle slowly. The nitrogen-content of the super- natant plasma was then determined by Kjeldahl's method, and from tables which these observers constructed the volume of the plasma and hence that of the red cells could be determined, from the quantity of N obtained. While the results obtained with the hematocrit by several observers indicate that the normal volume of red cells varies between 40-66 percent, Bleibtreu's method gave normal variations in cadaveric blood between 25.15 and 55.8 percent (Bleibtreu, Pfeiffer), v. Lim- beck obtained very low volumes with Bleibtreu's method (24-28 per- cent) which he refers to the use of highly oxidized blood, in which he believes the red cells are reduced in volume. The lengthy discus- 30 GENERAL PHYSIOLOGY AND PATHOLOGY. sion which has prevailed regarding the above points, indicates that the vokimc of the red cells is subject to a considerable variety of changes, the origin and significance of which are little understood. It has been shown that in order to prevent N from leaving the red cells during sedimentation, the exact isotonic tension of the plasma must be determined in each instance, and a corresponding solution of salt used. The isotonic tension of plasma is rarely so low as .G-percent NaCl. Moreover, supposing that the red cells remain intact during sedi- mentation, the pathological variations in the N-content of the plasma, depending on several variable nitrogenous bodies, are too frequent and marked to permit of any fixed formula to give the volume of the serum from its content of nitrogen. (Bleibtreu's method has been sharply criticised by Hamburger, Eyckman, Hedin, Biernacki and others.) The following table of results obtained by Biernacki well illustrates the unreliability of comparisons between results obtained by different procedures and the general inaccuracy of all indirect methods of esti- mating the number of red cells or percentage of Hb, from the volume of the cells and the specific gravity : Case. Content of water. Red Cells. Number. Volume f. Hb. Normal Normal Cancer Esophagus Kheumatism Chlorosis Phthisis Nepliritis Tabes... Chlorosis Phthisis Ulcer of Stomach Chlorosis Nephritis 77.18% 77.50 79.58 79.02 80.99 82.37 82.73 83.09 83.04 84.59 85.43 89.36 89.46 037 mil. 487 175 902 958 672 800 512 250 975 825 456 184 56.3 53.6 52.7 49.1 50.0 40.9 40.0 47.6 35.4 30.9 37.1 20.0 13.6 105% 100 80 85 70 60 70 65 55 50 45 25 20 Grawitz determines the volume of red cells in blood drawn in con- siderable quantities by venesection. The specific gravities of the whole blood (Dj), of the centrifuged serum {D^, and of the sedimented red cells {D^, are first determined, from which the volume percentage of red cells (;r) may be computed by the following formula : x= 100 ( A -A) ESTIMATION OF THE NUMBER OF BLOOD CELLS. The Hematocytometer. — The instrument now in use for counting blood cells is that of Thoma, who combined and improved several features of instruments previously devised by Hayem, Gowers, and Malassez. This apparatus consists of a mixing pipette and a counting chamber. ESTIMATION OF THE NUMBER OF BLOOD CELLS. 31 (a) The pipette is a capillary tube graduated in ten equal divisions, surmounted by a bulb of exactly 100 times the capacity of the tube, and to which is attached a rubber tube and mouthpiece. (See sketch.) AVhen the tube is filled with blood up to the mark 1, and this is mixed with a diluting fluid sucked up to the mark 101, a specimen of blood is obtained in the dilution of 1 to 100. By filling only one-half the tube with blood, up to the mark 0.5, the subsequent dilution is in the proportion of 1 to 200. The bulb contains a glass ball to facilitate the mixing of the blood. {b) The counting-chamber is constructed so as to secure a layer of diluted blood J^ millimeter in depth over a certain square area. On a thick glass slide is cemented a thinner glass plate, the central portion of which is cut out. In this central area is cemented a circular glass shelf the surface of which is exactly J^ millimeter lower than Fig. 6. Tlif Tlioma hematocytoineter. the surface of the glass plate. When a drop of diluted blood is placed on the shelf and covered with a cover-glass, a layer of fluid is secured, which is exactly J^ millimeter deep. Between the edge of the shelf and the surrounding plate is a moat into which the blood may run, but if the fluid should run over the moat and beneath the cover- glass, the latter will be elevated and the resulting layer of fluid will be more than ^L millimeter deep. The shelf is accurately ruled as shown in Fig. 8. The entire ruled area is nine square millimeters, but only the cen- tral square millimeter is used in counting red cells, the others being required in counting leucocytes. It will be seen that this central square millimeter is subdivided into 400 small squares (1 6 blocks of 25 each), so that each small square is -j^q- sq. mm. Beginning at the lower left-hand corner of this area, it will be seen that every fifth square, above and to the right, is subdivided by an extra line, which 32 GENERAL PHYSIOLOGY AND PATHOLOGY. | is added merely to assist in counting the squares. The outlying square millimeters are variously ruled. The above description applies only to the so-called " Zappert " chamber Avhich should always be secured, preferably of Zeiss' manu- facture. The older chambers* cannot well be used for counting leuco- cytes. Diluting Fluids. — Of the various diluting fluids, Toisson's Mixtuee is to be recommended : Sodium sulphate 8 grrn. Sodium chloride 1 Glycerin pur 30 Aquadest 160 " Methyl violet 025 " This fluid keeps well, stains the leucocytes, and is of high specific gravity so that the red cells settle from it slowly. When counting leucocytes only, one may use with advantage a .6- percent solution of sodium chloride tinged with gentian violet (about 1 drop of saturated alcoholic solution gentian violet to 50 cc. of salt solution). This fluid, while readily prepared, does not keep well, and the red corpuscles settle from it so rapidly that it ought not to be used in counting these cells. It permits, however, of the identification of eosinophile cells and of certain degenerative changes in leucocytes. A reliable fluid for diluting and permanently preserving blood is found in Hayem's Mixture : Hvdrarg. bichlor 0.5 grm. Sod. sulphat 5.0 " Sod. chlor 1.0 " Aq. dest 200.0 " Directions for Using the Hematocytometer. (a) Filling the Pi- pette. — The finger tip of the patient is cleansed with soap and water, dried with alcohol, and freely punctured with a needle or a specially prepared acne-lancet. Using very gentle pressure only, a compact drop of blood is then expressed and the capillary tube is filled to the mark 1 or 0.5. In doing this the pipette must be held between the thumb and forefinger and the hand steadied against the hand of the patient. In well constructed pipettes the column of blood is easily controlled, and after filling, the end of the tube may be cleansed of adherent blood. The diluting fluid is then sucked up to the mark 101, taking care that no blood runs out of the tube when it is immersed in the fluid. The specimen is then thoroughly mixed by shaking. (6) FiLLixG the Couxting-chamber. — The counting-chamber and cover-glass are thoroughly dried and freed from particles of dust. One or two drops of diluted blood are first forced from the pi- pette and a third drop, the size of which can be learned only by expe- rience, is deposited on the central shelf. The cover-glass is then im- mediately adjusted, slipping one corner under the forefinger of the left ESTIMATION OF THE NUMBER OF BLOOD CELLS. 33 hand and controlling the opposite corner with the second finger of the right hand, and lowering the glass slowly so as not to inclnde air bubbles. Without raising the fingers, now quickly cover the other corners with the forefinger of the right and second finger of the left hand, and press the cover-glass firmly into position. If the application is successful and no dust particles have intervened, Newton's color rings will appear beneath the cover-glass. The formation and permanency of these rings may be facilitated by breathing very gently on the specimen before ap- plying the cover-glass. The specimen should now be held up to the light and examined closely to see that the red cells are evenly distributed. An uneven distribution is readily detected by the naked eye. After settling a few moments the specimen is ready for counting. The rapid and successful adjustment of the cover-glass is the most im- portant detail in the process of counting blood cells. The cover-glass must be rapidly adjusted because from the moment the drop is placed upon the shelf there is a rain of cells upon the ruled area out of a layer of fluid which is more than ^L mm. deep. The specimen must be discarded : If Newton's rings do not appear. If any air bubbles are inclosed. If the fluid runs underneath the cover-glass. If the shelf i& not well covered by fluid. If, on inspection, the cells are found unequally distributed. (c) Counting the Red Cells. — The specimen proving satisfactory the count may begin as soon as the cells have settled. Zeiss, D., Leitz, No. 7, Reichert's or Bausch and Lomb, ^, are the lenses best suited for this purpose, and a good mechanical stage is necessary for accurate work. Locate in the field the lower left-hand block of 25 small squares, begin at the lower and left square and passing to the right count all the cells lying in the first five squares. The fifth square will be found subdivided. In each square count all the cells lying on the lower and lefi side lines, leaving to be counted with the adjacent squares all the cells lying on the lines above and to the right. Proceed in this way till at least four blocks of 25 small squares, and at least 1,000 cells are enumerated. The more squares counted over the greater the accuracy, and when slight variations are to be demon- strated the entire square millimeter must be covered. If the cells now appear to be unevenly distributed the specimen should be discarded and another prepared after thoroughly shaking the pipette. (d) Computation. — Suppose that 1,280 cells are enumerated in 100 small squares, i. e., in ^ of the square millimeter. This number multiplied by 4 gives the number lying over one square millimeter. But the depth of the fluid is only J^ mm. so that we multiply again by 10 to get the number of cells in one cubic millimeter of fluid. Finally we must multiply by 100 because the blood is diluted in the propor- tion of 1 to 100. In short, after counting over 100 small squares the result is multi- 3 34 GENERAL PHYSIOLOGY AND PATHOLOGY. plied by 4.,000 to give the number of cells per cubic millimeter. (4 X 10 X 100 = 4,000.) If the capillary tube was originally filled to the mark 0.5, the di- lution is 1-200, and the multiplier 8,000. If 400 squares are counted over, the multiplier is 1,000. (e) Sources of Error in the Use of the Hematocytometer. 1. In Securing the Drop of Blood. — When much pressure is employed in expressing the drop of blood, tissue fluids are squeezed out with the blood and the number of red cells is reduced. Reinert found a reduc- tion of 722,000 from this cause, which is especially potent in cases of dropsy and of severe anemia. When the finger is cold, the circulation poor, or local stasis is pro- duced, as by a ligature, the red cells are increased in number. To avoid errors of this class, the circulation in the hand should be as ac- tive as possible, the finger warm, and the puncture liberal enough to permit the flow of blood with little pressure applied at some distance from the puncture. Unless these conditions can be secured it is hardly worth while to count the blood cells. 2. In Diluting the Blood and in Transferring it to the Counting-chamber there are numerous plainly evident sources of error, such as the inaccurate filling of the capillary tube, the entrance of air with the blood column, the failure to cleanse the tip of adherent blood, the escape of blood into the diluting fluid, the overfilling of the bulb with diluting fluid, the inadequate mixture of the blood, the fail- ure to discharge one or two drops before applying one to the shelf, the use of thin cover-glasses, and, above all, delay and inaccuracy in ad- justing the cover-glass. A little experience and constant care serve to eliminate all these difficulties. 3. In the Construction and Condition of the Apparatus. — The tendency to favor the Zeiss instruments is still probably well founded, although Leitz and Reichert are now making very excellent pipettes after Grawitz' model. Aside from inaccuracies in the gradu- ation of the pipette and construction and ruling of the counting-chamber, which are now reduced to a minimum, some pipettes are still on sale which are too short, their calibre is too large and is narrowed at the point so that the tubes cannot be cleaned, they require too much blood, and the short arm is so small that the mark 101 comes too close to the bulb. Accurate work cannot be performed with such instruments. The worker is at present advised to insist on having Grawitz' pipette made by Zeiss, Leitz, or Reichert. Many close observers find that their pipettes vary with the temper- ature. While accurate information on this point is not at hand it is just as well to avoid extremes of temperature in making the tests and in cleaning the instrument. It has been suggested that the polycy- themia of high altitudes is partly referable to variations in the hemo- cytometer due to changes in atmospheric pressure, but this suspicion has not been confirmed. The condition of the pipette is of prime importance. Absolute dry- OLIVERS HEMATOCYTOMETER. 35 Fig. ness of tube and bulb is essential. The collection of minute water drops in the tube and bulb is responsible for many of the shadow cor- ])nscles sometimes seen in the counting-chamber. Every few weeks a pipette should be cleaned out with concentrated nitric acid. (/) Cleaning the Apparatus. — After using the pipette, the rubber tube may be transferred to the long arm and the remaining fluid ex- pelled. The tube should then be cleaned thoroughly with water, then with alcohol and ether, or better, with pure ether. It must be thor- oughly dried before using again. The counting-chamber must be cleaned with water only, as alcohol and ether' dissolve the cement under the shelf and plate. ( g) The Limit of Error with the Hematocytometer. — Lyon, Thoma, (and Reinert), counting an average of 1,141 cells in 100 squares, with a dilution of 1-200, found an aver- age error of 1.82 percent in 24 preparations of the same specimen, and in another case, counting an average of 934 cells in 1 00 squares, 1-200 dilution, an average varia- tion of 2.71 percent, in 12 prep- arations of the same specimen. (Limbeck.) These results in the hands of experts using special care, indicate that a variation of 150,- 000 cells (3 percent) cannot be ac- cepted as of any significance. More accurate data are, however, seldom required by the clinician. Oliver's Hematocytometer. The Principle. — When a candle flame is viewed through a flat glass tube containing water a transverse line of bright light is seen which results from the blending of numerous images of the flame. The im- ages are produced by the minute longitudinal corrugations in the glass which reflect the light in various directions. When diluted blood is placed in the tube the fluid is sufficiently opaque to shutout the images until a certain dilution is reached when a bright streak of light becomes rather suddenly visible. Oliver believes that the appearance of this bright horizontal line is a very sensitive indicator of the proportion of red cells in the fluid, and by a long series of observations has devised an instrument for determining the number of red cells in blood according to this principle. Oliver's hematocytometer. A, measuring pipette; B, dropper; C, mixing tube graduated iu percentages ; I), mode of observation. 36 GENERAL PHYSIOLOGY AND PATHOLOGY. The apparatus consists of a measuring pipette (A) ; a dropper for Havem's fluid (B) ; a flat glass tube graduated in 120 degrees (C). The Procedure. — The capillary pipette is carefully filled with blood and washed into the tube by means of Hayem's fluid. A proper amount of fluid is then added to the diluted blood and the two are mixed by inverting the tube closed by the thumb, care being taken not to remove any diluted blood with the thumb. The test should be made in a dark room, the light being furnished by a Christmas candle placed about ten feet from the operator. When the blood is insuf- ficiently diluted the image of the candle is invisible when looking through the tube held horizontally, Fig. 7, D, but at a certain dilution the images begin to appear, and at the proper dilution a rather com- pact transverse line of light becomes visible. The bottom of the men- iscus is then read off on the graduated scale. Each degree of the scale represents 100,000 red cells, the mark 100 corresponding to 5 million cells, 80 to 4 million, 60 to 3 million, etc. There are both theoretical and practical objections to the use of Oliver's instrument. Theoretically, the method falls in the undesir- able class of indirect methods about which there are always a large series of unknown disturbing factors, which can only be eliminated by prolonged experience. Practically the difliculty of determining the exact dilution from the appearance of a compact line of light is very great. The method can- not be recommended until it has received much wider application than it has yet enjoyed. The Estimation of Leucocytes. The leucocytes may be counted by a method which requires a spe- cial mixing pipette, yielding a dilution of blood in the proportion of 1 to 10, and a diluting fluid (3-percent acetic acid, tinged with gen- tian violet) which dissolves the red cells, leaving only the stained leu- cocytes to be counted. The same chamber is used as for counting red cells, and the same procedure is followed. All the leucocytes in one square millimeter having been counted, the result is multiplied by 100, giving the number of leucocytes per cubic millimeter. The disadvantages early recognized in this method are the expense and inconvenieuce of an extra pipette, and a second diluting fluid, the time required in preparing a second specimen, the larger quantity of blood required, the difficulty sometimes encountered in distinguishing leucocytes from the detritus of red cells, and the impossibility of sep- arating and evenly distributing the cohesive leucocytes. This method has gradually been replaced to a large extent by the practice of counting leucocytes in the same specirnen prepared for counting the red cells. In 1892 the writer found that he secured more uniform results with the latter method, and has since found no inducement to return to the former. The Counting of Leucocytes in the Same Preparation with the Red. — This method requires the Zappert Chamber which was orig- THE ESTIMATION OF LEUCOCYTES. 37 inally devised by Elsholz for the estimation of eosinophile cells in fresh blood. Various modifications of the ruling in this chamber liave been employed, one of which, made by Leitz, at the writer's suggestion, is represented in Fig. 8. With this chamber using a Leitz lens. No. 7, it is possible to count over nine square millimeters which gives almost as many leucocytes as are counted in the other method. When the leucocytes are nor- mal or reduced in number, it is necessary to count all there are in Fig. 8. Palling of the Zappert chamber. the available 9 sq. mm., and if the number is very low it is advisable to prepare a second specimen in the chamber and count the white cells in 18 sq. mm. When the leucocytes are increased, 9 sq. mm., or in cases of leukemia, 6 sq. mm., will yield a number large enough to in- sure an accurate result. In order to make the leucocytes visible the Toisson's fluid or other solution should contain enough methyl-violet to stain these cells dis- tinctly. With a little practice the eye picks out the bluish highly refractive leucocytes very readily. 38 GENERAL PHYSIOLOGY AND PATHOLOGY. What has been said regarding the condition of the local circulation, and the effects of pressure in expressing the blood, is to be specially empha- sized when estimating the number of leucocytes in a specimen of blood. Computation. — Divide the number of leucocytes counted by the number of square millimeters traversed in the count and multiply by I 000. The result is the number of leucocytes per cubic millimeter of blood. If the original dilution is 1 to 200, which ought not to be employed except in cases of leukemia, the multiplier is 2,000. Thus if 54 leucocytes are counted in 9 sq. mm. (dilution 1-100) the number per cubic mm. is 6,000 (54 ^ 9 x 1,000). The Enumeration of Eosinophile Leucocytes, (a) In the same Preparation with the Red Cells. — When the blood is diluted, 1-100, with .6-percent salt solution tinged with gentian violet, the leuco- cytes retain their natural size and shape and eosinophile cells can be readily identified by their large, greenish, refractive granules. In cases of myelogenous leukemia this method is satisfactory, but when the eosins are present in their usual numbers (1—5 percent) one must count a larger number than can be found by this method. The usual expedient is to estimate their percentage from a dried specimen of blood, and then to calculate their number from the total number of all leucocytes counted by other methods. Thus, if the count shows 12,000 leucocytes per cubic mm. and the dried blood slide shows 2 per- cent of eosins their number will be 240 per cubic millimeter. This method is sufficiently accurate for clinical purposes. (6) By Means of Thoma's Special Pipette for the Enumeration of Leu- cocytes. — Klein, Mueller and Reider, and Elsholz have employed methods for the accurate estimate of eosins adapted to finer clinical work and to experimental research. They use the large pipette of Thoma which gives a dilution of 1-10. The capillary tube is filled with blood to the mark 1, and the bulb is half filled with the following solution : Watery eosin (2 percent), 7 parts ; glycerine, 45— ; Aq. dest., 55 — . After shaking 3-4 minutes the bulb is filled to the mark II with the following staining fluid : Aq. dest., 15 cc; gentian violet, cone. aq. sol. 5 drops, alcohol 1 drop. In specimens thus prepared both neutrophile and eosinophile leuco- cytes are readily distinguished, the eosins being particularly brilliant. The red cells are dissolved and the leucocytes concentrated so that a sufficient number of eosins may be counted. Zappert's extensive studies of eosinophile leucocytes were conducted with specimens diluted in the large pipette of Thoma, by the following solution : 1 -percent osmic acid sol., 5 cc, to which are added 5 drops of a filtered mixture — Aq. dest. 10 cc, glycerine, 10 cc, 1-percent watery eosin, 5 cc. THE ESTIMATION OF HEMOGLOBIN. 1. Gower's Hemoglobinometer. — This instrument has always been largely employed on account of its cheapness and simplicity, and the THE ESTIMATION OF LEUCOCYTES. 39 Fig. 9. ease and rapidity with which its results are obtained. Except with low percentage of Hb, it is tolerably accurate, but much less reliable than Fleischl's instrument. With low percentages of Hb it is well to use a double quantity of blood, halving the result. Aside from any inaccuracy in the construction of the apparatus, errors arise chiefly from the difficulty of adding exactly the proper quantity of water to the blood, and the imperfect comparison of red eolors in daylight. The apparatus consists of two glass tubes (A, B) of exactly equal caliber, one of which is partly filled with gelatine col- ored by picrocarmine so as to represent the color of a 1 -percent solu- tion of normal blood. The second tube carries a grad- uated scale from 10 to 120 and serves to hold the di- luted blood. The capil- lary pipette C measures 20 cmm., the quantity of blood to be used. In making the test the pipette is filled to the mark 20 cmm. with blood ob- tained under the usual pre- cautions. The specimen is quickly discharged into the tube B, in which a few drops of distilled water have previously been placed. The pipette must then be washed once or twice into the tube, the distilled water removing all traces of blood adher- ent to the inside of the capillary tube. Distilled water is now added drop by drop, until the solution of blood, carefully shaken and mixed, ex- actly matches the carmine-gelatine. The percentage of Hb is indicated by the height of the solution on the scale, reading from the middle of the meniscus. The colors match in daylight, and the eye may be as- sisted by holding the tube in front of white paper. 2. Fleischl's Hemoglobinometer. («) Apparatus. — This apparatus consists of a metal stand with plate and plaster mirror, 8, which casts diffused light through a circular opening in the plate. Beneath the plate, by means of a rack and wheel {T), slides a colored glass wedge fixed in a graduated frame, P. The glass wedge and graduated scale are arranged so as to indicate the percentage of Hb correspond- ing to the different portions of the wedge. In the circular opening of the plate fits a cylindrical metallic cell ((r), with glass bottom and Gower's hemoglobinometer. 40 GENERAL PHYSIOLOGY AND PATHOLOGY. ii metal partition, one compartment of which lies directly over the glass wedge. The other compartment {a) being filled with diluted blood, one is enabled to make a close comparison of the color of the dissolved blood with that of the glass wedge. The blood is measured by an au- tomatic capillary pipette, while a slowly running dropper is provided with which to add distilled water. On the handle of each pipette is stamped a number, indicating the cubic content of the tube. On the stand of each instrument is also a number, showing the capacity of the tubes Avith which it can be used. (6) Procedure. — One should first see that the automatic pipette is in working order, by blowing it out several times with water until it fills instantly and completely, after which it must be thoroughly dried. A drop of blood having been expressed under very strict pre- cautions against pressure, one end of the pipette is lightly touched to the drop wliich instantly fills the tube. There should be neither nega- tive nor positive meniscus to the column of blood, but the tube should be level full at either end. It should not be immersed in the blood drop, otherwise blood will adhere to the sides which cannot safely be removed. The tube of blood is immediately transferred to one com- partment of the cell which has been half filled with distilled water, and the blood is thoroughly Fig. 10. dissolved by moving the tube rapidly from side to side. On withdrawing the tube it should be washed into the cell with a few drops of distilled water. With the handle of the pipette the blood which collects in the cor- ners of the chamber is thoroughly mixed. Both chambers of the cell may now be filled level full with water. The thick round cover-glass should then be adjusted, avoid- ing the inclosure of air. Serious error may here result by forcing dis- solved blood over into the water compartment, or over the side of the cell. If the reading is made promptly the cover-glass need not be used, but after an exposure of 10-15 minutes the oxidizing action of the air may darken the blood and increase the reading 10-15 percent. The reading must be done in a dark room by means of candle- or gaslight. The colors do not match in daylight. The best results Fleischl's hemoglobinometer. THE ESTIMATION OF HEMOGLOBIN. 41 are obtained by placing the candle about eighteen inches from the stand, and by looking through an improvized paper tube which ex- actly fits the cell. With low percentages of Hb a very dim light is essential. When the cell is in place, and the light adjusted, the wedge is moved with quick rather than gradual turns until the color of the glass exactly matches that of the dissolved blood, when the percentage of Hb may be read on the scale. Several expedients are employed to assist the eye in the comparison of colors. It is well to relieve one eye with the other so as not to exhaust the color sense. The strength of the light may be varied by altering its distance, but the faintest distinct light is usually found to be the best. Two or three readings should always be made and the average taken. As the graduation of the instrument is more accurate and the matching of colors more exact in the middle of the scale, it is advisable to use a double quantity of blood when dealing with low percentages of Hb. (c) Limitations of Fleischl's Instrument. — With considerable experi- ence and constant care Fleischl's instrument yields results which are sufficiently accurate for most clinical purposes. One cannot attach any significance, however, to a variation of five percent, within which figure the ordinary error ought to be limited. All the causes of error mentioned as affecting the number of red cells may also disturb the percentage of Hb in the blood drop. In the procedure the principal difficulties relate to the even filling of the tube, the thorough cleaning of the tube, the overflow of blood solu- tion into the adjoining chamber, the proper adjustment and regulation of the light, and the accurate matching of colors. The chief sources of error in this method probably lie in the con- struction of the instrument. It is a common experience to find that different instruments give different results with the same specimen of blood, which Limbeck satisfied himself were referable principally to differences in the glass wedges. The writer tested his own blood with 16 different instruments, two of which were old (1892), the others of more recent importation (1897). With the two old instruments the Hb registered 87 percent and 90 percent; with the newer instruments it varied between 97 percent and 105 percent, except with one which gave 85 percent with different cells and different capillary tubes. The colored wedge with this instrument was broader and darker than usual, and when it was replaced by another wedge the instrument registered 100 per- cent with the same specimen. The error attributable to variations in different tubes and cells the writer finds is seldom greater than five per- cent among the newer instruments. Many old instruments read 10—15 percent lower than normal. The possible combination of these various defects in construction renders it desirable that one should test every new instrument for him- self using the blood of five or six healthy subjects. The newer in- struments will not be found to require much correction. Miescher's Modification of Fleischl's Hemoglobinometer. — 42 GENERAL PHYSIOLOGY AND PATHOLOGY. Miescher's improvements have removed some important defects in Fleischl's instrument, and as it now stands the improved hemoglobin- ometer yields results that leave little to be desired in point of accuracy. With this apparatus the blood is diluted by means of a graduated pipette very similar to that of Thoma, but yielding dilntions of 1-200, 1-300, and 1-400, according as the tube is filled with blood to the mark i, |, or ^. Two cells are provided, one with a depth of 15 mm., the other with a depth of 12 mm., the percentage of Hb being obtained with the deeper cell, and the other being used as a control specimen and giving only I (i|) of the actual percentage of Hb. These cells have a projecting jpartition dividing the compartments, Miescher's hemoglobinometer. along which a grooved cover-glass, D, may be slid, without fear of mix- ing the blood and water. If an excess of the blood solution overflows the side of the cell during the adjustment of the cover, no harm results as the blood is already properly diluted in the pipette, and the depth in the chamber will always be 15 mm. (or 12). These changes add very much to the facility and precision of the method. Finally, the cells are covered with diaphragms transmitting a ray of light which includes only three degrees on the scale, thus giving prac- tically a single color of the wedge for comparison with the blood. In the procedure, the blood is diluted in the pipette as with the hematocytometer, the diluting fluid being distilled water, or a filtered 1- percent solution of sodium carbonate. After shaking and clearing the tube of diluent, one chamber in each cell is filled with diluted blood. OLIVER'S HEMOGLOBINOMETER. 43 the opposite chamber with distilled water, and cover-glasses and dia- phragms are adjusted. Using a small candle and shielding the eyes from light, the readings with the two cells are carefully taken. The reading with the small cell should be ^ that with the larger. If there is any variation one reading may be used to correct the other. For example, suppose the readings to be : For the larger chamber (15 mm.) 64.0 For the smaller chamber (12 mm.) 50. If the first reading were absolutely correct the second reading should have been 51.2, since 64 x |- = 51.2. Or, assuming the second to be correct, the first should have been 62.5, since 50 x f = 62.5. The mean of 64 and 62.5, i. e., 63.25, should be taken as the true value. If the original dilution \Vas 1-200, the percentage of Hb is 63.25, the corrected result with the larger cell, but if the dilution has been 1-300, this result must be multiplied by Ih, or if 1-400, by 2. The technical difficulties of this method are so slight and the results so accurate that the instrument may be recommended over any other yet devised for this purpose. Oliver's Hemoglobinometer. — This instrument is constructed on the excellent principle of the tintometer, which is extensively used in various arts and industries. It consists of a series of six red glass discs, («), mounted upon white plaster mirrors in convenient frames. These discs represent the colors of twelve solutions of blood containing twelve different proportions of Hb. Two sets of discs are made, one for reading in daylight, the other for candlelight. The latter give more accurate results and should be chosen. The intermediate percentages of Hb are secured by means of '' riders " of colored glass to be placed over the discs, and which represent respectively 2|^ and 5 percent of Hb in the upper half of the scale, but twice that amount in the lower half. The discs are graded according to the specific dilution-curve of Hb. The blood is measured in an automatic pipette, c, and diluted in a cell, e, provided with a white, plaster, glass-covered bottom, and blue glass cover, which facilitates the reading. The blood is washed from the tube by means of a pipette, d, with rubber nozzle. The tube may be cleaned by means of a thread and needle. Procedure. — The capillary tube is filled by touching it to the blood- drop, and the blood is washed into the cell by attaching the rubber nozzle of the pipette, filled with distilled water, taking care not to as- pirate any blood into the pipette. The handle of the tube is then used to mix the blood, and water is carefully added till the cell is level full. The cover is then applied and should inclose a minute air bubble show- ing that the chamber has not been overfilled. The comparison of colors should be made by candlelight, the candle being placed at a convenient distance from the discs, and the eye shielded by means of a hood or paper tube. Oliver uses a special camera with green-glass eye-piece. If the color of the specimen matches any one of the discs the reading is completed. If it does not, the discs may be varied by using the 44 GENERAL PHYSIOLOGY AND PATHOLOGY. " riders," adding a clear glass disc to the specimen to compensate for the thickness of the " rider." Oliver's instrument presents the advantage of a comparison with a Fio. 12. Oliver's hemoglobinometer : a, set of standard colored discs ; b, lancet ; c, capillary pipette ; d, dropper ; e, mixing chamber. ^ngle color instead of with a scale of 10 percent +, as in Fleischrs. ihis difficulty is slight, however, and is overcome in Miescher's modi- hcation of Fleischl's instrument. The color of FleischPs glass wedge ESTIMATION OF THE IRON OF THE BLOOD. 45 varies accordiDg to the specific dilution-curve of the colored glass, which differs from the specific dilution-curve of Hb. How serious an error may arise with Fleischl's instrument from this cause has not been clearly shown. It is probably not great, as it is not apparent in the use of Miescher's modification in which different parts of the scale are employed. Yet the principle of Oliver's instrument is, in this respect, a distinct improvement. The technical difficulties are considerably less with Miescher's hemoglobinometer. Oliver's instrument is at present held out of the market on account of its excessive price. Estimation of the Iron of the Blood. Jolles' Ferrometer. — JoUes has devised a method and designed apparatus for the estimation of the iron of the blood, which is well adapted to clinical purposes. The apparatus is manufactured by Reichert, and full directions in German accompany each set. The procedure is as follows : By means of a pipette .05 cc. of blood is transferred to a platinum crucible, and the adherent blood washed out with a few drops of water. The blood is then evaporated and incinerated over a Bunsen Fig. 13. Jolles' ferrometer. flame. The ash is melted with .1 gr. water-free potassium bisulphate, until white fumes of sulphuric anhydride cease to rise from the dish. Papers containing the requisite amount of potassium bisulphate accompany the instrument. After cooling, the ash is washed into cyl- inder C (Fig. 14) with about 5 cc. of hot distilled water which is added till the whole quantity is 10 cc. In cylinder C, 1 cc. of the standard solution of iron (.00005 gr. iron oxide with potassium sulphate) is measured by a pipette and distilled water added to the mark 10 cc. Both cylinders are now 46 GENERAL PHYSIOLOGY AND PATHOLOGY placed in the stand (Fig. 14), and when at even temperature 1 cc. of dihite HCl (33 percent) is added to each. To cylinder C' should then be added 4 cc. of the solution of am- monium sulphocyanide, and to cylinder C, about 3.5 cc, and both are shaken after covering the ends with glass plates. Cylinder C containing the blood is now filled with sulphocyanide until it presents a positive meniscus when it is permanently covered and sealed by its glass plate. It may then be placed in the colorim- eter. In cylinder C the aluminum float is then adjusted, free from air bubbles, and it is also placed in position in the colorimeter. The further procedure consists in bringing the color of C" to match that in C, when viewed from above in the colorimeter. For this pur- pose the fluid in C is allowed to run out drop by drop through a stop- cock until the two colors are exactly Fig. U. alike. The comparison should be made in daylight. When the colors exactly match, the height of the necessary column of fluid in cylinder C may be read off and the per- centage weight of iron determined by reference to a table accompany- ing the instrument. The Hb may then be found according to v. Jak- sch's formula. Hb = 100 X m .42 in which la = the percentage weight of metallic iron. In order to secure accurate results with the ferrometer it is necessary to observe the same care that is re- quired in all quantitative chemical joUes' ferrometer. aualyscs, cspccially to avoid the loss of fluids by sputtering from the hot crucible, the use of wet and unclean instruments, and the inclos- ure of air bubbles, etc. Under most conditions the Hb may be accu- rately determined by means of the above formula, but from the consid- erations mentioned under " The occurrence of iron in the blood," it will be seen that there* may be considerable variations between the iron and the Hb-content. The ferrometer, therefore, usually gives a higher proportion of Hb than does Fleischl's instrument. THE HISTOLOGICAL EXAMINATION OF BLOOD. The greater part of the examination of blood is conducted in dry stamed specimens. To prepare such specimens for staining one re- THE HISTOLOGICAL EXAMINATION OF BLOOD. 47 quires only polished glass slides aud a Buiisen gas-burner. The glass slides must be thoroughly cleaned with soap and water, dried, and kept free from dust. A rather small compact drop of blood expressed from the finger tip under the usual precautions, is lightly scraped off with the polished edge of one slide and applied to one end of a second slide which should lie on firm support. When the blood has spread along the edge of the smearer it should be slowly and firmly drawn over the surface of the receiving slide. The drop should, if possible, be small enough to be exhausted in the smearing, and the thickness of the layer can be fully controlled by the degree of pressure. The blood should be pushed before the smearer and not trailed after. (See sketch.) Many prefer to use cover-glasses in spreading the blood. One polished cover-glass is touched to the drop of blood and applied to a second cover, all corners projecting. When the blood has spread to the edges the cover-glasses are gently slid apart without pressure. Fig. 15. Method of making blood smears. The cover-glasses should be handled with forceps, otherwise the mois- ture of the finger will often crenate many cells. The writer prefers to use glass slides, finding that beginners are much more successful with the slides than with cover-glasses, that, after very little practice, eveiy specimen can be spread successfully ; that forceps are not required ; that slides may be handled and trans- ported without fear of breakage ; that they need not be mounted and therefore do not fade like cover-glass specimens which require mount- ing in balsam ; that they may be restained, if necessary ; and above all that they may safely be fixed in the free flame. After spreading, all specimens should be well dried in the air. They may then be kept for weeks if wrapped in tissue paper and kept from moisture, but it is better io fix them at once. Fixation. 1. Heat. — In routine work one may discard all other methods for that of fixation in the free fiame of a Bunsen burner. The slide, specimen side up, is passed slowly through the flame until it is 48 GENERAL PHYSIOLOGY AND PATHOLOGY. decidedly too hot for the hand to bear. At this temperature, which probably varies between 110° and 150° C. fixation is complete in one to two minutes. A little practice will give the confidence necessary to heat the slides hot enouo-h, as one's initial failures from this method almost always result from incomplete fixation and subsequent vacuolization of the red cells. Overheated slides can usually be seen to change color in the flame, after which the red cells stain yellowish with eosin. The beginner is strongly recommended to perfect himself in this simple method of fixation. Small ovens, provided with a thermometer, are made for the fixation of blood slides, and may be used when many specimens are in hand, or when one does not care to risk the free flame. Specimens should be exposed 5-10 minutes to a temperature of 110°-120° C. 2. Alcohol. — Fixation for 10-30 minutes in 97-percent alcohol, or in equal parts of alcohol and ether, is a very reliable method, in very general use. Specimens may be left in alcohol for twenty-four hours, but do not then stain quite so well. There appears to be no advantage in adding ether to the alcohol, which even without mixture with the more volatile agent must frequently be replaced. Fixation in alcohol is to be specially recommended for the malarial parasite, but is unsat- isfactory when Ehrlich's triacid stain is to be used. 3. Fixation by Vapors. — Specimens may be fixed by being laid, speci- men side down, over a wide-mouthed bottle containing : 25-percent formalin, to which the exposure is five minutes, or, 2-percent osmic acid, to which expose two minutes. Both these fluids have to be re- placed frequently, they considerably alter the staining relations of the blood cells, and are inferior to other methods of fixation. Methods of Staining Dry Blood Specimens. 1. Eosin and Methylene -blue. — The solutions required are : A saturated alcoholic solution of Ehrlich's blood-eosin. A saturated watery solution (1-per- cent) of Ehrlich's rectified methylene-blue. The latter should be at least one week old, as fresh solutions lack selective quality and stain the specimen diffusely. After several weeks, methylene-blue in solu- tion diminishes in staining power, while the alcoholic eosin absorbs water, and becomes less selective and more powerful. In staining, flood the specimen with eosin for a few seconds and wash in water. If the stain is not effective add more eosin, but the water on the slide dilutes the alcohol and renders the second applica- tion of eosin much more powerful than the first. Next, flood the specimen repeatedly for one minute with methylene-blue, wash hastily in water, and dry. This method may be recommended for all ordinary examinations. The blood is stained as shown in Plate II., readily distinguishing the various forms of normal leucocytes. It does not stain neutrophile granules in leucocytes unless the action of eosin has been prolonged, in which case the neutrophile leucocytes can be distinguished from the eosinophile only by the size of the granules. Its chief advantage is ESTIMATION OF SPECIFIC GRAVITY OF THE BLOOD. 49 the clear differentiation of basophilic leucocytes and of nuclear struc- tures. It clearly demonstrates the malarial parasite but in this field is greatly inferior to Nocht's method. Its chief disadvantage is the danger of overstaining with eosin which prevents the full action of methy leue-bl ue. 2. Ehrlich's Triacid Mixture. — This agent has the following com- position : Sat. watery sol., Orange G 120-135 cc. Acid Fuchsine 80-165 Methyl -green 125 To these add Aqua 300 Absolute alcohol 200 Glycerine 100 The attempt to prepare this mixture is not always successful. The smaller quantities of orange G, and acid fuchsine are best employed, and the solution of methyl-green, well seasoned, should be added slowly, with stirring, to the mixture of the other dyes. The water should be added next, then the alcohol, and finally the glycerin, with constant stirring. After standing one week the mixture is ready for use. Griibler's preparation of this mixture is in the market and is reliable. In staining, it is only necessary to flood the specimen with the dye for one to two minutes, and wash hastily in water. It cannot over- stain, but overheated specimens are usually faint, and the red cells are yellowish. It stains neutrophile and eosinophile granules deep red, the latter being distinguished by their size. It is therefore indis- pensable in the diagnosis of leukemia. It is a poor nuclear stain, fails to demonstrate the structure of normal mononuclear leucocytes, and does not stain the malarial parasite. On account of the uniformity of its results many prefer it to eosin and methylene-blue as a routine method. 3. Demonstration of " Mast-cells." — The large basophilic granules of these cells retain basic dyes with tenacity, and may be demonstrated by a mixture of one of these dyes, with a strong decolorizer, which removes the stain from most other basophile structures. Ehrlich's dahlia solution is adapted to this purpose : Abs. alcohol 50 cc. Glacial acetic acid 12.5 cc. Distilled water 100. cc. Add dahlia to saturation. Stain several hours, wash in water, decolorize in alcohol or more rapidly in 20-percent Ac, till the nuclei fade, and wash in water. The nuclei of leucocytes are then very pale blue, the mast-cell granules very dark blue or black. ESTIMATION OF THE SPECIFIC GRAVITY OF THE BLOOD. Hammerschlag's Method is the most practical of the various indirect procedures devised for this purpose. A small urino- meter of suitable dimensions is partly filled with a mixture of chloro- 4 50 GENERAL PHYSIOLOGY AND PATHOLOGY. form (s. g. 1.526) and benzine (s. g. .889), of a gravity of about 1.060. By means of a pipette, such as the red cell mixer of Thoma, a drop of blood, expressed with the usual precautions, is transferred to the fluid. In expelling the blood the tip of the pipette should be submerged and no air should be allowed to pass out with the blood. The drop should not be very minute in size and should float on the fluid. If it is allowed to sink, it will often be lost by spreading out on the bottom of the vessel. By adding chloroform or benzine, drop by drop, as required, and carefully mixing by inverting the urinometer closed by the palm, a mixture is secured in w^hich the drop neither rises nor sinks, but which is of exactly the same density as the blood. The specific gravity of the mixture, and of the blood, may then be taken as with urine. The urinometer should be graduated up to 1,065, and should be tested in distilled water at 60° F. The apparatus should be clean and perfectly dry. The mixture may be filtered and used repeatedly. Great accuracy can hardly be expected from this method yet it is sufficiently reliable for clinical purposes. Errors arise from changes in the blood drop during and after its transfer to the mixture, from evaporation, from the escape of gases, from the inclosure or adherence of minute air bubbles, and perhaps also from the possible action of chloroform and benzine upon the blood which is at present an unknown factor. Accordingly, Hammerschlag's method gives a uniformly higher gravity than is obtained by more accurate direct estimates. Practically, in performing the test one finds that the behavior of the drops varies, for reasons which are not clear, some rising and others falling in the same mixture. A large drop should be secured and fol- lowed in the test. Since the specific gravity of the blood in simple anemia varies principally with the Hb, the percentage of Hb may in many cases be calculated with considerable accuracy from the specific gravity. Ham- merschlag has prepared the following table showing the relation of Hb to the specific gravity as determined by his method : Specific Gravity. Hb. 1033-1035 25-30 percent. 1035-1038 30 35 " 1038-1040 35-40 " 1040-1045 40-45 " 1045-1048 .....45-55 " 1048-1050 55-65 " 1050-1053 65-70 " 1053-1055 70-75 " 1055-1057 75-85 " 1057-1060 85-95 " The suggestion that Fleischl's hemoglobinometer be discarded for this indirect method of estimating Hb has not found favor. The changes in the plasma in severe anemia, leukemia, dropsy, and diarrheal dis- eases, renders this practice always unscientific and frequently very un- reliable. (Cf. Siegel, and Stintzing and Gumprecht.) ESTIMATION OF ALKALESCENCE OF THE BLOOD. 51 Schmaltz' Method. — A thin-walled capillary glass tube is prepared, about 12 mm. in length, aud of a calibre of about | mm. at the ends and li mm. at the middle. This is thoroughly cleaned, dried and weighed Ijefore and after filling with distilled water. After drying with ether the tube is filled with blood and again weighed. The weight of the lilood divided by that of the water gives the specific gravity of the former. Fig. 16. Capillary glass tube adapted to various details of blood analysis. This is the best of the methods requiring a small quantity of blood, and in experienced hands gives more accurate results than Hammer- schlag's method. It is to be especially recommended in experimental and laboratory work, but usually requires too much blood for its adop- tion as a routine clinical method, although the quantity required, two drops, is not great. The chief sources of error are in the difficulty of thoroughly cleaning the tube, and inaccuracy in the use of the scales. Schmaltz found by controlling his results by the use of salt solutions of known density that the error did not exceed .003 in the computed gravities. Comparing the percentage of Hb obtained by Fleischl's method with the specific gravity obtained by his own method, Schmaltz con- structed the following table : Specific Gravity. Hb. Specific Gravity. Hb. 1030 20 percent, dr 1049 60 percent. 1035 30 " 1051 65 " 1038 35 " 1052 70 " 1041 40 " 1053.5 75 " 1042.5 45 " 1056 80 " 1045.5 50 " 1057.5 90 " 1048 55 " 1059 100 " ESTIMATION OF THE ALKALESCENCE OF THE BLOOD. The determination of the alkalescence of the blood is attended with very great practical difficulties. This alkalescence being referable to the presence of carbonates, bicarbonates, and of albumens which are retained in solution by acid phosphates, it is always difficult to judge of the changes in these principles and the consequent variations in re- action produced by the procedures required in alkalimetry. If serum alone is titrated, the alkaline principles of the clot are left out of ac- count, and if " laked " blood is employed, there is an uncertain factor in the chemical changes produced, especially in the delicately balanced albumens and phosphates, during the solution of red cells. Nevertheless it appears from a considerable number of painstaking studies by Landois, v. Jaksch, Kraus, Tausczk, Lowy, Schultz- Schultzenstein, Limbeck, and many others, that in blood and in serum there is a fairly constant group of alkaline principles which 52 GENERAL PHYSIOLOGY AND PATHOLOGY. may be rather accurately measured and which have a distinct and im- portant relation to disease (cf. Blood in Fever). Four of the many methods employed may be recommended as most reliable. 1. Lowy's Method. — In a 50-cc. flask containing 45 cc. of 2-per- cent solution of ammonium oxalate, 5 cc. or less of fresh blood are accurately measured, and dissolved. Of the solution 5 cc. are titrated by means of a J^ normal solution of tartaric acid, using litmus paper as an indicator. The latter may be prepared by soaking prepared paper in an alcoholic solution of litmus to which dilute HCl has been added till a violet color appears. The end reaction is obtained by adding a drop of blood solution to the paper and closely inspecting the color of the outer zone into which the fluid diffuses. The result is not affected by changes in temperatures. Engel's Alkalimeter. — Engel has devised an apparatus for the clin- ical estimation of alkalinity according to Lowy's method. A large drop of blood is drawn into a special pipette up to the mark 0.05 and diluted with distilled water to the mark 5.0. After shaking, the dis- solved blood is discharged into a glass cylinder and titrated by -^-^ normal solution of tartaric acid (ac. tartar. 1 gr., aq. dest. 1 1.). The Fig. 17. Eugel's alkalimeter. acid is added drop by drop until a distinct red zone appears when a drop of blood is allowed to diffuse through litmus paper. In normal blood about 10 drops of acid bring the end reaction. Computation. — If .4 cc. of acid are required to neutralize .05 cc. of blood, 8 cc. of acid will be required to neutralize 1 cc. of blood. One cubic centimeter of j\ normal tartaric acid neutrahzes .533 mg. NaOH (Engel), so that 8 cc. of acid solution indicate the presence in 1 cc. of ESTIMATION OF OSMOTIC TENSION OF THE PLASMA. 53 blood of 4,264 mg. of NaOH, to wbich terms the alkalinity of the blood is usually reduced. 2. Method of Schultz-Schultzenstein. — By means of the pipette of the Fleischl hemoglobinometer, 5 or 7.5 mg. of blood are measured, which is dissolved in 12 cc, of distilled water. This is acidified by adding 1.5 cc. y^-g- normal H.,SO^. After careful mixing a drop or two of ethereal solution of erythrosin is added as an indicator and the solution titrated, with frequent stirring, by g-^^- normal solution of KOH. The end reaction is shown by the first appearance of a red color in the supernatant ether. The test must be performed speedily, to avoid obscuring the end reaction by a layer of fibrin precipitated by the ether. The requirement of a very small quantity of blood renders this method specially suitable for clinical purposes. 3. V. Limbeck's Method, — To about 200 cc, of boiled distilled water are added 5 cc, decinormal HCl solution, and drop by drop, with stirring, 5 cc, of serum spontaneously expressed from a clot. The stir- ring rod should be covered with black gutta percha. The resulting clear and slightly opalescent fluid is now titrated with deci-normal so- lution of NaOH. After adding a few drops a precipitate forms which soon dissolves. The end reaction is reached when the abundant pre- cipitate (of albumen) no longer dissolves, which is best determined by finding a flocculent precipitate persisting on the black rod. The same quantity (5 cc.) of fresh blood may be titrated by this method and the total alkalinity of the blood determined. In adding blood, however, it must be dropped very carefully into the hot water, to avoid coagulation. The chief technical difficulty with this method lies in detecting the end reaction. In case of doubt the precipitate may be dissolved by adding 1-2 cc. of acid, as above, and titrating as before. There is also the objection that it takes no account of the alkalinity referable to albumens. Limbeck doubts, however, if the capacity of albumens to neutralize acids ever comes into action in the body, and argues that his method estimates exactly, without regard to the quan- tity of albumen, the alkalinity of the salts present in the blood, 4. Wright has devised and extensively used one of the most practical clinical methods of estimating the alkalinity of the blood for the full explanation of which the reader should consult the original description. Determination of the Coagulability of the Blood. Vierordt and Wright have employed methods of determining the coagulability of the blood the results from which do not appear to be sufficiently accurate even for clinical purposes. (Limbeck.) Estimation of the Osmotic Tension of the Plasma. Although comparatively little attention is usually paid to the osmotic tension of the plasma, it is evident from the frequent occurrence of 54 GENERAL PHYSIOLOGY AND PATHOLOGY. hemoglobiaemia in various forms of severe anemia, malarial and espe- cially hemoglobinuric fever, jaundice, acute poisonings, etc., that the condition of the blood in this respect is of prime importance in the clinical and pathological study of these diseases. In Hammerschlag^ s method we possess ready and very exact means of investigating this problem. In a series of test tubes are poured small quantities, accurately measured, of the serum to be examined, and to each is added an in- creasing quantity of distilled water. The surface of the mixed fluids in each tube is then touched with a pipette holding a very little normal blood, and the tubes are allowed to stand twelve hours. By that time the cells have settled to the bottom in some tubes, while in others the supernatant fluid is tinged with dissolved Hb, and one notes that tube in which the first traces of Hb appear. Normal red blood-cells begin to lose their Hb in solutions of salt containing any less than .46 per- cent of NaCl. Estimated in terms of NaCl that tube of the series which shows the first faint traces of dissolved Hb, therefore shows the same osmotic tension as a .46-percent solution of salt. The dilution being known the tension of the original serum can be computed as fol- lows : Suppose that to 1 cc. of serum .9 cc. of water added, caused the solution of Hb. Then the tension of the original serum is equivalent to 1 -h .9 X .46 = .874 percent of NaCl. Bremer's Specific Reaction of Diabetic Blood. Saturated watery solutions of eosin (watery ?) and of methylene- blue are mixed in about equal proportions so that a neutral reaction is obtained, and the mixture produces little or no stain on litmus paper. A precipitate forms, soluble in alcohol, insoluble in water, which is filtered, washed, dried, and powdered. To 24 parts of this powder are added six of powdered methylene-blue and one of eosin. Of this mixed reddish brown powder .025 to .05 grm. are dissolved in 10 cc. of 33-percent alcohol, in which solution the specimen is stained for four minutes. The specimens are prepared by smearing tlie blood on glass slides or covers, and fixed by boiling in equal parts of alcohol and ether for four minutes. This may be accomplished by placing the bottle of alcohol and ether in hot water at 60° C. After washing the stained specimens in water, diabetic blood has a greenish tint, while normal blood is reddish violet, and on microscop- ical examination the erythrocytes of diabetic blood are found to be greenisli, while those of normal blood are red. Bremer found the reaction in 50 out of 51 cases of diabetes. The negative result was obtained in a well-established case, a boy of six- teen, in whom glycosuria began after an electric shock. He found that the reaction persisted in the absence of glycosuria ; that normal blood floated on diabetic urine for 15 minutes gave the reaction ; that blood treated with solutions of glucose failed to give the reaction ; DEMONSTRATION OF GLYCOGEN IN BLOOD. 55 that the blood iu glycosuria artificially produced in animals by pliloro- glucin gave the reaction, while in that produced by phloridzin it did not. The nature of the reaction is not understood. While some observers have convinced themselves that the presence of glucose is not sufficient to bring about the altered staining qualities of the blood, Hartwig concluded that the glucose first causes a change in the Hb which shows itself in the altered reaction of the red cells to anilin dyes. Various modifications of the above most approved method have been employed successfully by Bremer and others. One of these, em- ployed by Bremer, is as follows : 1 -percent solutions of Congo red or of methylene-blue stain diabetic blood very slightly, while 1 -per- cent solution of Biebrich-scarlet stains it intensely. A directly op- posite relation holds with normal blood. Rather thick smears of blood should be used with this procedure, and the colors compared by the naked eye. The value of Bremer's test has been confirmed by Le Goff, Eichner and Folkel, Lepine and Lyonnet, James, Jeanselme, Badger, and Hart- wig, but similar reactions have been found in normal blood in leu- kaemia, Hodgkin's disease, exophthalmic goitre, and multiple neuritis. A partial reaction has been obtained in cachectic conditions, and Bremer failed to find it in a case of glycosuria of neurotic origin. Yet in most conditions, other than diabetes, the reaction has been found, when present, to be inconstant, and to occur in a very small proportion of cases (Lepine, Eichner, Hartwig). The technical difficulties in carrying out the test are considerable, while a slight variation in technique appears to vitiate the result, as is indicated by the failures reported by Patella and Mori, after both Bremer's and Williamson's methods. Williamson finds that diabetic blood decolorizes solutions of meth- ylene-blue, while normal blood does not. His test is performed as fol- lows : 20 cmm. (2 drops) of blood are dissolved in 40 cmm. of water and to the solution is added 1 cc. of methylene-blue (1-6,000 solution) and 40 cmm. of Liquor Potassce (s. g. 1.058). The vessel is then placed in boiling water for four minutes ; diabetic blood decolorizes the solution, normal blood leaves it deep blue. Diabetic urine has the same effect. Williamson found this reaction in 6 diabetics and failed to find it in 160 cases of other diseases, including one of leukemia. Demonstration of Glycogen in Blood. Gabritschewsky's method may be employed. The blood smears thoroughly dried in the air, are stained for several minutes in : iodum pur. 1 ; KI, 3 ; aq., 100 ; acacia pulv. in excess. The presence of glycogen is indicated by the appearance of mahog- any-brown granules of variable size, in leucocytes and plasma. (See Plate X., Fig. 2.) 56 GENERAL PHYSIOLOGY AND PATHOLOGY. Czerny claimed that this method demonstrates the presence in the blood not of glycogen but of a carbohydrate more nearly related to amyloid, Huppert's later studies on the blood of animals support the behef that the substance thus demonstrated is really glycogen. It is probable that many colorless globules visible in leucocytes treated by ordinary staining methods are referable to the former pres- ence of glycogen, which is soluble in water and of which the reactions are largely destroyed by heat. Demonstration of Fat in the Blood. The blood smear is fixed in 1-percent osmic acid for 24 hours and counterstained with eosin. The fat particles are then stained black. Since all that blackens under osmic acid is not fat, a control prepara- FiG. 18. Pronounced lipemia. Specimen treated with osmic acid. Lower half shows extra-cellular fat globules, upper half having been cleared by oil of turpentine. (Gumpkecht. ) tion should be fixed 24 hours in alcohol and ether, then in 1-percent osmic acid for 24 hours, counterstained with eosin, and the extraction of the fat by ether demonstrated by the absence of black particles in cells and plasma. BiBLIOGEAPHY. Badger. Cited by Cabot. Bergmann. Cited by Limbeck, Inaug. Dis., Miinchen, 1884. Biernacki. Zeit. f. Physiol. Chem., Bd. 19, p. 179. Zeit. f. kl. Med., Bd. 24, p. 460. Bleibtreu. Pfliiger" s Archiv, Bd. 51, p. 151. BIBLIOGRAPHY. 57 Blix, Hedin. Scand. Arch. f. Physiol., 1890, p. 134. Brandenburg. Munch, med. Wocli., 1900, p. 183. Bremer. N. Y. Med. Jour., Vol. 63, p. 301. Ibid., Vol. 66, p. 808. Cent. f. inn. Med., 1897, p. 521. Buntzen. Virchow-Hirsch, Jahresb., 1879. Cheron. Compt. Eend. Acad. Sci., 1895, Vol. 121, p. 314. Cohnheim. Vorles. u. Allg. Pathol., Berlin, 1882, Bd. I., p. 397. Cohnstein, Zimtz. Pfliiger s Archiv, 1888, Bd. 42. Czerny. Archiv f. exp. Path. u. Pharm., Bd. 34, p. 268. Ibid., Bd. 31, p. 190. Daland. Fort. d. med., 1891, No. 21. Elzholz. Wien. klin. Woch., 1894, No. 32. Enqel. Berl. klin. Woch., 1898, p. 308. Eichner, Folkel. Wien. klin. Woch., 1897, p. 1103. Eyckman. Arch. f. ges. Physiol., 1895, Bd. 40, p. 340. Oabritschewsky. Arch. f. exper. Path., Bd. 28, p. 272. Gartner. Allg. Wiener med. Zeit., 1892, p. 513. Gartner, Romer. Wien. klin. Woch., 1892, No. 2. Grawitz. Zeit. f. klin. Med., Bd. 21, Bd. 22. Methodik d. kl. Blutuntersuch, Ber- lin, 1899. Hamburger. Die phvsiol. Kochsalzlos. u. d. Volumbestimmung d. korperl. Ele- mente im Blute. Cent.'f. Physiol., 1893, 1894, p. 656. Virchow's Archiv, Bd. 141, p. 230. Hammerschlag. Wien. klin. Woch., 1890, p. 1018. Hartwig. Deut. Arch. klin. Med., Bd. 62, p. 287. Hedin. Arch. f. ges. Physiol., Bd. 40, p. 360. Heidenhain. Versuch. zur Lehre von der Lymphbildung, Pfliiger's Archiv, 1891, Bd. 49, p. 209. Heissler. Arbeit, aus d. path. Instit. zu Miinchen, 1886, p. 322. Huppert. Zeit. f. Physiol. Chem., Bd. 18, p. 144. Jacobi, 31. P. N. Y. Med. Eecord, 1898, I., p. 933. V. Jaksch. Zeit. f. klin. Med., Bd. 13. Klein. Volkmann's Vortriige, No. 29. Knopf elmacher. Wien. klin. Woch., 1893, pp. 810, 886. Kraus. Arch. f. exper. Path., Bd. 26, p. 186. Landois. Text-book of Physiol., 1893. Lepine, Lyonnet. Lyon Med., T. 82, p. 187. Le Goff. React, chrom. du sang diabet, Paris, 1897. Leichtenstern. Unters. u. d. Hemoglobingehalt d. Blutes, Leipzig, 1878, p. 57. V. Lesser. Sitzungsber. d. k. Siichs. Gesell. d. Wissenschaften, 1873, p. 573 ; 1874, p. 153. Du Bois-Reymond's Archiv, 1878, p. 41. von Limbeck. Grundriss einer klin. Path. d. Blutes, Jena, 1897, p. 68. Lloyd Jones. Journal of Physiol., 1887, viii. Lowit. Studien z. Phys. u. Path. d. Blutes, etc., Jena, 1892. Lowy. Pfliiger's Archiv, Bd. 58, p. 462. Cent. f. med. Wissen., 1894, p. 785. Fort. d. Med. ,"'1898, p. 171. Ludwig. Cited by Lukjanow. Lukjanow. Grundziige ein. allg. Path, des Gefiiss-systems, Leipzig, 1894. Lyon, Thoma. Virchow's Archiv, Bd. 84, p. 131. Mitchell. Amer. Jour. Med. Sci., Vol. 107, p. 502. MuUer, Rieder. Deut. Arch. klin. Med., Bd. 48, p. 96. Oertel. Klinische Zeit- u. Streitfragen, Wien, 1889. Patella, Mori. Gaz. deg. Osped., 1896, IL, p. 1441. Pepper, Stengel. Cong. f. inn. Med., 1896, p. 631. Pfeiffer. Ibid., 1895, No. 4. Reinert. Zahlung der Blutkorp, Leipsic, 1891. V. Recklinghausen. Handb. der allg. Path. d. Kreislaufs u. der Ernahrung, Stutt- gart, 1883, p. 179. Schmaltz. Arch. f. klin. Med., Bd. 47, p. 145. Schidtz-Schultzenstein. Cent. f. med. Wissen., 1894, p. 801. Siegel. Wien. klin. Woch., 1891, No. 33. Stintzing, Gumprecht. Deut. Arch. f. klin. Med., Bd. 53, p. 265. Tauczk. Ungar. Arch. f. Med., 1895, p. 359. Thayer. Johns Hopkins Hosp. Bulletin, 1893, p. 37. Thibault. Bull. Acad. Med. Bruxelles, 1894, p. 112. Valentin. Reporter, f. Anat. u. Physiol., 1838, Bd. 3, p, 281. 58 GENERAL PHYSIOLOGY AND PATHOLOGY Vierordt. Arch. f. Heilk., 1878, Bd. 19, p. 193. Wagner's Handb. d. Physiol. 1842, Bd. 1, p. 84. Wenddstadt, Bleibtreu. Zeit. f. klin. Med., Bd. 25, p. 204. Wick. Wien. med. Zeit., 1887. Winternitz. Cent. klin. Med., 1893, Bd. 14, p. 177, 1017. Worm-Mulkr. Transfusion u. Plethora, Christiania, 1875. Wright. Lancet, 1897, II., pp. 719, 814, 833. ^ Lancet, 1892, I., pp. 457, 515. Zappert. Zeit. f. klin. Med., Bd. 23, p. 227. CHAPTER II. CHEMISTRY OF THE BLOOD. THE CHEMICAL COMPOSITION OF RED BLOOD CELLS. In the analyses of Schmidt, Hoppe-Seyler, and Judell, the red cells were separated from the serum by the addition of salts (Na,,SO^, NaCl), a procedure which considerably alters the chemical composition of cells and plasma. The accuracy of their results, which are the best avail- able, is therefore only approximate. The specific gravity of red cells is usually about 1.088. They con- tain about 90 percent of oxyhemoglobin and a small proportion of a globulin-like albumen (nucleo-proteid) coagulating at 75° C. There are also traces of lecithin and cholesterin. The principal salts are phosphates of Na, K, Ca, and Mg, and chloride of K, whereas in the serum the chief salt is NaCl. In cholera, dysentery, and dropsy, Schmidt found the red cells to be of increased specific gravity in proportion to the duration of the exudative process, while their chemical analysis showed that they par- ticipate in the changes which first affect the serum, losing first water, then salts, and finally albumen. More recently, v. Jaksch,^ after es- tablishing the normal content in N of the red cells, followed the varia- tions in this principle in various diseases. He found very marked and irregular variations both in health and disease, which indicate that the method used (determination of total N) was unreliable. Biernacki, estimating the dry residue of red cells settled after the addition of sodium oxalate, found a normal residue of 29.28—30 per- cent ; in carcinoma of esophagus 27.9 percent ; in tabes with anemia, 25.51 percent ; in chlorosis (Hb 25 percent) 22.24 percent ; in chronic nephritis (Hb 20 percent) 22.88 percent. Among the important ob- servations of Biernacki's are, the increased content in water of the red cells in hydremia ; the general parallelism between P.,0., Fe, and K, and between water and NaCl, in the red cells. Hemoglobin (Hb) and its Derivatives. Hemoglobin belongs to the group of proteids, containing about 96 percent of albumen, and 4 percent of an iron-holding pigment, hemo- ckromogen. Hb exists in the red cells in combination probably with the nucleo-proteid of the stroma. The nature of this union is not certainly understood, but it renders Hb comparatively insoluble, greatly concentrated, and capable of actively forming unstable com- pounds with oxygen. (Hoppe-Seyler, Gamgee.) 60 CHEMISTRY OF THE BLOOD. Its chemical composition is very complex and apparently variable, but its spectroscopic relations are constant. In the circulation it exists principally as reduced Hb in the veins, and in molecular union with oxygen as oxyhemoglobin, in the arteries. One gram of saturated oxy- hemoglobin yields about I.IG cc. of oxygen, but in the circulation the degree of saturation with oxygen varies and in health is never com- plete. Meyer and Biarnes found in the arterial blood of a dog 76 per- cent of saturation with oxygen ; after a large hemorrhage it rose to 85 percent of saturation, while after inducing CO poisoning it rose to 90 percent. Limbeck found 97 percent of saturation with oxy- gen in the blood of a dog poisoned with potassium chlorate. It thus appears that with a relative loss of functionating Hb, the demands of the system may cause a more complete saturation of the blood with oxygen than exists in health. Oxyhemoglobin is bright red in color, and forms with difficulty yellowish-red rhombic crystals. These crystals are very soluble in water and in very dilute solutions of alkaline carbonates, but when treated with strong alcohol they are modified and become insoluble (parheraoglobin of Nencki). They are insoluble in ether, chloroform, benzol, or carbon disulphide. Oxyhemoglobin is absolutely non-dif- fusible. (Gamgee.) Hemoglobin (Reduced Hb) is dichroitic. In thick layers, or in thin layers of concentrated solutions, it presents a dark cherry-red color, while very dilute solutions exhibit a green tint. (Gamgee.) This dichroism is characteristic of the blood of simple asphyxia. Reduced Hb is more soluble but less easily crystallizable than oxyhemoglobin. Methemoglobin is brownish red in color, readily soluble in water, and crystallizes in brownish red needles, prisms, and hexagonal plates. It contains the same proportion of oxygen as oxyhemoglobin, but in much firmer and probably in chemical union. Methemoglobin is found in the blood chiefly in cases of poisoning by a variety of sub- stances, the so-called -'blood poisons." Tschirkoflf claims to have found it in Addison's disease, and Ruyter has recognized a very similar but not identical coloring-matter in a case of malignant oedema. Hemoglobinemia, a solution of Hb in the plasma, a normal con- dition in the blood of some vertebrates, is in man always pathological, and results from lessened resistance on the part of the red cells, and from abnormal states of the plasma. The former condition is probably concerned in cases of paroxysmal hemoglobinemia and in the destruc- tion of blood which follows general burns (Lichtheim, Murri, Chvostek, Silberman), while the latter condition exists in the hemoglobinemia of acute poisonings and in cases referable to increased globulicidal power of the serum. Carbonic oxide Hb is a firm compound of Hb and CO, and im- parts a bluish red or rose-red color to the blood. Its crystals are similar to those of oxyhemoglobin but more bluish and less soluble. In cases of poisoning by the inhalation of fire-damp or illuminating gas, CO-Hb is formed in considerable proportion and the respiratory THE CHEMICAL COMPOSITION OF LEUCOCYTES. 61 capacity of the blood is largely destroyed. CO-Hb persists in the blood for several days, in cases that recover, and for a mnch longer period in fatal cases. Hem.vtix, one of the advanced decomposition products of Hb, is a dark brown or blackish, non-crystalline solid, decomposed at 180° C. It is insoluble in water, dilute acids, alcohol, ether and chloroform, but dissolves in acidified alcohol and ether, and readily in dilute alkalies. It is found in old bloody transudates, in the feces where digestive fluids have acted npon blood, and in the urine in cases of arsenical poisoning. Crystals of hnemin, or the hydrochlorate of hem- atin, are formed in Teichmann's test. Hematoidix is a crystalline derivative of Hb, found in old blood extravasations. It crystallizes in orange-colored rhombic plates, is soluble in chloroform, ether, carbon disulphide, and ammonium sul- phide, contains no iron, and gives no absorption bands although absorbing most of the violet end of the spectrum. It is generally regarded as identical with bilirubin. Hemosiderix is an amorphous iron-holding product of the decom- position of Hb, and is abundantly found in the viscera in disease accom- panied by much destruction of blood. It is probable that when red cells are disintegrated hemosiderin is formed by the action of living cells upon Hb, while hematoidin originates apart from any cellular activities. (Perls, Thoma, Ziegler.) Hemosiderin is blackened by am- monium sulphide and turned blue by acidified potassium ferrocyanide. Melaxix is a yellowish brown or black pigment produced by the action of the malarial parasite upon Hb. It is insoluble in water, alcohol, ether, chloroform, carbon disulphide, and acids in moderate strength, but is destroyed by heat, and is soluble in strong alkalies and in ammonium sulphide. It fails to yield the micro-chemical reac- tions of iron, and probably contains no iron. The term " melanin " is also loosely applied to a variety of black pigments occurring in the body, some of which contain iron but whose origin is uncertain. From " antialbumid," formed by heating egg-albumin witli 3-pei'cent HjSO^ for 10 liours at 100° C, and fi*om hemipeptone, Chittenden and Albro produced black melanin, by digesting these substances with 10 -percent H2SO4 and boiling for 79 hours. The change they believe to consist in a process of hydrolytic cleavage. Sulphur and fatty acids were thrown off in the process and the melanin was found to contain C, H, N, and S. THE CHEMICAL COMPOSITION OF LEUCOCYTES. Attempts to determine the chemical composition of leucocytes have been made by analyses of leukemic blood, of pus, and of fresh lymph- oid tissue. In all of these it cannot be doubted that abnormal products were constantly present, so that the chemical composition of normal leucocytes cannot be learned by such methods. Probably Lilienfeld's analyses^ of lymphoid tissue furnish the most reliable data. 62 CHEMISTRY OF THE BLOOD. In cadaveric leukemic blood various fatty acids have been isolated, and lecithin, a normal constituent of blood, has been found in excess. Glycogen has been repeatedly isolated in this and other conditions. In fresh pus, Miescher found five diiferent forms of albumen, an abundance of lecithin, and cholesterin, phosphates of Na, K, Ca, Mg, Fe, and chloride of Na. In analyses of the thymus gland in calves Kossel and Lilienfeld isolated various albumens, principally nucleins, lecithin, and chole- sterin, besides about 3 percent of inorganic principles. In older cells they believe that secondary products develop in the form of glycogen, protagon, and fats. The nuclein of Kossel is a combination of the organic nucleinic acid which contains phosphorus, with albumen, and in the leucocytes is found in com- bination with another albuminous body, histon, with which it forms nucleo- histon. Under certain conditions Kossel and Lilienfeld believe that nucle- inic acid becomes free in the cells, and exerts bactericidal action before and during phagocytosis. The nucleins are bodies obtained from animal (or vegetable) cells after digestion with pepsin to which they are resistant. They are rather insoluble in water, alcohol, and ether, give the biuret and Mil- Ion's reactions, and acting as strong acids and uniting with bases, they may be identified from the basic albumens by their tinctorial qualities. Boiled with dilute acids they yield " nuclein-bases " or xanthin-bodies ; or treated with alkali they yield albumen and nucleinic acid. They are rich in phosphorus and iron. Nucleinic acid is of variable com- position, but that from the calf's thymus is a combination of a xanthin body with a complex phosphoric acid. (Hammarsten.) THE CHEMICAL COMPOSITION OF BLOOD PLATES. Since the true nature of blood plates has been at least partly shown, the chief inquiry concerning them has related to their chemical com- position. Lowit has strongly maintained that they are composed of globulin, a claim which is probably true of some of the bodies which appear in coagulating shed blood and which must be classed with the blood plates of different origins. Lilienfeld ^ regards the substance of blood plates as belonging to the nucleo-albumens, and identifies them with the remnants of the nuclei of leucocytes, basing this opinion on their content in phosphorus, their resistance to digestion by pepsin, and their micro-chemical reac- tions. These observations may be regarded as definitely settling the question of their chemical composition, but do not prove their exclu- sive origin from leucocytes. THE SERUM. The Albumens of the Serum. — Blood serum contains two albu- minous bodies, serum albumen and serum globulin, which together THE SERUM. 63 form 7,62 percent (Hammarsten), or 8.26 percent (Schmidt) of the weight of normal serum. The quantitative changes in these principles in disease are not usu- ally very marked, owing to their relatively slight diffusibility. From the studies of Becquerel and Rodier, Schmidt, v. Jaksch, Lim- beck and Pick, it has been shown that the albumens of the serum are considerably diminished in severe anemias, and in nephritis or endo- carditis with dropsy, but that in most severe infectious diseases they are but slightly reduced and bear a fairly constant relation to the other solids of the serum. If temporarily reduced by some exudative lesion, as in diarrhea, dysentery, etc., they are soon replaced if the disease continues. Becquerel and Rodier, however, found the albumens much reduced in puerperal septicemia, and Schmidt found a marked increase in the concentrated blood of cholera, v. Jaksch," determining the total N of the serum, found slight variations in many acute and chronic diseases, but a well-marked loss in leukemia, pernicious and secondary anemia, and chlorosis. Limbeck, using the same method, found less marked changes in cases of anemia. The ratio of serum albumen to serum globulin, on the other hand, varies much more than does the total quantity of albumen in the serum. The normal limits are placed by Limbeck and Pick, for globulin 16.9-38.3 percent, for albumen 61.7-83.1 percent of the total albu- men of the serum. When such Avide limits may be found in nor- mal subjects it is difficult to attach much importance to very consider- able changes demonstrated in disease, as reported by Estelle, Hoffmann, Halliburton, and Mya and Viglezio. In a considerable series of ob- servations in various diseases Limbeck and Pick succeeded in showing only that serum globulin, the less diffusible principle, is less subject to change than is serum albumen, but they were unable to establish any other general rules. The inorganic salts of the serum include phosphates, chlorides, carbonates, and sulphates, but in what proportion or form the alkalies and earths are combined with these acids is not definitely known. Sodium and potassium are combined especially as neutral chlorides, partly as phosphates and carbonates. Pathological variations in the phosphates of the serum are of slight degree and importance, so far as is known, probably on account of their occurrence in scant quantity and their relative lack of diffusibility. A retention of phosphates in the blood of pathological grade has not been demonstrated. The variations in the chlorides of the serum investigated by Schmidt, Biernacki, and Limbeck have not been found extreme, nor of notable pathological import, although this principle is chiefly responsible for the isotonic relations of cells and serum. A high percentage of chlo- rides is usual in anemias. The specific gravity of the serum in disease varies slightly from the normal limits, 1.025-1.030. (Hamraerschlag, Limbeck.) While Becquerel and Rodier and Hammerschlag found the specific gravity 64 CHEMISTRY OF THE BLOOD. of the serum normal in chlorosis but reduced in secondary anemia, Limbeck found it low in both (1.021-3). In infectious diseases Bec- querel and Rodier found the serum of normal gravity. The color of human serum is yellowish with a slight greenish fluor- escence, but after a hearty meal the increased quantity of fat ruay yield a cloudy or whitish color. The coloring matter of normal human serum belongs to the group of luteins or lipochromes (Hammarsten). It is extracted by alcohol and by ether, yields a blue color with iodine and sulphuric acid, and on spectroscopic analysis causes an absorption of the violet and part of the blue, which is unaffected by reducing or oxidizing agents. Icteric serum causes more complete absorption, be- ginning sharply in the blue. (Limbeck.) When, from lessened resistance of the red cells or abnormal con- ditions of the plasma, red cells are dissolved, the serum contains Hb in solution (hemoglobinemia), as occurs principally in malaria, par- oxysmal hemoglobinuria, septicemia, and with the blood poisons. In jaundice, the serum has a characteristic orange-yellow color and bili- rubin is readily detected. The Globulicidal Property of Serum. — When the serum of dis- eased subjects is added to normal blood the red cells are often rapidly dissolved and the normal color of Hb is replaced by a greenish tint. On spectroscopic analysis, such a specimen shows the presence not of Hb but of hematoidin. Maragliano, who first investigated this property, found marked globulicidal action in the serum of all grades of primary anemia, leukemia, purpura, pneu- monia, malaria, erysipelas, typhoid fever, cirrhosis, and nephritis. Limbeck has noted a greenish color by the naked eye, and demonstrated hematoidin, in the serum of pneumonia, typhoid fever, and purpura hemorrhagica. Buchner demonstrated a relation of the globulicidal activity of serum and its con- tent in salts, finding more marked activity in serum poor in salts and finding that the addition of salts diminishes the effects. Buchner was also the first to call attention to the close relation between the globulicidal and bacteri- cidal properties of serum. Castellino noted the toxic effects of the serum of one animal upon the blood of another, and found that they depended upon globulicidal and coagulative properties. According to Mairet and Bosc, 15 cc. of normal human serum suffices to kill a rabbit weighing two pounds. Heating to 52-58° C. or a slight addition of salt, largely inhibit the coagulative effect, but only partly destroy the toxicity of serum, from which facts, principally, Mairet and Bosc, and Hayem, conclude that the toxicity of serum depends on the presence of an albuminoid body. Castel- ino finds that the globulicidal, coagulative, and general toxic properties vary uniformly in specimens of serum and all are proportionate to the con- ent of nuclein. THE WHOLE BLOOD. Albumens. — The albuminous principles of normal circulating blood include hemoglobin, serum albumin, serum globulin, and fibrinogen. In shed blood a nucleo-proteid exists which is probably derived from the nuclei of leucocytes, probably also from red cells, which combines with calcium salts to form the fibrin-ferment, and which is called THE WHOLE BLOOD. 65 '^ prothrombin." Traces of albumose have been found in the- blood in various diseases, and other nitrogenous principles, occasionally pres- ent and commonly reckoned with the albumens, are urea, uric acid, and xanthin bodies. Varying considerably with the method of deter- mination, the proportion in which these elements exist in the blood has been placed for the entire group by v. Jaksch^ at 22.62 percent ; by Limbeck at about 25 percent; by Schmidt at 10.82-16.63 percent. A relative increase of total albumen is seen, according to v. Jaksch, in diseases accompanied by marked loss of fluids, as in cholera and severe diarrhea, but an absolute increase has not been demonstrated. In infectious diseases the albumens are moderately reduced, even when the number of red cells remains normal (typhoid fever). In nephritis there is usually but not always a reduction. In endocarditis there is little change. In chlorosis, pernicious anemia, and leukemia, the total albumens of the blood are constantly diminished, but in secondary anemia the diminution usually noted sometimes fails, v. Jaksch found his lowest proportion, 8.46 percent, in a case of gastric cancer. Peptone has been demonstrated in the blood of leukemia, by Ludwig, V. Jaksch,^ and Freund and Obermayer, and deutero- albumose by Mat- thes. These observations were made on blood from the cadaver. Devoto and Wagner, who examined blood obtained during life, both failed to find peptone. The suspicion that the peptone previously demonstrated was of postmortem formation was followed up by v. Limbeck in a case of myelogenous leukemia. In the fresh blood no peptone was found, but in a specimen that had stood at room tempera- ture for forty-eight hours, albumose was demonstrated, but its exact character was not determined. The patient died three weeks later and deutero-albuniose was demonstrated in the blood of the cadaver, v. Jaksch states that peptone (albumose ?) is found in leukemic blood only when eosinophile cells are abundant. Freund's observation that peptone is found in the blood in cases of sarcoma but not in cases of carcinoma, has not yet been confirmed. The addition of small quantities of peptone to plasma in vitro re- duces its content of CO^ and its coagulability, and it seems probable that if peptone (or albumose) exists in the circulating blood a similar influence is exerted, intra vitam, by its presence. Quantitative Estimation of Total Albumen of the Blood may be ac- complished by one of two methods. 1. By precipitating the albumens by excess of alcohol, and weighing the dried precipitate. 2. By estimation of total N by Kjeldahl's method and multiplying the result by 6.25 (v. Jaksch). By the first method an uncertain quantity of inorganic matter is car- ried down which cannot be thoroughly removed by washing, and which therefore disturbs the result. In the second method, N is derived from other principles besides albumens, and, as Limbeck and Pick have shown, there may be a dif- ference of 10-20 percent in the results of the two methods. 5 66 CHEMISTRY OF THE BLOOD. Most of the nitrogenous principles other than albumens are soluble in alcohol and some of the sources of error in this method are there- fore removed by estimating the total N of the alcoholic precipitate. V. Jaksch has determined that by multiplying the percentage of N by 6.25 the average percentage of weight of albumen is obtained. Lim- beck and Pick, after investigating the ratio between N and the weight of albumen from which it was derived, concluded that v. Jaksch's fig- ure is sufficiently reliable in the great majority of cases, but that con- siderable inaccuracies may result in the cases of venous stasis, nephri- tis, and especially uremia, in which nitrogenous principles in the blood other than albumens are increased. The determination of total N by Kjeldahl's method is generally em- ployed in estimating the albumens of the blood. The method may be found fully described in Halliburton, " Chem. Phys. and Path.," or Sutton, " Volumetric Analysis." Inorganic Principles of the Blood. The Blood-Ash. — After in- cineration of two specimens of normal blood Schmidt found .84 per- cent and .91 percent of ash. This quantity was increased to 1 per- cent in a case of cholera, but the relation of ash to dry residue remained normal, 3.1-3.5 percent. In the watery blood of nephritis, however, the ash was relatively much increased (6.5 percent). Chemical analyses of the blood-ash in health and disease have shown that pathological importance attaches principally to variations in the chlorides, phosphates, and the iron of the blood. With regard to the chlorides (principally NaCl) the law has been established, that the large?' the proportion of plasma the greater is the percentage of chlorides in the blood. In pneumonia the chlorides are low, probably owing to diminished ingestion, and the effects of an ex- udative process which drains the blood of a considerable quantity of salts. In typhoid fever and erysipelas a reduction usually exists which is neither so marked nor so constant. In nearly all forms of anemia the proportion of chlorides is high, following the rule above stated. Yet Limbeck found normal proportions in two cases of chlorosis, which he refers to diminished salts in the red cells, and Biernacki refers to cases of severe anemia with normal chlorides, which he also explains from the loss of red cells. Becquerel and Rodier found considerable variations, but a normal average in chlo- rides in six cases of chlorosis. From the known influence upon the urinary chlorides of diminished ingestion of food, vomiting, diarrhea, and exudation, it appears prob- able that these factors may to some extent aifect the chlorides of the blood, for in many of the available analyses of the blood-ash in gen- eral diseases, there are numerous contradictory results (tuberculosis, syphilis, cancer). The phosphates exist as neutral or alkaline salts of Na, Ca, Mg, in the plasma and in various combinations (lecithin, nuclein), in the red and white cells. Few facts of importance have been established in regard to the variations in these principles of the blood. I INORGANIC PRINCIPLES OF THE BLOOD. G7 Biernacki demonstrated a considerable diminuton of P2O5 iu anemia, and at the same time a certain parallelism between P2O5 and K in these condi- tions. In leukemia, Freund and Obermayer found an increase in P2O5, and Na, while the K was diminished. The alkalies Na and K are principally combined with CI in the blood, but Na is united in considerable proportions as phosphates and carbonates. The sodium salts, being found principally in the plasma, are subject to marked quantitative variations, being usually increased in watery blood. Potassium being located chiefly in the red cells is diminished in most hydremic conditions, and not alone in scurvy as suggested by Garrod. Sodium carbonate is probably the next most abundant salt of the plasma after sodium chloride, and to this salt mainly the plasma owes its alkalinity and its power to absorb CO2. Iron. — The iron of the blood is principally found in the hemoglo- bin, a compound of albumen and iron containing, in human blood, about .42 percent of iron. It also occurs in traces in the plasma, and is found in relatively large proportion in nuclein. (Hammarsten.) The percentage of iron in normal blood is placed by Limbeck, from the results of several analyses, between .056-.058. Jolles found vari- ations between .0413-.0559 percent and Hladik an average of .0425 percent, using the ferrometer devised by the former. While the principal depot of iron is in the Hb of the red cells, and the quantity of iron is closely propjortionate to the percentage of Hb, the ratio between the tivo is often disturbed, because all iron of the blood does not form colored compounds. A further discrepancy between the Hb percent (after Fleischl) and the iron content of the blood results from the presence of iron-free pigments, as the lutein of Thudicum or the hydrobilirubin of Maly. Thus Biernacki found that direct quantitative estimation always yields more iron than the computation from the percentage of Hb, after Fleischl. Sometimes twice as much iron was found as was to be expected from the Hb percentage. Similar results have been obtained by Jolles and Jellinek. It seems probable that the genei'al introduction of a practical clinical method of estimating the iron of the blood may develop some new facts of interest in the pathology of the blood. Thus Jellinek found in a case of purpura Jiemorrhagica that iron-free pigment was apparently absorbed from extrava- sations, since the Hb registered 50 percent, after Fleischl, but only 38 per- cent as computed from the iron-content. Also in a case of malaria, imme- diately after a paroxysm the Hb sank 10 percent, while the iron remained constant. Estimation of the Inorganic Principles of the Blood. — It falls outside the scope of the present volume to consider the details of inorganic quantitative analysis of the blood. For the estimation of Fe, however, reference should be made to the clinical method and results of Jolles. This method permits the esti- mation of Fe within 15 minutes, and with considerable accuracy. The apparatus is offered by Reichert's agents in this country and is accom- panied by full directions for use. Mackie has described clinical methods of estimating Fe, and phos- 68 CHEMISTRY OF THE BLOOD. phoric acid, in a drop of blood, to which the reader is referred. (Iron, Lancet, 1898, Vol. I., p. 219 ; Phosphoric Acid, Lancet, 1899, Vol. IL, p. 484.) Urea in the Blood. Urea occurs in traces in normal blood (Picard) and in increased quantity in fevers (Gscheidlen), and when, as in nephritis, its excretion by the kidneys is imperfect. In uremia Munzer found .4 percent of urea in the blood, but the amount is regarded as insufficient to cause the toxic symptoms of this condition. Demonstration of Urea in Blood.— v. Jaksch recommends the following procedure. 200-300 cc. of blood are precipitated with 3-4 times as much alcohol, and after 24 hours the precipitate is repeatedly washed in alcohol. The alcohol is then evaporated, the residue treated with nitric acid, and the crystalline mass, secured after some hours, is dried between filter paper, dissolved in water, treated with barium carbonate till CO2 ceases to form, dried on a water-bath, and extracted with hot alcohol. On evaporation urea is found in slender rhombic prisms. If secured in sufficient quantity the crystals may be treated with nitric acid and evaporated, when charac- teristic crystals of urea nitrate form. Or the biuret test may be employed, dissolving the crystals in a little caustic potash, and adding a drop of dilute cupric sulphate. Various other methods may be found described in text-books of physio- logical chemistry, e. g., Gamgee. Uric Acid in the Blood. Scanty traces of uric acid have been demonstrated in normal blood. (Picard, Abeles.) Garrod found .025-. 145 percent during acute attacks of gout, and Salomon also found an increase in acute gout. V. Jaksch regards their methods as inexact. A moderate increase (.08 percent) has been demonstrated in pneumonia and anemia (Sal- omon,^ v. Jaksch ^), in cardiac and other forms of dyspnoea, and in nephritis (v. Jaksch). Klemperer ^ and Weintraud noted moderate increase in leukemia (.09 percent), in nephritis (.06 percent), and in uremia (.19 percent), but failed to find any trace in three healthy subjects and in one case of pneumonia. Estimation of Uric Acid. — Garrod took 10 cc. of serum from 30-35 cc. of coagulated blood, added 10 percent of dilute acetic acid, and obtained crystals of uric acid on a thread placed in the fluid. In blood containing not less than .025 per 1,000 of uric acid the thread was covered with crystals in 24-48 hours. These crystals should be submitted to the murexide test. V. Jaksch recommends the qualitative and quantitative estimation of uric acid by means of the Ludwig-Salkowski method employed in urinary anal- ysis. The blood is prepared for this method as follows : 100-300 cc of blood are diluted 3-4 times with water, heated on the water bath till coagu- lation begins, when it is feebly acidified by acetic acid (s. g. 1.0335). After 15-20 minutes it is filtered, the precipitate washed with hot water, boiled again after feebly acidifying, cooled and filtered. A little sodium phos|»hate is now added to the filtrate, which is then submitted to the Ludwig-Salkow- ski method. For a full description of this method see Simon " Clinical Di- agnosis." GLYCOGEN IN THE BLOOD. 69 Glucose in the Blood. Normal blood always contains traces of glucose, which may be in- creased by diet rich in carbohydrates, and is diminished by muscular exercise and hunger. (Seegen, Chauveau and Cavazanni, v. Mering.) Limbeck found in the blood of two healthy subjects, five hours after eating, .075 pei'cent and .089 percent of glucose. Freund and Trinkler find that glucose is very considerably increased in the blood in cases of carcinoma, readily reducing cupric oxide when freed from albumens, which cannot be accomplished with the blood of any other disease except diabetes. In one case of cancerous cachexia Trinkler found .3 percent of glucose, which reaches the limit stated by Claude Bernard to be capable of producing diuresis. In diabetes the quantity of glucose in the blood is subject to great variations, according to the character and progress of the disease. Hoppe-Seyler found in one case .9 percent. Estimation of Glucose in the Blood. — A weighed quantity of blood is freed from albumeu by boiling with an equal quantity of sodium sulphate and filtering. The precipitate is well washed and the presence and quantity of glucose in the filtrate determined by Fehling's solution or by polarimetry. A certain amount of sugar is carried away by the precipitated albumens, and unless the blood is fresh, the action of the glycolytic ferment may cause serious error in the result. The presence of a glycolytic property in normal blood has long been known, but it has never been shown on what this property depends. Lepine found that it depended upon the red cells more than upon the plasma, and based upon its presence a theory that diabetes results from the diminution or absence of this glycolytic property or ferment. It has been shown, however, that the glycolytic power of diabetic blood is not distinctly inferior to that of normal blood. (Minkowski, Kraus.) The experiments of Seegen and of Arthus indicate that this property is not exerted during life, but is a cadaveric phenomenon, and that the extent of its action depends upon the length of time consumed in the examination of the specimen. Glycogen in the Blood. The long discussion regarding the feasibility of demonstrating gly- cogen in the blood by chemical methods has led to many contrary opinions but in recent years positive results have been more constant. (Salomon, Frerichs, Cramer, Lepine and Barral, Huppert, Czerny.) The microscopical test on specimens dried in the air and stained by iodine, after Ehrlich's suggestion, has given more uniform results. Gabritschewsky, using this method, found intra- and extra-cellular gly- cogen in the blood of both healthy and diseased persons. Extra-cel- lular glycogen, in the form of fine or coarse granules (1-6/^.) he found to be the only form usually present in normal blood ; and it was increased in diseases in which intra-cellular glycogen was abundantly present. He states that extra-cellular glycogen is derived from the disintegra- 70 CHEMISTRY OF THE BLOOD. tion of leucocytes, but offers no evidence to prove this very unlikely orio-in. The glycogen was found in the bodies of neutrophile leuco- cy t*es in cases of diabetes and leukemia, and in the plasma in a con- siderable variety of other diseases. The results of his experiments, injecting sugar and peptone into the blood of animals, indicate that the leucocytes are capable of transforming both sugar and peptone into glycogen, Livierato also found extra-cellular glycogen in normal blood but failed to discover any in the leucocytes in diabetes. From his clinical observations he concludes that the glycogen of the blood is increased in febrile cases with an active exudative lesion and with leucocytosis. In typhoid fever he found only extra-cellular glycogen. Czerny's studies added considerably to the knowledge of the subject. He found the increase of glycogen in exudative processes with leucocytosis. In the blood of children with cachectic leucocytosis he found an almost equal quantity. In 2 or 3 dogs exposed to prolonged cold he found that glycogen appeared in the leucocytes after 24 hours, and persisted for several days. From the severe disturbances of respiration following section of both vagi, or pneumothorax artificially induced, as well as in severe anemia from hemorrhage, he found a marked increase in glycogen. An interesting in- quiry of Czerny's related to the exact chemical nature of the brown staining granules demonstrated by iodine, a reaction common to both glycogen and amyloid. That these brown granules are glycogen is indicated by (1) their reaction to iodine, (2) by the disappearance of their brownish stain on heating (Barfurth), and (3) by their complete digestion by saliva. (Czerny.) On the other hand, Czerny points out that pure glycogen is soluble in water, while the glycogen of leucocytes is insoluble in water (as the writer, also, has found), that iodine with dilute sulphuric acid colors these granules violet, which is the reaction of amyloid ; and finally that in dogs suffering from prolonged suppuration whose blood continually shows abundance of " gly- cogen," the viscera, on the other hand, develop extensive amyloid changes. These observations indicate that the brownish staining granules are not pure glycogen^ but a comparatively insoluble compound of glycogen with some other substance, or else an intermediate product between glycogen and amyloid. Caminer failed to find any glycogen in normal blood, and in disease very rarely found it in the plasma. He distinguished three stages of the deposit in leucocytes, (1) the presence of a light diffuse mahogany stain ; (2) the pi-esence of a few isolated globules ; (3) the complete transformation of the body of the cell into glycogen. In cases of extreme sepsis, all stages were seen ; in pneumonia glycogen was usually present but never in the third stage of its formation ; in phthisis only the first stage was found ; and in rheumatism it was absent. In cases of carcinoma it was found only when suppuration occurred. In four diabetics none was found, but in a case of diabetic coma it was abundant. It was absent in most cases of carcinoma, in chlorosis, in two cases of leukemia and in secondary anemia from hemor- rhage. From experimental studies, he finds that three factors are concerned in the glycogenic degeneration of leucocytes : (1) fever, (2) leucocytosis, (3) toxemia, of which the last is most potent. Further studies are required to reconcile the above conflicting ob- servations and to establish the clinical importance of this form of de- generation of leucocytes. The frequent presence of extra-cellular gly- cogen must at present be doubted, and it must be remembered that other principles such as myelin, lecithin, and amyloid, stain brown with iodine. ACETONEMIA. LIPACIDEMIA. 71 The Diastatic Ferment of the Blood. The study of the property of fresh blood to digest starch has lately been extended to the blood iu various diseases. This property may be demonstrated by adding 1 cc. of blood to 50 cc. of a solution of starch, allowing the mixture to stand a few hours in the thermostat, when glucose may be demonstrated by Fehling's solution. Castellino aud Pracca found that 2 cc. of normal human blood added to 50 cc. of starch solution produced .07 percent of sugar iu twenty-four hours, at 30° C. Arterial blood is more active than venous, and the fermentation, most active between 30-38° C, is inhibited at 75° C, and diminished by slight changes in reaction. (Cavazzani.) Cavazzaui and Pracca found con- siderably increased fermentative power of the blood in anemia, chlorosis, leukemia, pneumonia, malaria, nephritis, cirrhosis, and carcinoma, while in other cases of these and other diseases it was diminished. The principal source of the diastatic ferment is placed in the red cells by Tiegel and Plosz, in the serum by Bial, and in the leucocytes by Castellino and Pracca. The last named observers find that the diastatic property is closely related to the globulicidal and coagulative power of the blood, that it increases after the blood is shed, that it is inhibited by the addition of nucleiu, and increased by sodium sulphate and chloride. Lipemia. The occurrence of free fat (palmitin, stearin, olein) in the blood, both in health and in disease, has repeatedly been observed. While usually present in such small quantity as to be recovered in very small amounts from the ethereal extract, and recognized with difficulty by the microscope, it is sometimes so abundant as to give the blood a milky appearance. (Gumprecht.) Its physiological variations probably exceed the pathological, as it has been found very much increased in healthy individuals after a hearty meal. In disease its occurrence ap- pears to follow no general rule, so that its real pathological significance remains doubtful, v. Jaksch, extracting the blood with ether and thereby including fat, lecithin, cholesterin, and a trace of nitrogenous compounds, found in three cases of diabetes .05— .16 percent, in ne- phritis .1— .5 percent, in typhoid fever .16 percent, and in pneumonia .15 percent. It has also been found in increased quantity in phthisis, poisoning by carbonic oxide, and in fat embolism after traumatism. The presence in the blood of a fat-splitting ferment (lipase) has re- cently been claimed by Hanriot. Acetonemia. Lipacidemia. Deichmuller and v. Jaksch, by extracting the blood with ether and by distillation, have isolated a principle which gives the reaction of acetone, and which they found increased in many processes, especially in fevers. Fatty acids have been found in the blood by v. Jaksch, in diabetic coma, leukemia, acute yellow atrophy of liver, and acute in- 72 CHEMISTRY OF THE BLOOD. \ fectious diseases : /9-oxybutyric acid, in the cadaveric blood of diabetes, by Hougounenq; and sarcolactic acid, in normal blood, by Gaglio, Spiro, and Berlinerblau. Limbeck doubts the reliability of these re- sults, believing that fatty acids may develop from lecithin during the technical procedures followed in isolating these principles. Cholemia. The poisonous symptoms developed in cholemia have been referred by most authorities to the presence of biliary acids. Flint's statement that cholesterin is the poisonous agent has not been accepted, although it has been found in considerable traces in icteric blood. To the naked eye, icteric blood may appear of yellowish-red color, while in the serum or its foam small quantities of bile pigment are readily detected by the peculiar yellowish tinge. On repeated heating the yellowish-red bilirubin changes to the green biliverdin. Diminished isotonic tension and increased resistance of the red cells are a peculiar character of the blood in jaundice. Limbeck found the tension of the cells reduced to .4, .38 and .32 percent NaCl, and that of the serum, .76, and .864 percent NaCl. The well-attested fact that in intense jaundice red cells are frequently dissolved by biliary principles, can, with some difficulty, be reconciled with this markedly hyperisotonic quality of the serum. Limbeck believes that bile-acids affect the union of Hb with the stroma of the red cells, rendering the Hb more easily soluble, and thinks that the solution of red cells in jaundice, as well as in other conditions, depends on other than simple osmotic factors. Other characters of icteric blood are, according to Limbeck's analy- ses, an increase of nitrogenous bodies (3.29, 3.52 percent) ; diminu- tion of chlorides of both blood and serum, which he refers to their displacement by biliary acids ; and a well-marked increase in the vol- ume of red cells (Bleibtreu's method). Detection of Biliary Principles in the Blood. — Well-marked cholemia may be detected by naked eye inspection of serum or its foam. On heating to 50° C. bilirubin may be changed to biliverdin. V. Jaksch has been able to demonstrate bilirubin in the blood when none was to be found in the urine, or when only urobilin existed in the urine, by the following procedure. A little blood obtained by a wet-cup is allowed to coagulate and after one to two hours the serum is drawn off and forced by aspiration through an asbestos filter. The broth of the filtrate is yellow if any bilirubin is present and this yellow tinge becomes green on heating two or three hours at 35° C, if only minute traces of bile pigment are present. Or the blood may be coagulated slowly at 78-80° C. when the serum be- comes greenish in the presence of minute traces of bile. Biliary Acids may be demonstrated by Pettenkofer's method on serum prepared as follows : Albumens are removed by boiling or by alcohol, and the filtrate is treated with lead acetate, and with ammonia, which precipitates the acids with the lead compounds. The acids are then recovered, by washing the precipitate THE SPECIFIC GRAVITY OF THE BLOOD. 73 on a filter, boiling in alcohol, filtering, and decomposing the lead salts by carbonate of soda. The solution is again filtered, evaporated to dryness and the acids extracted by boiling in absolute alcohol. Finally on evaporating the alcoholic extract, biliary acids crystallize out, or an amorphous substance remains from which the crystals may be obtained by extracting with ether. THE SPECIFIC GRAVITY OF THE BLOOD. The specific gravity of the blood changes with the content of water, its most variable constituent, the proportion of salts, which are less variable, and the percentage of albumens, which are the last principle affected by pathological processes. The normal limits have been placed by Becquerel and Rodier between 1.058-1.062 for men, 1.054-1.060, including both sexes; by Ham- merschlag, between 1.056-1.063, including both sexes ; by Lloyd Jones, between 1.045-1.066. Some of these discrepancies are doubtless refer- able to the differences between the methods employed. There are considerable physiological variations in gravity. Accord- ing to Lloyd Jones the blood of newborn infants shows the highest gravity, averaging 1.066 ; after the second week of life up to the second year, the gravity sinks, 1.048-1.050 ; rising with men between 35-45 years, to 1.058, with women after the climacteric, to 1.054; in old age the blood of both sexes approaches the initial high gravity of infancy. Limbeck finds an explanation of these progressive changes in the decreasing capacity of the tissues to absorb water. From 44 estimations on his own blood Schmaltz found minimal variations at diferent hours of the day, from 1.061 from 7-8 A. M., to 1.058, from 2-8 P. M. Muntz found a marked increase, 1.038 to 1.058 in the gravity of the blood in over-fed sheep. Muscular activity, if accom- panied by sweating, slightly decreases, sleep slightly increases the gravity. (Schmaltz, Jones.) Schmaltz found that menstruation is followed by a slight increase of gravity, while the slightly lower gravity of pregnant and parturient women has been frequently observed. For somewhat obscure reasons, the gravity of the blood is increased by residence in high altitudes and a considerable difference (.015) has been observed in the blood of animals pastured on mountain tops and those grazing in the valleys. (Muntz, Viault, Glogner.) In pathological conditions lowered specific gravity is a constant character of the blood in anemia. In chlorosis the change is usually referable to and proportionate with the loss of Hb, but Stintzing and Gumprecht and Siegel have shown that this parallelism is not invariable. In pernicious anemia the specific gravity and dry residue suffer more than the Hb, owing to loss of albumens from the serum. Extremely low gravity, in com- parison with the Hb-content, is characteristic of this condition. In leukemia the gravity is reduced as in simple anemia, but extreme re- ductions are seldom observed, owing to the increase of white cells and the presence of abnormal products soluble in the plasma. In second- 74 CHEMISTRY 01 THE BLOOD. ary anemias there are numerous exceptions to the parallelism between specific gravity and Hb-content. These are found especially in dis- eases in which an exudative process drains the blood of albumens (dysentery), or when from edema there is relative hydremia of serum or red cells. In the infectious diseases the specific gravity of the blood depends upon associated conditions, such as profuse perspiration, diarrhea, ex- udation, etc., more than upon any specific property of bacteria to in- crease the water of the blood. That the power to impoverish the blood varies however with different infectious agents is strongly in- dicated by clinical observation. Grawitz' claim that tuberculin and the diphtheria toxin tend to increase the gravity of the blood, while the toxines of streptococcus and staphylococcus pyogenes tend to diminish it, is frequently confirmed by comparison of the blood in cases of mili- ary tuberculosis and uncomplicated diphtheria with that of septicemia. Although acute stasis is usually followed by marked increase in the gravity of the blood, yet in chronic endocarditis the variations observed are very irregular. General edema is usually associated with lowered gravity. The same observation applies to the blood of nephritis^ normal gravity being observed in many cases of chronic interstitial nephritis (Hammerschlag), while anemia and low gravity of the blood and edema of the tissues are nearly constantly associated with the large white kidney. Numerous studies of the blood of pulmonary tuberculosis, including that of Grawitz, have failed to bring to light any uniform variations in gravity peculiar to the blood in this disease. As a rule extreme reductions are not observed. In the cachexia of malignant neoplasms some of the lowest gravities on record have been observed (1.030, 1.032, Lyonnet), especially in ulcerating and bleeding tumors of the stomach and uterus. In certain skin diseases (pemphigus, eczema, psoriasis, prurigo, mor- bus maculosus) an increased gravity of the blood has been noted by Schlesinger. In cases of general burns Tappeiner, Baraduc, and Schlesinger have found very high gravities, 1.065-1.073, which, in cases that recovered, fell to normal in twenty-four hours. Finally, various drugs appear to exert a moderate but in no sense peculiar influence on the gravity of the blood. Purges, diaphoretics, and diuretics, remove water from the system, and when this loss is not immediately replaced from the tissue the blood is temporarily con- centrated. The action of mercury is somewhat uncertain, but in syphilis it appears to diminish the gravity after a short initial period of increase. (Schlesinger.) To recapitulate, it has been found that considerable variations in the water of the blood may under ordinary conditions be promptly equal- ized by the action of the tissues on the one hand replacing a loss, or of the kidneys and skin removing an excess, so that the change in the blood is very transitory. Only when there is interference with these THE OSMOTIC RELATIONS OF THE BLOOD. 75 processes or wheu the demands upon them are excessive does a more or less permanent alteration in the gravity of the blood follow. On the other hand, when the albumens of the blood are affected, more marked and permanent changes are produced. In general, the content of the tissues in water and that of the blood are very closely interdependent. THE OSMOTIC RELATIONS OF THE BLOOD. AVhen a drop of blood is placed in distilled water, the red cells arc promptly dissolved, but when blood is placed in a solution of salt of a certain concentration, the red cells retain their Hb and sink to the bottom of the fluid. The solution of the cells results from the law of os- mosis, that when two solutions of different concentration are separated by an animal membrane the solutions pass through the membrane, until the quantity of salt in each is equal. The force which leads to this interchange is called " osmotic tension," and two fluids with equal content of salt are said to be " isotonic " or of equal " isotonic tension." Fluids are likewise said to be hyper isotonic or hypisotonic when they contain greater or lesser quantities of diffusible salts than other fluids, and are capable of drawing water from or yielding it to such fluids, according to the laws of osmosis. A solution containing .46 percent of NaCl is just sufficient to prevent the solution of red cells in the average specimen of normal human blood (Limbeck), and the isotonic tension of human red cells may, therefore, be said to be .4.6 percent NaCl. Yet Avhen red cells are placed in a .46- percent solution of salt, they absorb water and swell, although they do not yield up Hb, and when placed in strong solutions of salt, red cells shrink, yielding water to the fluid. Hamburger finds that a .9- percent sohUion of salt causes neither swelling nor shrinkage of the red cells. TJiis solution, therefore, represents the isotonic tension of the blood plasma, and is properly called the "normal salt solution.'' Any considerable lowering of the osmotic tension of the plasma must therefore lead to swelling of the red cells and eventually to their solu- tion. The hyperisotonic quality of the plasma with reference to the red cells is a physiological necessity, otherwise the ingestion of a con- siderable quantity of water would cause the solution of many red cells. The exact limits of osmotic tension between which the red cells suffer no alterations of volume are not known, but it is certain that they are often exceeded in pathological conditions. Changes in osmotic tension which affect the volume of the red cells may occur not only in the plasma but also in the red cells. Thus if the red cells in chlorosis are for some developmental anomaly deficient in salts they would shrink in plasma of normal (.9 percent) osmotic tension, while normal cells would swell in the watery plasma supplied after hypodermoclysis for hemorrhage. The isotonic relations of the blood do not apparently depend en- tirely upon the presence of salts, but are affected also by the presence 76 CHEMISTRY OF THE BLOOD. of Other diffusible principles, as the albumens. Limbeck finds only .2 percent of salts in red cells, yet their isotonic tension is equivalent to at least .46 percent NaCl. Hamburger finds that albumens, phosphates, and chlorides behave dif- ferently under changing osmotic conditions. When a little acid is added to blood, albumens and phosphates pass from red cells to serum, while chlorides pass from serum to cells, but when alkali is added the opposite transfer is induced. Similar physical effects are produced by the passage of oxygen and carbonic acid through the blood and Hamburger suggests that these factors take important part in the metabolic exchanges in the capillaries. One of the chief physiological relations of isoton'iG tension of the blood is its influence in confining Hb to the red cells. Hamburger, who was one of the first to study this subject, regarded the fixation of Hb as the result solely of osmosis on the fluid Hb lying within the membrane of the red cell. Yet the opinion of Hoppe-Seyler that Hb enters into chemical or molecular union with the stroma of the red cell, and the fact that the existence of a membrane about the erythrocyte has not been satisfactorily proven (Limbeck), render this belief uncertain. Limbeck offers evidence to show that there are other than purely physical influences con- cerned with this important function, viz., the chemical union of Hb with other elements of the red cells, and the influence of albumens on osmosis. Physiological variations in isotonic tension of the blood are numerous. That of venous is slightly higher than that of arterial blood (.02 per- cent). The addition of CO, COj, hydrogen, nitrogen, arsenic, or a trace of acid, increases isotonic tension, while oxygen and traces of al- kali diminish it. (Limbeck.) In pathological conditions, from a series of fifty-four observations on blood from venesections Limbeck con- cludes : During acute infections, especially typhoid fever, erysipelas, and pneumonia, the isotonic tension of the blood is frequently much increased, but not always or constantly so. In general disturbances of nutrition the tension of the blood is very variable ; in diabetes and osteomalacia it was normal, in leukemia it was much increased, in jaundice it was low, while in chlorosis it was low, and in severe sec- ondary anemia higher than normal. In pregnancy and lactation Vicarelli found a distinct increase in osmotic tension, i. e., a diminished resistance of the red cells to water (.6-. 66 per- cent NaCl). While Limbeck and Castellino found the red cells less resist- ant in typhoid fever, Bianchi and Mariotti found that experimental injec- tions of B. typhosus lowered the isotonic tension of the blood although filtered cultures of this and other bacteria had the opposite effect. Cavazanni found that injections and inunctions of mercury slightly increased the resistance of the red cells. The tension of the serum has been investigated by Hamburger, using another method, who found no change after bleeding. Viola and lona, during seven hours following venesection, found a moderate diminu- tion, while Limbeck, in three portions of blood, taken at intervals during the exsanguination of a dog, found nearly constant conditions, and Adler found little variation in the tension of the serum in various diseases. ALKALESCENCE OF THE BLOOD. 77 It will thus be seen that the knowledge of this subject is yet in a very rudimentary condition, although its importance in hematology invites further investigation. ALKALESCENCE OF THE BLOOD. It is an established principle of physiology that the capacity of the blood to absorb CO2 depends on its alkalescence. When one compares the results of direct alkalimetry obtained by any of the recognized methods with the proportion of CO^ obtained after the dissociation of carbonates by strong acids, marked discrepan- cies are observed. The volume of CO^ differs greatly from the degree of alkalinity obtained by direct titration. To explain this fact it must be supposed that certain basic properties of the blood are brought into action by alkaliraetric methods which are not active during life, or else that under changing conditions of metabolism the blood is required to absorb varying quantities of CO2, which are by no means necessarily proportional to the capacity of the blood to absorb this element. Thei-e are good physiological grounds for supposing that both of these con- ditions actually exist. It is probable that the existing methods of titration are sensible to the alkaline carbonates and phosphates, which are principally concerned in the alkalinity of the blood, and, as well, to some acid-neutral- izing albuminous principles that are liberated from plasma and red cells during alkalimetric procedures. (Limbeck and Steindler.) To what extent the phosphates and albumens are concerned in the physiological functions centered in the alkalinity of the blood is not known, but it has seemed to the writer that the opponents of the titration method have failed to consider the possibility that other important functions besides that of absorbing CO2 may depend on the alkalinity of the blood {e. g., fibrin formation), and that the results of the titration method therefore deserve recognition in the study of the pathology of the blood. On the other hand, the attempt to measure the alkalescence of the blood by the volume of CO2 recovered after addition of acids involves a needless confusion of the problem, as such estimations include both the CO2 loosely combined with Hb, and that more firmly united with the alkalies. Yet the physiological significance of each of these combinations is probably very different, the former representing the respiratory activity of the blood and the metabolic activity of the tissues, while the latter is related to other less definitely known functions. Moreover, it by no means follows, as has been said, that the volume of CO2 recovered from the blood represents the total quantity that the blood is capable of absorbing, or is any indication of its acid-neutralizing power. Bunge calculates that after allowing for the amount of sodium required to saturate the only strong mineral acid of the plasma (HCl), there is enough sodium left to fix 63 volumes percent of CO2 as carbonate and an equal additional amount as bicarbonate, which is far more than the amount of CO2 actually present in the blood. Again, the attempt to estimate the alkalescence of the blood, considered from either the biological or chemical standpoint, by the content in CO2, in- volves several probable errors. There may, first, be a diminished produc- tion of CO2 in the tissues in pathological conditions. There may be a diminished absorption of CO2 by the blood owing to chemical changes in the tissues or mechanical impediments in the circulation. Finally, there may be simple absorption of COo uncombined with acid neutralizing principles. 78 CHEMISTRY OF THE BLOOD. (Schafer.) Of the actual existence or importance of these factors it is, with the present knowledge of physiology, difficult to judge, but since they stand as uncertain quantities, it seems unwise to rely upon any such very indirect method of judging of the alkalinity of the blood. It would seem therefore that both alkalimetry and the estimation of CO2 furnish important information in regard to the state of the blood, but there are no good a priori reasons to suppose that both measure the same property of the blood, while the practical results obtained positively dis- prove such a view. When laked blood is titrated, a high degree of alka- lescence is obtained, as this method takes account of all acid-neutralizing principles, carbonates, phosphates, and albumens of plasma and red cells. When serum is titrated, the acid-neutralizing principles of red cells are ig- nored, and lower grades of alkalinity are obtained. When the CO2 is esti- mated account is taken only of the carbonates, but albumens and phosphates are ignored, and the presence of any dissolved CO2 disturbs the computation, while the possibility still remains of accidental variations in the ratio between the CO2 actually present and total capacity of the blood to absorb this acid. The studies of the CO^-content of the blood have given the following chief results, as summarized by Limbeck. Venous blood is always richer than arterial in CO2, and both are sub- ject to minor physiological variations (33.37-45.3 volumes percent, Schafer, Kraus). Febrile processes are generally accompanied by diminution of COg in the blood, often in proportion to the height of the fever (34.18-20.9 volumes percent, Geppert). This condition, according to Geppert and Minkowski, is referable to the abnormal production of acid metabolic principles in the blood, a view which is supported by Kraus who, in fevers, along with a diminution of CO2 (10-20 volumes percent), found an increase in the acid principles of the blood. In tlie cachexia of carcinoma Limbeck and F. Klemperer found marked diminution of the COg of the blood (9.67—20.5 volumes per- cent), but did not estimate the acid principles. Similar results fur- nished by other observers indicate that in diabetic coma tiie blood is frequently very deficient in CO,. (Minkowski, Stadelman, Kulz, Hallervorden, Kraus.) Yet Kraus in one case found a normal quan- tity. A simultaneous increase in acid principles has been demon- strated in one case of diabetic coma by Kraus, and on these grounds it has been concluded by various authors that in diabetes there exists an acid-intoxication probably from oxybutyric and diacetic acids. Limbeck accepts this view and concludes from the various studies that in acute febrile infectious diseases, in cancerous cachexia, and in dia- betic coma, a diminution in CO2 and an increase of acid principles of the blood have been fully demonstrated, pointing in all probability to the existence of an acid-intoxication in these diseases. In leukemia Kraus ^ found a slight deficiency of CO., (20.29 volumes percent). From the direct alkalimetry of the blood or serum much less uni- form results have been obtained. A fairly constant normal alkalinity has been established with titration methods, v. Jaksch ^ placed the alkalinity of normal blood at .26-.30 percent NaOH ; Kraus, at .226 percent ; Jeffries, at .2 percent ; Drouin, at .206 percent ; Freudberg, BIBLIOGRAPHY. 79 at .2-. 24 percent ; Limbeck, .220— .256 percent. In laked blood it has been found much higher by Loewy, .449 percent ; by Berend, .45-.5 percent. Peiper found slightly greater alkalinity in the blood of women than in that of children, and in the blood of men over that of women. A constantly diminished alkalinity in the blood during fevers has been reported by the above authors, by Rumpf, and others, and in carcinoma (Rumpf), anemia (v. Jaksch, Peiper), leukemia (Peiper, Rumpf), uremia, cirrhosis, and osteomalacia (v. Jaksch). On the other hand, the results obtained by Loewy and by Limbeck and Steindler, using their particular methods, were extremely variable, and these variations were about equal in health and disease. Further observations are therefore necessary to determine the true significance of the changes demonstrable by titration methods. THE ACIDITY OR BASIC CAPACITY OF THE BLOOD. There are certain unsaturated salts in the blood (NaHCOg, NaH^PO^ and probably Na2HPO^, Limbeck), which although alkaline to litmus are acid to phenolphthalein, and are capable of uniting with bases. While fresh blood is alkaline, serum reacts as acid to phenolphthalein. The capacity of the blood salts to neutralize bases has been called its basic capacity, by Kraus,^ who devised a delicate and somewhat diffi- cult method of measuring this capacity. In normal venous blood Kraus found a basic capacity of .162-.232 percent NaOH, which increased in febrile conditions to .209-.272 per- cent NaOH. His demonstration of a marked increase of basic ca- pacity (.347 percent) in diabetes, has been regarded as strong evidence of an acid intoxication in this disease. Bibliography. Chemistry. Abeles. Wien. med. Jahrbiicher, 1887, p. 479. Adler. Cited by Limbeck, p. 165. Arthus. Arch, de Physiol., T. 24, p. 337. Baraduc. Union Med., May, 1863. Barfurth. Arch. f. micr. Anat., Bd. 25, p. 305. Beal. Pfliiger's Archiv, Bd. 54, p. 72. Becqiierel, Bodier. Untersuch. u. d. Zusammenset d. Blutes in gesund. u. krank. Zustiinde, Erlangen, 1845. Berend. Zeit. f. Heilk., 1896, p. 351. Berlinerblau. Arch. f. exper. Path., Bd. 23, p. 333. Bianchi, Mariotti. Wien. med. Presse, 1894, No. 36. Biernacki. Zeit. klin. Med., Bd. 24, p. 460. Buchner. XI. Cong. inn. ^led., 1892. Bunge. Cited by Schafer, Text-Book of Physiol., 1898, p. 157. Caminer. Deut. med. Woch., 1899, No. 15. Ca&telUno. ' II Morgagni, Ann. 37, May-June, 1895. ^Gaz. d. Osped., 1891, No. 20. Castellino, Pracca. II Morgagni, Aug., 1894. Cavazanni. Arch. p. 1. sci. med., T. 17, No. 6. Chauveau, Cavazanni. Cent. f. Physiol., VIII., p. 22. Chittenden, Albro. Amer. Jour. Physiol., 1899, p. 291. Chvostek. U. d. Wesen d. parox. Hemoglobinfemia, 1894. Cramer. Zeit. f. Biol., 1888, p. 97. 80 CHEMISTRY OF THE BLOOD. Czemy. Arch. f. exper. Path., 1893, Bd. 31, p. 190. Deichmulkr. Cited by v. Jaksch, Clinical Diagnosis. Devoto. Revista Clin., T. 30. Brouin. These de Paris, 1892. Ehrlich. Zeit. f. klin. Med., 1883, p. 33. Estelle. Rev. mensuelle, 1880, p. 704. V. Frerichs. Zeit. f. klin. Med., 1885, p. 33. Freund, Obermayer. Zeit. f. physiol. Chem., Bd. 15, p. 310. Freund. Wien. med. Blatter, 1885, No. 9. Freiidberg. Virchow's Archiv, Bd. 125, p. 566. Gabritschewsky. Arch. f. exper. Path., Bd. 28, p. 272. Gaglio. Arch. f. Physiol., 1886, p. 400. Oamgee. Text-book of Physiol. (Schafer), 1898, I., p. 185. Garrod. Med. Chir. Transac, 1854, p. 49. Geppert. Zeit. f. klin. Med., Bd. 2, p. 364. Gescheidlen. Stud. u. d. Ursprung d. Harnstoffs im Thierkorp, Leipsic, 1871. Glogner. Virchow's Archiv, Bd. 126. Gmwitz. 2Deut. med. M^och., 1893, No. 51. Giimprecht. Deut. med. Woch., 1894, p. 756. Halliburton. Text-book of Physiol. Chem., 1893. Hallei-vorden. Arch. f. exper. Path., Bd. 12, p. 237. Hamburger. 'Arch. f. Anat. u. Physiol. Phvs. Abt., 1886, p. 476^1887, p. 31. Zeit. f. phys. Chem., 1890, VL, p. 319. Virchow's Archiv, Bd. 140, p. 50.3. 2 Cent, f. Phys., Bd. 9, No. 6. Hammerschlag. ^Zeit. f. klin. Med., Bd. 21, p. 472. 2 Wien. klin. Woch., 1890, p. 1018. Zeit. f. klin. Med., Bd. 20. Hammarsten. Physiol. Chemie. Hanriot. Compt. Rend. Soc. Biol., 1896, p. 925. Hayem. Compt. Rend. Soc. Biol.. 1894, p. 227. Hladik. AVien. klin. Woch., 1898, No. 4. Hoffman. Arch. f. exper. Path., Bd. 16, p. 133. Hoppe-Seyler. Physiol. Chemie, 1881. Hougounenq. Maly's Jahresber., 1888, XVII., p. 430. Huppert. Zeit. f. phys. Chem., Bd. 15, p. 335, 546. V. Jaksch. 'Zeit. f. klin. Med., Bd. 24, p. 429. 2 Ibid., Bd. 23, p. 187. "Zeit. f. physiol. Chem., Bd. 16, p. 243 and Clinical Diagnosis. * Zeit. f. klin. Med., Bd. 11, p. 307. 5 Ibid., Bd. 13, p. 353. Jeffries. Boston Med. and Surg. Jour., 1889. Jellinek. Wien. klin. Woch., 1897, No. 47. Jolles. Deut. med. Woch., 1897, No. 10. Judell. Ptluger's Archiv, Bd. 6-7. F. Klemperer. Charite Annalen, Bd. 15, p. 151. Klemperer. ' Deut. med. Woch., 1895, No. 40. Kossel. Zeit. f. phvsiol. Chem., Bd. 22, p. 74. Kraus. » Zeit. f. klin. Med., Bd. 21, p. 315. 2Zeit. f. Heilk., Bd. 10, p. 106. Arch. f. exper. Path., Bd. 26. Kulz. Zeit. f. Biol., Bd. 20, p. 165. Lepine, Barral. Comp. Rend. Soc. Biol., 1891. Lepine. Lyon Med., 1890, No. 3. Liehtheira. Volkmann's Vortrage, No. 134. Lilienfeld. > Zeit. f. Phvs. Chem., Bd. 18, 473. 2 ibid., Bd. 20, p. 155. Limbeck, Pick. Prag. med. Woch., 1893, Nos. 12-14. Limbeck, Sleindler. Cent. f. inn. Med., 1895, No. 27. Livierato. Arch. f. klin. Med., Bd. 53, p. 303. Lloyd Jones. Jour, of Physiol., Vol. 8, p. 1. Loeuy. Arch. f. d. ges. Physiol., Bd. 58, p. 498. Cent. f. med. Wissen., 1894, p. 785. L'owit. Studien u. Phys. u. Path. d. Blutes, Jena, 1892. Ludwig. Wien. med. Woch., 1881, p. 122. Lyonnet. De la densite du Sang., Paris, 1892. Mairet, Bosc. Compt. Rend. Soc. Biol., 1894, pp. 487, 543, 568, 586. Maragliano. XI. Cong. inn.'Med., 1892. Berl. klin. Woch., 1892, No. 31. Matthes. Berl. klin. Woch., 1894, Nos. 23, 24. Mering. Pfluger's Archiv, Bd. 37, p. 348 ; Bd. 39, p. 21. Meyer, Biarnes. Compt. Rend. Soc. Biol., 1893, p. 821. Miescher. Med. Chem. Untersuch. (Hoppe-Seyler), p. 411. P BIBLIOGRAPHY. 81 Minkowsk-y. ' Berl. klin. Woch., 1892, No. 5. 2 Arch. f. exper. Path., Bd. 19, p.. 233. 3 Ibid., Bd. IS, p. 33. Muniz. Compt. Rend. Acad. Sci., T. 112, p. 298. Miinzer. Cited by v. Jaksch.'' Murri. Klin. Zeit. n. Streitfragen, Bd. 8, H. 7. 3Iya, Viglezio. Ri vista clin., 1888. Nencki, Sieher. Bei'. d. Deut. cliem. Gesell., 1885, Bd. 18, p. 392. Peiper. Virchow's Archiv, Bd. 116, p. 337. Picard. Virchow's Archiv, Bd. 11, p. 189. Rumpf. Cent. f. inn. Med., 1891, p. 441. de Ruyter. Arch. f. klin. Chir., Bd. 37, p. 766. Salomon. ^ Zeit. f. phvs. Chem., Bd. 2, p. 65. Charity Annalen, 1880, V., p. 137. 5 Deut. med. Woch., 1877, pp. 92, 421. Schlesinger. Arch. klin. Med., Bd. 50, p. 407. Schmaltz. Deut. Arch. klin. Med., Bd. 47, p. 145. Deut. med. Woch., 1891, No. 16. Schmidt. Zur. Charact. d. epideni. Cholera, Leipsic, 1850. Schaffer. Sitzb. kais. Acad. Wissen. Wien., Bd. 41, Abt. 3, p. 589. Seeaeu. i Ptiuger's Archiv, 1884, 1885, 1886. ^ wien. klin. Woch., 1892, Nos. 14-15. Siigl. Prag. med. Woch., 1892, No. 20. Wien. klin. Woch., 1891, No. 33. Silberman. Virchow's Archiv, Bd. 119, p. 88. Spiro. Zeit. f. physiol. Chem., 1887, I., p. 110. Stadelman. Arch. f. exper. Path., Bd. 16, p. 419. Stiiitzing, Gumprecht. Deut. Arch. klin. Med., Bd. 53, p. 265. Tappdner. Cent. f. med. Wissen., 1881, Nos. 21-22. Tiegel, Plosz. Pfliiger's Archiv, Bd. 6, p. 249 ; Bd. 7, p. 391. Tschirkoff. Zeit. klin. Med., Bd. 19, SuppL, p. 87. Trinkler. Cent. f. med. Wissen., 1890, p. 498. Viault. Compt. Rend. Acad. Sci., T. Ill, p. 97. Vicarelli. Cited by Limbeck, p. 165. Viola, Jona. Arch. d. Physiol., VIII., 1, p. 37. Wagner. Cited by v. Jaksch." Weintraud. Deut. med. Woch., 1895, V. B, p. 185. CHAPTER III. MORPHOLOGY AND PHYSIOLOGY OF THE RED CELLS. In fresh normal blood the red cells appear as homogeneous bicon- cave disc-shaped bodies, with opaque yellowish rim and nearly trans- parent center. They show a strong tendency to cohere to one another by their flat surfaces, forming long rows (rouleaux). This property has been referred to the presence of a fatty material surrounding the red cells. In the capillary circulation they exhibit remarkable elas- ticity, folding, indenting, and greatly elongating, without rupture. After early infancy they are invariably non-nucleated. In dry specimens when thinly spread and rapidly dried they are cir- cular in outline, their concavities are obliterated and they stain evenly throughout. When rather thickly spread the rouleaux are retained, the concavities persist, and the centers may be nearly transparent while the thicker rims are densely stained. Neither membrane nor reticulum have been fully demonstrated in the human red blood cell, although both, from analogy, have been sup- posed to exist. It is more probable that the hemoglobin is held in compact form by a chemical union with other albuminous constituents of the cell. (Schafer.) Staining Reactions. — The living red cell is achromatic, but when fixed, is markedly acidophile in quality, a ])roperty ])robably residing exclusively in the Hb. In the center of the cell there is an achro- matic constituent which fails to stain with acid dyes, but under some circumstances this central substance may become partially separated or even extruded from the cell, when it stains lightly with methylene- blue, and yields the specific reaction of chromatin (Nocht's method). Two views regarding the structure of the red blood cell have been main- tained. According to the first the red blood cell is vesicular, consisting of colored semifluid contents, hemoglobin, surrounded by a membrane and supported by a stroma composed of various substances, principally lecithin, cholesterin, and cell-globulin. (Schafer.) According to the second view, the red cell is not vesicular but a viscous solid mass, consisting of a colorless elastic stroma v/hich is condensed at the periphery and which supports, i»artly by mechanical means and partly by chemical union, the hemoglobin and other constituents of the cell. (Rollett, Brucke.) It is not impossible to partly reconcile these opposing views, regarding the peripheral condensation of stroma as in many respects similar to a mem- brane. That the cell is not strictly vesicular with separable fluid contents is shown by the fact that it may be subdivided without escape of contents, each fragment assuming the discoidal shape. PLATE I. Normal Blood. (Triaeid Stain.) Fig. I. Normal red cell, flatly spread, and evenly stained. Fig. 2. Normal rouleau. Fig- 3. Normal red cells, varying slightly in size, thickly spread, showing central clear areas. Fig. 4. Normal red cell, of slightly altered shape. Fig. 5. Lymphocyte, medium size. Fig. 6. Large mononuclear leucocyte, incurved nucleus. Fig. 7. Polynuclear neutrophile leucocyte. Fig. 8. Eosinophile leucocyte. Separate nuclear lobes. I SIZE OF RED CELLS. 83 According to Foa the hemoglobin is limited to a peripheral layer of (iiauules under which lies homogeneous cell protoplasm, and this view is sup- ported by the occasional escape in pathological conditions of a central achro- matic substance leaving the hemoglobin more or less intact. Size of Red Cells. — In normal subjects, the average diameter of red CI lis varies between 7 and 8 /Jt, depending considerably upon the method employed in preparing the specimen. The following table compiled by Limbeck shows the results obtained by different writers : Limits. 4.5-9.5 fi Welcker Valentin Malinin Hayem 6 -8 Malassez Laache 6 -9 Bizzozero 7.0-7.15 Gram I 6.7-9.3 1 7.7-8 Average. 7.2-8.1/1. 7.0 7.7 7,5 7.6 8.5 From 75 to 90 percent of the cells fall within average diameters. Sex has no influence upon the size of red cells, and after infancy the diam- eters remain constant. In normal infants' blood the variations in the size of the cells are considerable, 3.3-10.3 fi. (Hayem.) Ill path^lojlc'd coivUtion-t, variations in the size of red cells are one of the most important of morphological features. In general, varia- tions in the size of the cells indicate a severe and chronic anemia, while in mild and in very acute anemias, the diameters are usually le altered. MiCROCYTES are cells distinctly under the normal size and may show a diameter of 4 to 1 /i. Several influences appear to be concerned in the formation of mi- crocytes. It is probable that cells poor in Hb, lying in plasma of high gravity and osmotic tension may be reduced in size, from the direct loss of substance and the concentrating influence of a fluid of high os- motic tension. These conditions are present in some cases of chlorosis, but the attempt to show that the cells in this disease are uniformly undersized has not been successful. Many microcytes probably result from the splitting of portions of degenerating and ameboid red cells, as in pernicious anemia. The occasional occurrence of nucleated red cells of small or minute size indicates that some microcytes arise by division of undersized mother-cells. This mode of origin is probably limited to pernicious anemia. The clinical significance of the presence of undersized cells or dis- tinct microcytes is rather indefinite. In some cases of chlorosis the cells appear to be uniformly smaller than normal, but this character is not invariable. In some cases of chronic anemia of pernicious grade, the majority of cells are undersized and deficient in Hb. In primary pernicious anemia, microcytes with increased or diminished Hb are constantly present. Tallquist believes that the presence of many mi- crocytes is an indication of rapid destruction of blood. 84 MORPHOLOGY AND PHYSIOLOGY OF THE RED CELLS. Megalocytes are cells distinctly larger than normal. They usu- ally vary between 10 and 20 ft in diameter. Most of these cells form from the division of the large nucleated red cells which appear in red marrow in severe chronic and especially in pernicious anemia. In addition to this mode of origin it is probable that many slightly en- larged cells and megalocytes are produced by the swelling which, under the laws of osmosis, results when red cells lie in plasma of low gravity and low osmotic tension. Megalocytes may show an excess or a deficiency of Hb, and frequently exhibit a brownish stain after eosin. The presence of megalocytes indicates an anemia of severe grade and usually of rather long standing. Their appearance denotes regenerative activity on the part of the blood. (Tallquist.) Some authorities arbitrarily suggest that when megalocytes form more than 10 percent of the cells the case should be classed as pernicious anemia. In secondary pernicious anemia the megalocytes are frequently defi- cient in Hb, while in primary pernicious anemia they usually contain an excess of this principle. Nucleated Red Blood Cells. — While the presence of nucleated red cells in the blood at any period of extra-uterine life may be considered pathological, it must be admitted that prolonged search in the blood of infants shortly after birth will usually disclose the presence of a few of these cells. Normoblasts are nucleated red cells of normal size and character (7—9 //). They usually show a normal or slightly increased quantity of hemoglobin, fail to show a biconcave form, do not cohere in rouleaux, and exhibit a central densely staining spheroidal nucleus occupying one-third of the diameter of the cell. Much less frequently the nucleus is subdivided into two or three lobes, more or less closely con- nected, while mitotic figures in normoblasts have several times been described. (Luzet, Troje, Askanazy.') Normoblasts are most commonly seen, in very variable numbers, in the milder forms of anemia, chlorosis, acute anemia from hemorrhage, early leukemia, etc. They may be very abundant in the anemia of children, especially in v. Jaksch's type. In severe anemia they are usually associated with larger nucleated red cells. In children any severe disturbance of the circulation, especially if accompanied by af- fection of the bone marrow, may cause the appearance in the blood of a few normoblasts. Severe leucocytosis in children and even in adults may draw a few normoblasts into the circulation. In anemic subjects, ante-mortem leucocytosis may be of this character. In chlorosis, a periodical increase of normoblasts, lasting 4-6 days, repeated at inter- vals apparently favored by rest in bed and successful medication, and accompanied by a marked increase in red cells, has been described by V. Noorden, under the term " blood crisis." This phenomenon chiefly has given rise to the belief that the appearance of normoblasts is of favorable import in anemia. This opinion may be accepted, not be- cause the normoblasts appear in the circulation, but because their ap- pearance indicates active reproduction and discharge of red discs from PATIIOLOQWAL CHANGES IN RED BLOOD CELLS. 85 the marrow, and demonstrates that the type of blood formation is physiological. Megaloblasts are nucleated red cells of larger than normal dimen- sions (10—20 fj.). When exceeding 20 /j. in diameter, they are usually called gigantoblasfs. They are usually circular in outline, but are readily deformed in spreading. Megaloblasts may show a deficiency of hemoglobin, but usually contain an excess, in which case they are apt to stain brownish with eosin (polychromatophilia). The nuclei of mega- loblasts show a great variety of forms. They may be found in : (1) the ordinary vesicular form of the resting stage with intranuclear net- work and nodal thickenings, but without nucleoli, or they may be (2) pyhiomorpliotts consisting of one or several densely staining lobes, either distinctly separate or grouped in irregular rosette form ; or (3) they may exhibit phases of normal or pathological mitosis, with irregu- lar wreaths or two or three unequal asters ; or (4) they may show many stages of karyorhexis, with fragmentation, vacuolation, fading of segments of nuclei, as well as minute subdivision into fine basic stain- ing particles widely scattered in the cell. (See Plate V.) The significance of the presence of megaloblasts in the blood depends on the conditions with which they are associated. Megaloblasts are seen in smears of the marrow of many young in- fants, and when seen in the circulation of such subjects in small num- bers and with a majority of small nucleated cells, they have the same significance as normoblasts. In cases of severe anemia of adults a few nucleated red cells of moderately large dimensions are sometimes seen with a larger number of normoblasts, when no special significance can be attached to their discovery. When the majority of a considerable number of nucleated red cells are megaloblasts, especially if gigantoblasts are present and unequal mitotic figures are observed, the diagnosis of primary pernicious anemia is justified. The presence of a few megaloblasts alone is usually found only in primary pernicious anemia, but may occur in pernicious secondary anemia. Possibly the discovery of a single gigantoblast may warrant the diagnosis of primary pernicious anemia. In general, the appearance of megaloblasts in the blood indicates that a pathological type of blood formation has been established in at least a part of the red marrow. Whether this change represents a re- version to the embryonal type of blood formation, as suggested by Cohnheira, cannot yet be determined, and will be considered later. Probably the low osmotic tension of the watery plasma may be re- sponsible for some of the increased size of the nucleated red cells of pernicious anemia. Pathological Changes in Red Blood Cells. Simple loss of Hb is one of the commonest changes occurring in red cells, this being the chief cellular lesion in chlorosis and mild sec- ondary anemia. When deficient in Hb the cell exhibits an en- 8G MORPHOLOGY AND PHYSIOLOGY OF THE RED CELLS. kirged central clear area of variable outline. In extreme cases the cell may be reduced to a very thin ring of Hb, inclosing a wide and per- fectly transparent central area. In dry specimens, such cells when thinly spread and rapidly dried may fail to exhibit a marked central clear area, but the whole cell shows a uniform pallor. Occasionally the central clear area is irregular, owing either to ameboid changes in the red cell or to the unequal loss of Hb in different parts of the cell, or, frequently, to artificial influences. Thus may arise minute clefts and the heaping of remnants of Hb in the center of the cell. (Plate YI.) Poikilocytosis is a term applied to the appearance in blood of misshapen red cells. Probably the chief mode of origin of poikilo- cytes is through the effects of ameboid motion of a portion or the whole of the cell, which has been described as among the phenomena occurring in degenerating red cells. A great variety of shrunken and misshapen cells are produced artificial/ ij in dried specimens of anemic blood. The true poikilocyte is usually a pear-shajjed cell with a short globular projection at one or more poles. Extremely irregular cells usually result from shrinkage or trauma. Very small cells may be formed by the separation of fragments of poikilocytes. (Ehrlich.) Poikilocytosis is an indication of a severe anemia with degenerative changes in the red cells. Ameboid motion of entire cells or of fragments of cells, or of central colorless masses in necrobiotic cells may be observed, in slight degree, in many severe anemias. This phenomenon, which is more commonly observed in the blood of the low vertebrates, is probably respon- sible for the production of many poikilocytes, and for the subdivision of red cells, but is otherwise lacking in special significance. (Cf. Askanazy.) Variations in size of red cells is an important feature of degenerative processes which is considered elsewhere. In cases of acute poisoning in man and animals, a considerable va- riety of degenerative changes in red cells has been described by many writers, but these do not yet admit of accurate classification. Among them may be mentioned : (1) the liemoglobineniic der/cneration of Ehrlich, which consists in the appearance of droplets of Hb within the cell ; (2) the subdivision of the cell with formation of microcytes, as seen in some forms of poisoning (Litten) and after severe burns ; (3) the specific re- action of diabetic blood described by Bremer ; and (4) the changes in color to be noted after poisoning by hydrochloric acid, hydrocyanic acid, chlorate of potash, and some coal-tar products. The malarial parasite produces a peculiar series of degenerative changes in the red cell which will be considered later. Anemic or Polychromatophilic Degeneration. Polychromasia. Polychromatophilia. — Under the term polychromatophilia two ap- parently distinct conditions have been described. The term is applied by Ehrlich ^ to the appearance, especially in cases of chronic anemia, of cells which, after hematoxylon and eosin. ANEMIC OR POLYCHROMATOPHILIC DEGENERATION. 87 instead of staining light red take a bluish red or violet tinge, or, in extreme examples, even take a deep blue stain. These various abnor- mal tints are referred by Maragliano, Castellino, and Ehrlich, to a pro- gressive coagidat'ion-necrosis of the cell which thereby loses its power to retain Hb and gradually yielding this element to the plasma, loses also its normal affinity for acid dyes. This opinion is based on the following grounds: (1) The appearance of cells showing an advanced stage of the change, in which the frequent breaks in contour " convince every practiced eye that the cell is approaching solution." (2) By the abundance of such cells in starving animals where new formation of red cells is not to be expected, and (3) by their abundant presence within 23 hours after hemorrhage, when normoblasts are only excep- tionally seen, and (4) by the constant polychromatophilic tendency of megaloblasts as compared with normoblasts. Gabritschewsky, Smith, Askanazy,^ Engel, Dunin, and many others believe that polychromatophilic discs are not degenerating hut on the contrary are the youngest red cells of the blood. This view is supported by the abundance of such cells shortly after severe hemorrhages and in many other conditions where red cells are being rapidly formed. As- kanazy found polychromatophilia of all the nucleated red cells in the marrow of a rib just after its excision and before degenerative changes could have set in, and states that a large proportion of cells in the liver of the seven months' fetus are polychromatophilic. Walker also found as high as 34.75 percent of basophilous red cells in the blood of a fetal puppy, while in the marrow of various healthy domestic animals they varied from 12 to 62 percent. Engel has shown that there is a physiological polychromasia be- longing to a portion of the red cells of the embryo, and Ehrlich ^ accepts this fact without yielding his opinion that, in the adult, polychromasia is a sign of degeneration. According to Engel, poly- chromatic cells appear when rapid new formation of red corpuscles is demanded, and when there is not time for the complete shrinkage of nucleus and development of orthochromatic cells. The bearing of the first studies of Maragliano and Castellino is here of importance. These observers found that normal blood in carefully prepared fresh specimens begins to show distinct alterations of the red 'cells after standing 10-12 hours. These consist in widening of the central colorless area, appearance of ameboid activity in this area, for- mation of poikilocytes and development of a basic staining quality in a part or the whole of the cell. In various pathological conditions they found these changes present immediately after shedding and the more rapid their appearance the graver was the general condition of the patient. The authors belie v^e that these changes may occur in the cir- culating blood ; that they are referable to decreased stability of cells and increased globulicidal activity of plasma ; and that they affi^rd val- uable clinical evidence of destruction of blood. They depict cells which show the polychromatophilic tendency of Ehrlich, and other cells in ivhich there is the formation of a central mass lohich stains bluish by meth- 88 MORPHOLOGY AND PHYSIOLOGY OF THE RED CELLS. ylene-blue ^ and which has been shown to herald the extrusion of a por- tion of the red cell in the form of blood plates. (See Plate II., Fig. 2.) The processes represented by these two kinds of altered cells maybe re- lated but are certainly not identical, and since the chief evidence in favor of the degenerative nature of Ehrlich's polychromatophilic cells is found in the studies of Maragliano, it would seem that further investigation is necessary before one can accept the view that such cells are necrotic. Finally, it should be mentioned that Sherrington refers the brownish staining reaction frequently seen in red cells in severe anemia to in- complete oxidation of Hb. To summarize the evidence in this field, it appears that the term poly- chromasia should be limited to a diffuse brotvnish staining qnality of the red cells occurring in various forms of anemia, and abundantly in normal marrow, the significance of which is undetermined. The appearance of bluish staining granules and areas in red cells, often described under this same term, should be classed with Graivitz' granular degeneration of red cells. (Cf Plates Y., Fig. 6, and VI., Figs. 4, 5.) Granular Degeneration of Red Cells (Grawitz). Punctate Ba- sophilia of Red Cells. — The attention of pathologists has long been attracted by the presence of bluish staining granules and masses in red cells and they have repeatedly been interpreted as remnants of a nucleus. In specimens stained by the Nocht-Romanowsky method these masses usually give the reaction of chromatin. Lately the for- mation and significance of these bodies have been fully considered by Maximow, who calls them " nucleoids " and whose conclusions regard- ing their nuclear character can hardly be doubted. In 1893 Askanazy ^ described the appearance of many fine baso- philic granules in the bodies of nucleated red cells in pernicious anemia, regarding them as evidence of karyorhexis. In 1894 Schumann saw such cells in the anemia of hothriocephalus, and in 1896 Lazarus de- scribed them in 20 cases of pernicious anemia, while failing to find them in other diseases. Later they were described in pernicious anemia and variously interpreted by Klein, Zenoni,^ Lenoble, and Grawitz.^ Very similar changes have been described in secondary anemia, normal blood, in birds' blood, and in that of the human embryo by Askanazy, Gabritschewsky, Klein, Pappenheim, and Engel.^ Plehn found them abundantly in the blood of malarious subjects and regarded them as remnants of the parasite. Grawitz has recently described the appearance and studied the nature and occurrence of such cells in human blood, concluding that they represent a specific form of degeneration. He did not find any evi- dence of increased karyolysis in the normoblasts of marrow when de- generating cells were found in the blood, except in very severe anemia. The exact origin of the bluish staining granules he is unable to explain, but thinks they have no connection with a process of karyolysis. In pernicious anemia and leukemia Grawitz could always find many cells •Ehrlich warns against the use of methylene-blne in demonstrating polychromato- philia, using only the triacid mixture, or heraatoxylon and eosin. ( Die Anemie, p. 34. ) THE NUMBER OF RED BLOOD CELLS. 89 in granular degeneration, but in 11 cases of severe chlorosis they were missing. In carcinoma they were present when the tumor was in such a situation " as to favor the absorption of toxic substances." Thus they were present in 10 cases of gastric and esophageal cancer, but absent in two cases of uterine carcinoma. They were very abundant in chronic lead poisoning, as has also been noted by Behrendt. They were not found in numerous cases of tuberculosis and syphilis. The writer has frequently noted these changes in red cells in per- nicious anemia, and is convinced that in megaloblasts they often result from karyorhexis, having seen many transitional stages of the frag- mentation and fine subdivision of these nuclei. In the hydremic blood of malarial cachexia they are extremely common. In some cases, after overstaining in methylene-blue, nearly every red cell shows such bluish points, sometimes connected with fine threads. The Numbers of Red Blood Cells. — It is hardly necessary to point out that the present method of enumerating the red cells in a unit of volume of blood gives an imperfect estimate of their total number in the body. Yet the conclusions that are constantly being drawn from such information could be justified only in case there were a uniform relation between the number of cells per cubic millimeter and their total number in the body. The considerations mentioned in the Introductory Section show that in pathological conditions no such relation uniforndy exists. On the other hand, since the total quantity of blood in healthy subjects is properly adapted to each individual, the bulk is of less practical importance than the proportion in the unit of volume, and the enumeration of red cells in tolerably healthy cases, therefore, warrants the same conclusions that could be drawn from the estimation of their total numbers in the body. The original estimates of Welcker (1854), that in the cubic milli- meter of blood there are normally 5,000,000 red cells in men and 4,500,000 in women, have not been seriously disturbed by many sub- sequent observations by means of Malassez' method. Perhaps the chief contribution of later observers using Thoma's instrument has been the proof that the numbers are more apt to exceed rather than fall below these averages, especially in men, a fact that has become more certain from the more careful estimates of the last decade. Thus the average obtained bv Helling was 5,910,000 ; by Fredrichson 5,072,- 000 fZaslein, 5,010,000 ; Neubert, 5,603,000; Graber, 5,081,000; Stierlin, 5,752,000; Heinicke, 5,209,667; Andriezen, 6,000,000; Hayem, 5,500,000. The variations in the number of red cells referable to physiological and accidental causes are so numerous and distinct as to require their consideration in detail. 1. Time of Day. — The variations found by some observers at different hours of the day may probably, as Limbeck suggests, be re- ferred to the influence of digestion and exercise. Reinert found a tendency toward diminution of red cells and increase of leucocytes as the day progressed with healthy subjects. 90 MORPHOLOGY AND PHYSIOLOGY OF THE RED CELLS. 2. Digestion. — With but few reports to the contrary, the observa- tions of Yierordt and Duperie, Reinert, and Limbeck, indicate that within one-half to one hour after the ingestion of a full meal the pro- portion of red cells begins to diminish, falling usually 250,000-750,- 000 per cubic mm. The numbers diminish for 2-4 hours and then gradually approach the normal. The percentage of Hb falls pro- portionately. These eifects are more prominent after a largely fluid meal, and are probably referable to the absorption of fluids with dilu- tion of the blood. For the same reasons, hunger has been found by many observers to cause in animals a considerable increase in red cells. (Bidder and Schmidt, Voit, Subotin, Panum, Groll, Hosslin, Raum.) The observations on man by Raum, Senator, Luciani, and Grawitz,^ do not, however, show that such an increase is either constant or in- variable in fasting men. Twenty-four hours' fasting usually suffices to raise the red cells of dogs 400,000-500,000, and in captivity they may emaciate from starvation while the red cells continue above normal (Hosslin), but in man there has been a lack of uniformity in the changes observed. Senator and Luciani finding an irregular increase in the number of red cells in the professional fasters Cetti and Lucci, while Raum and Grawitz ^ found a distinctly anemic condition es- tablished in healthy fasting men. 3. Sex. — The relative anemia of women is not found until men- struation is established at about the fourteenth to sixteenth years. Indeed, Stierlin, who examined the blood of ten boys and ten girls under 15 years, found a slight and rather uniform difference in both red cells (350,000) and Hb (2.5 percent) in favor of female children. While the figures of Welcker, 500,000, have been generally accepted as representing the difference between the blood of men and that of women, Leichtenstern, from 191 observations, placed this difference at about one million cells. Later reports (Ziegler, Stierlin), as well as common clinical experience, would seem to indicate that the difference does not usually amount to 500,000, and that when the number falls below 4,500,000 in women there is some distinctly pathological con- dition to account for the relative anemia. There are at hand no systematic comparisons of the number of red cells in women at differ- ent ages, apart from the immediate effects of menstruation and preg- nancy. Stierlin has shown that both red cells and Hb are subject to more extensive physiological variations in women than in men. In specific gravity Dieballa found that with the same red cells and Hb the blood of women averages .0025 less than that of men. Mexstruatiox. — Since a hemorrhage which removes less than one- fiftieth to one-twentieth of the total volume of blood fails to noticeably affect the blood (Vierordt), the menstrual flow, normally about 100- 200 cc. and extending over a period of some days, usually fails to re- duce the red cells. On the contrary both red cells and Hb have frequently been found slightly increased after normal menstruation. (Vierordt, Hayem, Duperie, Scherpf, Reinl, Reinert.) This result PLETHORA. POLYCYTHEMIA. 91 has been referred to stimulation of blood-forming organs. A slight leucocytosis has also been noted by Hayem, Moleschott, and Re inert. The clinical rule that normal menstruation in health}' subjects fails to induce anemia but may do so when associated with other causes of anemia, is borne out by Reinert's observations on a chlorotic girl. In this patient a moderate flow reestablished through the use of iron failed to lower the blood count, but at two succeeding periods more profuse discharge was followed by a reduction in 7 days of 700,000 cells. The changes which occur under the most favorable circumstances are probably illustrated by the results of Sfameni, who examined the blood of six healthy women before, during, and after menstruation. He found that normal menstruation reduces the Hb 4—15 percent and the red cells 220,543. During the menstrual period the leucocytes are usually slightly increased, average 330, and the decrease of red cells is less marked (122,443). In the inter- menstrual period the red cells slowly increase, reaching a maximum three days before the succeeding floAV. In spite of the above largely negative results, it can hardly be doubted that a slightly disordered menstruation is a most important factor in the etiology of chlorosis. Pregxaxcy. — After passing through many phases of conjecture, opinions regarding the condition of the blood in pregnancy are now based on numerous observations extended and summarized by Schroeder, which show that pregnancy in itself has little influence on the Hb or red cells of the blood, and that if anemia exists it is to be referred to unhygienic conditions. Post-partum anemia should not continue longer than 10—14 days, and may be succeeded by a slight increase of red cells above the normal. PLETHORA. POLYCYTHEMIA. Although the subject of plethora belongs properly to the physiologist or general pathologist, the clinical importance of the condition requires for it some notice in a work of the present character. It has already been shown that there are good grounds for accepting the old belief that there are considerable physiological and pathological variations in the total quantity of blood in the body, and it remains to consider here those factors which directly affect the proportion of red cells. Pohjcythemia may arise either from an increase in the number of red cells or from a decrease in the volume of the plasma. Practically it can hardly be said that a true polycythemia from the first-named cause is ever observed, as the blood-forming organs are never known to produce an excess of red cells, and the effects of transfusion of blood are very transitory. Polycythemia from decrease in the volume of plasma, however, is very frequently encountered, and among its causes may be mentioned diarrhea, dysentery, cholera (Schmidt, Hayem) or any disease attended with excessive watery exudate ; conditions marked by insufficient aeration of the blood and venous stasis, as endocarditis, 92 MORPHOLOGY AND PHYSIOLOGY OF THE BED CELLS. emphysema, asphyxia, and in the peculiar permanent polycythemia, observed at high altitudes ; the administration in therapeutic or poison- ous doses of various drugs, lyraphagogues (Grawitz*), phosphorus (v. Jaksch), pilocarpine, eserine, etc. (Hamburger), and cold baths. Various other influences leading to local polycythemia have been mentioned in the Introductory Section. The danger of overlooking the true cause of polycythemia is well illustrated in Timofjewsky's experiments, when after the intra-venous injection of pus an increase of red cells of 1.5-2 million w^as found to result from concealed exudates from mucous and serous membranes. The Polycythemia of the New-Born. — The marked relative poly- cythemia of the blood of new-born infants has been constantly attested. The following averages have been observed : Hayem 5,368,000 Sorensen 5,665,000 Otto 6,165,000 Helot 5,531,000 Schiff 5.825,000 Bidone 6,500,000 While the averages of considerable numbers of cases reported by the above authors are rather uniform, very marked variations have been noted, referable to changes at different periods of the day, and to individual peculiarities. Hayem and Helot found that too early liga- ture of the cord might cause a reduction of one-half to one million cells. During the first weeks the polycythemia, which is usually highest be- fore the first nursing, gradually disappears at the rate of about 250,000 cells weekly, but while the average fall of a considerable number of cases is very regular, there are marked daily variations in individuals of from one-half to one million. (Schiff.) No definite relation between the changes in body weight and the number of red cells has been traced, but a marked reduction usually follows each nursing. The cause of the polycythemia of the new-born is doubtless to be found in a temporary concentration of the blood in which a number of factors are concerned. The loss of water during the first hours of respiration has been regarded as an important influence (Preyer), but can hardly account for the condition of the blood in the first hours after birth. The writer finds that the polycythemia bears a rather close relation to the degree of cyanosis exhibited by the expressed blood-drop, and believes that the concentration of blood is principally referable to a state of relative stasis which is then established in the peripheral capillaries. The Polycythemia of High Altitudes.— The observations of Bert in 1882 that the blood of animals from the Bolivian plateaus showed much increased capacity to absorb oxygen led Viault to examine his own blood while sojourning in the Andes and to find that at an eleva- tion of 4,392 meters the red cells had increased from five to eight mil- lions, many of them being undersized. In the blood of animals he found a similar polycythemia but no increase in oxygen. THE POLYCYTHEMIA OF DIARRHEAL DISEASES. 93 These observations were vei'ified by Egger, Koppe, Wolif, and many others, who found that the increase begins almost immediately upon the change in altitude, amounting at times to one million cells in 24 hours, reaching its limit in about two weeks, after which the numbers remain permanently high. On returning to sea level the polycythe- mia promptly disappears, even more rapidly than it is established. (Mercier.) The percentage of Hb appears to be less affected, and the volume of red cells not at all. Egger found no change in the dry residue of the blood of two rabbits, but Grawitz found the residue distinctly increased in a rabbit which had been confined in a rarefied atmosphere. The blood of anemic and phthisical patients is less markedly and per- manently affected than that of healthy individuals. The significance of these changes has not been fully explained, but it seems most probable that the polycythemia, if it really exists at all, is an effect of concentration of the blood, brought about by Nature's effort to adjust the respiratory function to the smaller proportion of oxygen in the rarefied atmosphere. That the state of the blood is one of simple concentration is, however, rendered uncertain by the low percentage volume of red cells. Egger and Koppe maintain that there is a formation of new red cells, but the microcytes found cannot figure as such evidence. Normoblasts ought to be present if, as Grawitz says,' about one liter of new blood has been produced in 24 hours, while the polycythemia promptly subsides at the sea level without yielding signs of blood destruction. The possibility that the cells are concentrated in the capillaries is negatived, since Egger found a simi- lar increase in the large internal arteries of rabbits, while the sugges- tion of Fick that the red cells may enjoy a lengthened existence in the circulation does not commend itself. Several observations, recently reviewed by Starcke, indicate that the results obtained by the hematocytometer are markedly dependent upon temperature and barometric pressure, and suggest that many, if not all, of the peculiarities of the blood at high altitudes may possibly be referable to this cause. Residence at the sea coast is frequently followed by increase in the number of red cells as a part of improvement in general health. In the tropics most Europeans become more or less anemic until fully inured to the various unhygienic influences obtaining there. The Polycythemia of Diarrheal Diseases. — The concentration of the blood which results from depletion of fluids through the intestine is seen to a slight extent in severe simple diarrhea, more distinctly in acute dysentery and at its height in cholera. An increase of two millions in the number of red cells has been ob- served by Hay within two hours after the administration of 21 grm. of Glauber's salts in concentrated solution, and Brouardel demonstrated less clearly a marked increase after purgation by croton oil and jalap. When given in dilute solution the effects of the salines are of course less evident. Clinically, a moderate polycythemia has been observed after severe diarrhea or vomltinoj, but the effects are usuallv transient. 94 MORPHOLOGY AND PHYSIOLOGY OF THE RED CELLS. In dysentery, polycythemia may be expected only when the diarrhea is acute and the quantity of blood lost is slight. Usually an anemia is established by direct loss of blood. In typhoid fever the progress of the anemia is often greatly ob- scured by the concentration of blood in cases marked by severe diar- rhea. In cholera the conditions are specially favorable for watery depletion of the blood and tissues and remarkably high counts have been recorded in the late stages of this disease. Polycythemia of Endocarditis, Venous Stasis, etc. — Very numer- ous observations have shown that at some stages of endocarditis a marked polycythemia is established, and largely through the studies of Oertel and of Grawitz it has been shown that this concentration is found in conditions marked by chronic stasis, cyanosis, edema, and is more marked in the capillaries than in the veins or arteries. The same condition is seen in chronic stasis from other causes. (Peiper, Grawitz.'^) Phosphorus Poisoning. — In cases of phosphorus poisoning of mod- erate intensity, but fatal, Taussig observed an increase in red cells to 8.5 millions, v. Jaksch an increase from 4.3 to 8.25 millions, and Lim- beck an increase to nearly 8 millions. The fact that the polycythemia does not appear until the toxic symptoms are marked and usually subsides with them about the fourth or fifth day, indicates that it re- sults largely from depletion of the blood through the vomitus. Von Jaksch found no simultaneous increase in the proportion of albumens. The leucocytes are usually diminished, sometimes slightly increased, and the Hb-index is lowered. Illuminating Gas Poisoning. — Munzer and Palma and Limbeck have observed moderate polycythemia in three cases (5.7, 6.63, 5.7 millions), with slight leucocytosis, 13,300 in one case. The condition must be referred to venous stasis. Polycythemia After Cold Baths. — An increase of red cells of .5 to 1.5 millions has been observed by many writers to occur after the application of cold baths. (Toenissen, Winternitz, Knopfelmacher, Thayer.) The change is observed immediately after the bath and con- tinues after cyanosis has disappeared and while the patient is shiver- ing, but soon disappears thereafter. It is often missing after 15-20 minutes. The leucocytes also are usually increased (15,000-20,000). The condition is ])robably referable to concentration of the blood due to stimulation of the vaso-constrictors, and is probably more marked, if not exclusively present, in the peripheral capillaries. Yet some have suggested that a new formation of cells may be indicated by the change. It may be noted here that warm baths have a contrary effect. Influence of Therapeutic Measures on the Red Cells. Iron. — The remarkable effect of this agent upon the blood in suitable cases of anemia is seen in a rapid increase in the number of red cells and in a later but rather more uniform increase of Hb. According to INFLUENCE OF THERAPEUTIC MEASURES ON RED CELLS. 95 Hayem the effects of iron in chlorosis may be divided into two periods, in the first of which the increase in red cells outstrips that in Hb until a normal number is reached, while in the second period the Hbof the new red cells is gradually brought to the normal. The appearance in such cases of many small red cells deficient in Hb has been commonly noted, but Stifler finds they are not equally abundant at all periods of the regeneration of the blood, that the increase of cells occurs period- ically, and that at the height of the periods many pale cells are found which increase in Hb-coutent while the numbers of cells remain nearly constant. Again, Willcox finds a difference in the effects of small and of large doses of iron. When small doses are used the red cells outstrip the Hb but with large doses both increase in equal proportion. Laache also has observed equal progress in cells and Hb in cases of chlorosis. It would seem therefore that the regeneration of the blood under iron may follow any one of the above three courses, depending upon various circumstances which cannot here be further discussed. Hayem and Reinert have called attention to the tendency of the red cells to diminish slightly after reaching a normal number, until the restoration of Hb becomes complete. Of the two Reinert believes that the progress of the Hb is the more uniform. Limbeck finds that the red cells usually diminish for some days when the administration of iron is begun. Indications for the Use of Iron. — The specific effects of this drug being centered primarily on the Hb, the chief indication for its use and the best results are obtained in cases of pure chlorosis with marked loss of Hb and moderate reduction in cells. In secondary anemia its curative action is less certain. Witli the appearance of larger cells icith normal or increased Hb, the use of iron becomes much less effective, and when the Hb-index is above normal it seems to be entirely valueless. A large part of the effects attributed to iron may be obtained solely by the improved hygienic and dietetic conditions which are usually en- joyed during its administration. The influence of rest in bed, selected diet, baths, oxygen, massage, etc., has long been recognized as a very powerful adjunct in the treatment of anemia. But numerous observa- tions go to show that while the numbers of cells may be greatly in- creased by these means yet the complete restoration of Hb in severe chlorosis is seldom as complete or rapid as when iron is freely given, so that a specific influence of this drug must be admitted. Of the great variety of preparations employed at various times, prob- ably the more irritant preparations in large doses are still acknowledged to yield the best results when well borne by the stomach. The curative action of iron in anemia is still lai-gely unexplained, but the theory of Binz appears to remain the most rational. Binz holds that in chlorosis, as a result of prolonged malnutrition, the absorption of iron from the food becomes deficient while its excretion through the liver and bile re- mains undiminished, and the blood becomes impoverished in this principle. 96 MORPHOLOGY AND PHYSIOLOGY OF THE RED CELLS. By administering iron in large doses it is more readily absorbed and the de- ficiency is overcome. Buno-e's theory, based on Sir Andrew Clark's belief that chlorosis is an autointoxication of intestinal origin, assumes that the iron administered is not absorbed in increased quantity but exerts its curative influence by neu- tralizing toxic agents in the intestine. These assumptions have been largely disproved by v. Noorden, Rethers and Morner, who have demonstrated that abnormal putrefactive processes in the intestine are not usually present in chlorosis, and by the demonstration that increased ingestion of inorganic iron is followed by its increased absorption principally or exclusively in the duodenum. (Kuiikel, Woltering, Hall, McCallum, Hochaus and Quincke, Cloetta Hoffman.) v. Noorden's^ conclusions seem to embody the results thus far obtained, that while iron in the nucleo-proteid combinations of the food is constantly available in chlorosis, it is either not absorbed or is not appropriated by the blood, whereas when given in large doses it is absorbed in excess and exerts a specific influence upon the bone marrow. The prob- lem of the nature of this influence remains of course untouched. Arsenic. — While iron is contraindicated by the appearance in the blood of megalocytes with increased Hb, under these circumstances arsenic frequently exerts an almost specific eifect in increasing the number of red cells and in stimulating the production and more uniform distribution of Hb. Its field is therefore chiefly in the severe and chronic anemias, but in the simpler forms the action of iron is often accelerated by combination with arsenic. Its therapeutic effects in disease are rendered more obscure by the fact that in health it causes diminution in the number of red cells. (Stierlin, Delpeuch.) Stockman and Greig claim to have found evidences of increased for- mation of red cells in the marrow of rabbits after long use of arsenic, yet the number of red cells in the blood remained normal. Mercury. — Numerous observations upon the effects of mercury in syphilis dating from those of Wilbouchewitch, and Keyes, in 1874-6, to those of Lezius, and Martin and Hiller, in 1890, have failed to demonstrate that this drug, in therapeutic doses, has any direct effect upon the red cells. It has been shown, however, that in doses beyond a limit which varies with individuals and circumstances, mercury usu- ally causes a prompt reduction in both cells and Hb, and distinct anemia. It does not appear, from these studies, that the improved condition observed in the blood of syphilitics using mercury is separ- able from the removal of the virus and the consequent general im- provement in health. A reduction in Hb and cells has been frequently observed soon after beginning mercurial treatment or after the drug has been pushed and seems to continue until the sy.stem becomes in- ured or the dose has been reduced to a proper limit, so that the ex- amination of the blood, as shown by Martin and Hiller, may be made a valuable guide in the administration of this remedy. Most of the conflicting results reported may be explained on the grounds that in doses beyond a variable limit, especially at the beginning of treatment, mercury reduces the cells and Hb ; that when the original disease is being successfully combated the anemia usually improves ; and that apart from its antisyphilitic powers the drug has little effect upon the blood. (Cf. section on Syphilis.) OLIGOCYTHEMIA. 97 Lead. — In cases of chronic lead poisoning the blood commonly shows a moderate grade of secondary chlorotic anemia. Zinn records an acute case with hemorrhages in which the red cells numbered 4 mil- lions, leucocytes 3,300, Hb 58 percent, s. g. 1.048. More severe cases have been reported Avith anemia usually in proportion to length of the disorder by Malassez, 2.2 million red cells; by Limbeck, 2.2 millions ; and by Brochin, 1.3 millions. Hayem states that the anemia of chronic plumbism is characterized by its close resemblance to chlorosis, by the rapid decrease in cells while the patient suffers from colic, and by the absence of leucocytosis. Hayem saw no case with increased Hb-index but Malassez reported megalocytes and megaloblasts as abundantly present in some cases. This anemia is commonly attributed to gastro-intestinal disturb- ance, but Hayem referred it to destruction of cells from direct action of lead. Maragliano demonstrated increased globulicidal activity of the serum in chronic plumbism and Grawitz has noted considerable degrees of granular degeneration of the red cells. OLIGOCYTHEMIA. Hemorrhage causes a peculiar type of impoverishment of the blood in which the oligocythemia is associated with other important changes in the plasma. The progress of events in the circulation which follow a severe loss of blood involves the consideration of many physiological processes. One of the first effects is a lowering of blood pressure, which is followed by an increased flow of lymph and by a rapid trans- fusion of fluids from the tissues through the capillary walls. The fluid which replaces the lost blood is necessarily of different composition from the plasma and its addition to the circulation markedly affects the composition of the blood after hemorrhage. The lymph, being rich in salts and poor in albumens, and water from the tissues passing through capillary walls more readily than albumens, the blood when restored to its normal volume is found to be low in albumens, rich in salts, and poor in red cells. After rapid hemorrhages the alkalinity of the blood falls (Zuntz), and its content in sugar increases (Bernard, Mering), this principle being derived from the glycogen of the liver (Schenk), while the coagulability is much increased. The hydremia affects principally the plasma, but Herz found the relative volume of the red cells tripled ten hours after a very severe hemorrhage, indi- cating that they had absorbed much water. In small animals (rabbits) the restoration of the voliuiie of blood takes place very rapidly and oligocythemia is observed almost immedi- ately after the hemorrhage, but in larger animals (dog, man) a distinct interval is required before the fluids have replaced the lost blood, and, according to Limbeck, 35—40 minutes may elapse before a distinct re- duction in red cells is observed after moderately severe hemorrhages. The changes in the number of red cells following hemorrhages are 7 98 MORPHOLOGY AND PHYSIOLOGY OF THE RED CELLS. somewhat irregular, but from the observations of Huherfauth, Lyon, Otto, Koeppe,^ Viola and Jona, it has been shown that in dogs and man a single large hemorrhage reduces the red cells in proportion consider- ably less than the effect upon the volume of blood, beginning about one-half hour after the operation, reaching a maximum effect in 3-4 days, and followed by a restoration to the normal number in 19-34 days. (Lyon.) In small animals (rabbits) the rate of decrease is more rapid, the lowest point being reached in a few hours, and recovery also following earlier. With slow hemorrhages of large extent (3-4 percent) the minimum is reached in 1-9 days, depending largely upon the size of the animal, and recovery is complete in 14-22 days. (Huhnerfauth.) After lesser hemorrhages (1-3 percent) full recovery follows in 5-14 days, while small losses leave no trace after 2-5 days. (Lyon.) These periods depend much upon the state of nutrition, and it has been found that full diet, free supply of water, and transfusion of salines, greatly accelerate recovery. Ehrlich believes that the continuous reduction in cells is partly referable to the solution of many of them as the plasma becomes more watery. During both phases, the Hb falls behind the changes in the cells (Otto, Reinert), owing to the formation of numerous microcytes by the splitting of red cells. (Koeppe.) Grawitz however does not believe that such splitting of red cells can occur. Nucleated red cells of nor- mal or very small size appear within a few hours and in large numbers after severe hemorrhage, and even after small losses, if sudden. That they are drawn into the circulation mechanically is indicated by their usual absence after slow bleediugs of much larger extent. (Zenoni.^) Changes in the size and shape of the cells are conmionly observed especially after large bleedings. Microcytes and slightly enlarged megalocytes usually appear in 2-4 days, at the period of greatest oligo- cythemia. (Koeppe.) The extent and duration of these changes de- pend largely on the grade of anemia established. Polychromatophilia is usually marked within 24 hours. Healthy men recover rapidly from moderately severe hemorrhages and so completely that the red cells may be more numerous than be- fore, from apparent over-stimulation of the marrow. (Otto, Hall and Eubank.) Small and repeated hemorrhages, on the other hand, have led to some of the most severe forms of anemia ever recorded, in which the morphological changes of pernicious anemia are pronounced, but the prevailing feature of the blood is the loss of Hb. The leucocytes are usually, but not always, much increased (20,000— 40,000), the neutrophile cells being most affected, the eosinophils sometimes appearing in considerable numbers (Lyon, Hall and Eubank), and the small lymphocytes showing a relative increase. (Rieder.) Ehrlich and Lazarus found 13.7 percent of myelocytes in a marked post-hemorrhagic leucocytosis, and in one of the writer's cases of splenectomy with profuse hemorrhage the blood for some days closely resembled that of myelogenous leukemia. OLIG OCYTHEMIA. 99 The post-hemorrhagic leucocytosis is referred by Virchow to the re- tention of cohesive leucocytes in the small vessels, by Ehrlich to their active new production in bone marrow, and by Rieder to the outpour of the lymph. It is usually persistent in some degree till the regenera- tion of the blood is nearly complete. From very extensive observations in Mickulicz' clinic, Bierfreund found that the regeneration of the blood is most rapid in male sub- jects between 20-40 years of age, is considerably earlier in males than in females, and in adults than in children or old persons. A loss of 5 percent Hb was found to lengthen the period of regeneration 2—8 days. The minimum percentage of Hb was reached, after a loss of 10-1 5 percent, in 3.5 days; after 16-20 percent, in 5.8 days; after 21-25 percent, in 6.5 days ; after 26 percent, in 9.6 days, and in women usually one day later than with men. It is significant that in Bierfreund's tables there are no averages on patients losing more than 30 percent of Hb, indicating that when a surgical operation causes a loss of 30 percent Hb it is usually fatal. Mickulicz attempted to deduce from this fact the rule that no opera- tion should be undertaken unless the patient could stand such a loss, but it is difficult to see any great practical application of such a rule. Bierfreund's Table, Showing Average Time of Full Regeneration OF Blood After Hemorrhage. Loss of Blood Sex. Age(yrs.). m>f. M. F. 1-10 10-20 20-30 30-40 40-50 50-60 60+ 10-15 12 7 17 9 20" 3 27 17.1 days. 23.5 23.5 31.3 20 21 22.5 13.2 19.0 21.3 10.8 12.5 17.6 10.4 16.2 16 21 19 17. 22.7 28.0 16-20 21-25 23.8 27. 26 Anemia Following Repeated Small Hemorrhages. — While very marked changes in the blood following a single large hemorrhage may be rather promptly restored, the regeneration is very greatly retarded after even one repetition of the hemorrhage, and when they are frequently re- peated, there is soon established a severe form of chronic anemia. It does not appear that this anemia differs greatly from other equally se- vere secondary forms, but it is usually characterized by marked reduc- tion in the albumen (less of the globulin) of the plasma and serum ; by lowering of the dry residue and ash, due to loss of Fe, P and K, which more than balances the increased Na ; and by lowered alkalinity. Besides the actual withdrawal of albumen, repeated hemorrhages seem to have almost a specific effect in increasing fat formation at the ex- pense of albumen both of the blood and tissues. (Leube.) The loss of red cells is here an accurate measure of the grade of anemia and all ordinary forms of their degeneration are observed. In mild cases the loss of Hb is relatively slight but in some severe cases the Hb-percentage and index are very low. On the other hand there can be no doubt that the state of the blood following repeated small hem- 100 MORPHOLOGY AND PHYSIOLOGY OF THE RED CELLS. orrhages cannot always be distinguished from that of primary pernicious anemia with increased Hb-index. Anemia from hemorrhages is one of the secondary forms most commonly marked by leucocytosis, yet this feature is often not marked and is not infrequently absent. With man and animals the clinical effects of hemorrhage vary greatly among individuals. Bechampand Huhnerfauth find that dogs withstand the loss of 3-4 percent of the body weight, 30-40 percent of the total volume of blood, if the loss is gradual, but may perish after smaller but more rapid bleedings. As a rule the smaller the animal and the more rapid the hemorrhage, the more severe are its eifects. Adult male subjects usually die if losing rapidly 50 percent of the blood, but may survive much greater losses from slow bleeding. Children are much more susceptible and may die after very small hemorrhages, if rapid, while women withstand effects of hemorrhage much better than children or men. The regeneration of the blood however is more rapid i)i men than in women. (Bierfreund.) The quantity of blood which may be lost without causing death varies greatly among individual men and animals, as the following table will illustrate : Limits of Hemorrhage from which Recovery has been Observed. Author. Animal. Percentage of body weight lost. Percentage of red cells remaining. No of red cells remaining. Dog. ti (( n (( (( Kabbit. Man. Woman. n 2 hemorrhages in 6 days. 4.58 (not fatal. ) 4.33-5.55 4.3-7.3 5.48-6.57 average 5.12 5.4 5.44 (even (always chance. ) fatal. ) 4.5 3. 50 50 32% 11% Hayem Kireeff Maydel Schram Landerer Feis ■t Andral, Behier. Laache 1,598,000 1,415,000 Haveni TRANSFUSION OF SALT SOLUTION, SERUM, AND DEFIBRINATED BLOOD. Siegel and Schram both found no improvement in the regeneration of the blood from the transfusion of salt solution or of serum in ani- mals after bleeding, and while it has since been shown that the regen- eration is somewhat more rapid and complete after salt infusion, yet this procedure must be regarded as of more value as a means of saving life than as a stimulant to blood formation. The direct effects on the blood of infusion of defibrinated blood seem to be much more favorable. On the other hand, after the transfusion of blood the above ob.servers. ORIGIN OF SECONDARY ANEMIA. 101 and others, have found a rapid increase in the number of cells, and Bizzozero reported the same effect after transfusion of defibrinated blood in animals. Quincke was one of the first to note an increase in red cells in pernicious anemia as a result of transfusion of blood, and similar observations have been made by Ziemsen in anemia and scurvy. It would seem from the observations of Bizzozero and of Bareggi that the red cells are quite resistant to the process of defibri- nation and injection. Of the immediate effects of salt infusion upon the blood of the human subject there are a few reports at hand, which indicate that it has considerable influence in lowering the numbers of red cells and increasing the leucocytes. ORIGIN OF SECONDARY ANEMIA. The anemia associated with acute and chronic diseases is a result of many factors of varying prominence in different conditions, and re- quiring consideration in detail. Diminished Nutrition. — Since the digestive powers are largely in abeyance in severe febrile states, the loss of the usual supply of albu- minous principles must figure to some degree in the impoverishment of the blood. Although the complete withdrawal of food in healthy men leads to a simple " atrophy of the blood," with other tissues, marked by diminution in bulk, polycythemia, etc., but not to anemia (Hei- denhain, Voit), it by no means follows that similar results would occur in disease, while distinctly anemic conditions are known to result from improper quality of food. (Lichtenstern.) Grawitz (p. 69) has found that food deficient in albumen may cause a diminution in the albumen of the plasma, which may be demonstrated within 4—8 days, is more marked when the subject is exercising, and later affects also the num- ber of cells. Diminished Activity of Blood-forming Organs. — Simultaneously with the loss of digestive powers, it can hardly be doubted that the normal process of red cell formation must be interrupted in most infectious dis- eases. This conclusion may be based upon the known effects of tox- emia upon cellular processes, or the absence of normal stimulus from absorbed food products, and upon anatomical conditions demonstrated in the bone marrow. Regarding the last factor it may be said that the leucocytosis of pneumonia is associated with proliferation of myelo- cytes in the marrow at considerable expense of normoblasts, and in typhoid fever the normal relations in the marrow cords are greatly dis- turbed by congested, dilated, or even ruptured sinuses. Increased Consumption of Albumens. — The toxic destruction of albu- mens in fever has been fully demonstrated by Hallervorden and Leube,^ and their products noted in the increased excretion of urea, of urinary pigment (Jaffe), and potash salts (Salkowski). While the tissue proteids suffer most in this process it is but a step to the source of these prin- ciples in the blood. That the albumens of the blood and tissues suffer in some degree in afebrile cachexias is indicated by the results of Miiller's studies on metabolism in carcinoma. 102 MORPHOLOGY AND PHYSIOLOGY OF THE RED CELLS. Increased Destruction of Red Cells. — The increased globulicidal activity of the serum in infectious and septic processes is undoubtedly a most potent agency in the diminution of red cells, and has been demonstrated by Maragliano in carcinoma, pneumonia, typhoid fever, erysipelas, tuberculosis, and many other conditions. The very rapid deo-lobulization observed in malignant endocarditis and septicemia leaves abundant evidence of this peculiar process also in the pigment deposits found in the viscera. Of the sources of the globulicidal agent little is known, but it is certain that the products of disordered tissue metabolism and bacterial toxines are capable of dissolving red cells when injected into the circulation. Fischer and Adler saw the red cells fall from 6 to 1.3 millions in 7 days after injections of streptococcus cul- tures into a rabbit; and Grawitz could find only ;>00,000 red cells in a remarkably rapid case of streptococcus septicemia in an adult woman. Bianchi-Mariotti found a marked loss of Hb in direct proportion to the quantity of culture used, after injections of cultures of B. tiiphoxm, cholera, anthrax, etc. There is abundant evidence to show that in many conditions asso- ciated with acute infection and resulting from chronic cachexia the red cells are abnormally fragile. Of such evidence may be mentioned the occurrence of paroxysmal hemoglobinemia on slight exj)()sure to cold, and the rapid development of anemia in many mercurialized subjects. Loss of Blood by Hemorrhage and Exudation. — Perhaps the majority of severe cases of secondary anemia result from the complication of the original disease by single or repeated hemorrhages. Post-typhoid anemia is usually slight and sometimes absent if there have been no bleedings from the intestines, and cancerous cachexia is very promj^tly aggravated when ulceration and hemorrhage supervene. Much of the anemia in anchylostomiasis is referable to n)inute hemorrhages. Of the depleting effect of albuminous and bloody exudates goo>.-^. Normal Blood. (Eosin and Methylene Blue.) Fig. I. Normal rouleau. Fig. 2. Red cell, extruding slightly basophilic (nuclear?) contents as blood plates. Fig. 3. Lymphocyte. Fig. 4. Large mononuclear leucocyte with finely reticular cytoplasm. Fig. 5. Polynuclear neutrophile leucocyte, granules unstained. Fig 6. Eosinophile leucocyte. LEUCOCYTES IN PATHOLOGICAL BLOOD. 107 lymphemia medium-sized, strongly basophilic hyaline leucocytes with incurved or subdivided nuclei, which must be classed with lymphocytes. 2. Large mononuclear leucocytes may be only slightly larger than lymphocytes, but many of them are the largest cells seen in normal blood. Their protoplasm is slightly basophile and very finely reticular, with nodal thickenings often resembling granules. Their nuclei are vesicular, rather coarsely reticulated, with one or two central nodal thickenings resembling nucleoli, and may be circular, or horseshoe- shaped (transitional leucocytes), or elongated. In Ehrlich's triacid solution the bodies of mononuclear basophile cells stain very faintly red. (Plate I.) 3. Polynuclear ^ leucocytes are two or three times as large as the red cell. Their protoplasm is reticulated and possesses as inte- gral parts of the reticulum protoplasmic neutrophile granules in con- siderable number. The reticulum is otherwise very slightly basophilic and may be demonstrated, with basophilic nodal thickenings, by methylene-blue. (Plates IV., VI.) Their nuclei are elongated and constricted, or composed of two or more lobes usually connected by threads of chromatin. These lobes ai'e coarsely reticulated and usually possess a central nodal thickening. They may become completely separated from each other. 4. Eosinophile leucocytes vary in size from that of lymphocytes to that of polynuclear leucocytes. Their protoplasm contains large strongly acidophile granules which are believed to be integral parts of a cytoreticulum. (Heidenhain, Gulland.) Their nuclei are coarsely reticulated aud usually bilobed, the lobes are more often separate than in the neutrophile cells, aud they stain rather faintly with nuclear dyes. Leucocytes in Pathological Blood. — Besides the above forms of colorless cells w^iich alone are present in normal blood, other types of cells are seen in the circulation in disease. 1. Myelocytes are mono- nuclear cells with neutrophile or with eosinophile granules. Three types of myelocytes should be distinguished, (a) Ehrlich's mye- locyte is a medium-sized cell with pale, usually central nucleus, and neutrophile granules. It is found in many diseased conditions, especially in leukemia and secondary anemia. (Plate VII.) (b) Corxil's myelocyte is a large cell, much larger than a poly- nuclear leucocyte, with pale eccentric nucleus and neutrophile gran- ules. It is found almost exclusively in myelogenous leukemia, and less frequently in v. Jaksch's anemia. (Plate VII.) (c) Eosinophile MYELOCYTES may resemble the eosinophile cells of normal blood except that their nuclei are single. Such cells are abundant in myelogenous leukemia, occur not infrequently in v. Jaksch's anemia, and have been found in myxedema by Mendel, in some infectious diseases by Turck, and in pernicious malaria by Bignami. Or their granules may be of ' The attempt to displace Ehrlicli's old term "polynuclear" by the unfortunately derived "polymorphonuclear" seems to tlie writer ill-advised. Tiie former designa- tion has become tirraly established, never leads to confusion, and seeing that polymor- phism characterizes nuclei of other mononuclear cells, and that the lobes of the neu- trophile cell are often separate, there appears to be no gain in accuracy by such a change. 108 THE LEUCOCYTES AND LEUCOCYTOSIS. excessive size, in which case they are pathognomonic of myelogenous leukemia. (Plate VII.) 2. Mast-cells are mononuclear or polynuclear cells of different sizes, whose characteristic feature is the presence of large and small strongly basophile granules. These granules do not reflect the pure color of many stains but are metachromatic, especially with thionin. An iso- lated mast-cell is not infrequently seen in normal blood, but when present in any considerable numbers they are pathognomonic of myelo- genous leukemia. (Plate X., Fig. 1.) The Classification of Leucocytes. Although the leucocytes were discovered by Nasse in 1835 the first scheme of classification based upon their supposed points of origin was offered by Yirchow who divided the colorless cells into lymphocytes derived from the lymph-nodes, splenocytes from the spleen, while poly- nuclear cells he regarded as developmental forms of the mononuclear. Similarly, Einhorn, a pupil of Ehrlich's, classed the leucocytes as : (1) lymphocytes, small and large, derived from the lymph nodes, (2) myelogenous cells (eosinophiles) from the marrow, and (3) large mono- nuclear, transitional, and polynuclear cells, from spleen or marrow. In 1865 M. Schultze described the leucocytes as (1) non-granular (large and small mononuclears), (2) finely granular (neutrophiles) and (3) coarsely granular (eosins and mast-cells). He believed that all granules and cells represent developmental forms of one series. Lowit's classification, based on the morphology of the nucleus, in- cluded small and large mononuclear cells (lymphocytes), "transitional " leucocytes, and polynuclear leucocytes. Regarding solely the nucleus, Lowit found no difficulty in deriving all leucocytes in one series of cells. It remained for Ehrlicli to establish the essential distinction be- tween leucocytes by demonstrating specific microchemical reactions in the granules of Schultze. Ehrlich divided the aniline dyes into three main groups. (1) Basic dyes, as hematoxylon, methylene-blue, thionin, etc., act as bases, unit- ing, with selective power in the order named, with the acid principles of cells (nucleinic acid). (2) Acid dyes, as eosin, fuchsine, aurantia, act as acids, and unite with the basic principles of cells. (3) Neutral dyes. When certain basic and acid dyes are mixed a compound is formed of modified staining qualities which unites with certain cell structures not readily stained by other methods. Such a mixture Ehrlich calls a neutrophile stain, an example of which is the triacid mixture. Neutral red seems also to fall in this class. According to their reactions to these dyes the granules of leucocytes in human blood may be divided into three main groups, basophile, acidophile (oxyphile), and neutrophile, and on these grounds the present classification of leucocytes is based, as follows : 1. Basophile Cells: Mast cells (j-gramdes). Granules strongly basophile. Lymphocytes. Protoplasm strongly basophile. Large MONONUCLEAR LEUCOCYTES. Protoplasm slightly basophile. FINER STRUCTURE OF LEUCOCYTES. 109 2. Neutrophile Cells : Polynuclear leucocytes. Neutrophile [e) granules. Myelocytes. Neutrophile granules. 3. Eosinophile Cells : Eosinophile leucocytes. Large acidophile granules, {a-gramdes.) Eosinophile myelocytes. Large acidophile granules. Ehrlich also described granules which are araphophile, stain by both basic and acid dyes, and are found in some marrow cells, and o-granules which are small basophile granules said to occur in some mononuclear cells. It is impossible here to consider all the objections raised against the specific quality of the reactions demonstrated by Ehrlich and espe- cially the limitations shown to exist in the application of his principles to other tissue cells. At present it must be admitted that while there are rather narrow limits between which the reactions of granules may vary, yet each class of granules is peculiar to one cell and transitional stages have not been demonstrated, while the separation of leucocytes according to the chemical character of their protoplasm is a much more exact classification than one on any other basis. While the original views attributed to Ehrlich have suffered considerable modifi- cation since it became evident that the granules of many tissue cells though of variable composition may give identical reactions to dyes, and that the granules of the same leucocytes may stain differently, yet the wisdom of his classification is shown by the evidence which has accumulated regarding the significance of these granules. Finer Structure of Leucocytes. Recent studies have considerably increased the knowledge of the minute structure of leucocytes. For many years opinions have been at variance regarding the struc- ture of the protoplasm of basophilic cells. It is now evident that the bodies of these cells show the reticular structure of ordinary protoplasm and that any granules normally present in them are fine nodal thicken- ings of this reticulum. Moreover such nodal thickenings when most prominent do not have the sharp character of the neutrophile or eosin- ophile granules, and these cells are properly regarded as non-granular. It has been shown by Arnold, Gulland, and others, that the neutro- phile and eosinophile granules are connected by threads with the retic- ulum of their cells and are thus integral parts of this reticulum (plas- matic). Under suitable conditions, in thinly spread, rapidly dried, densely stained cells, it is apparent that the cytoreticulum of large mononu- clear, neutrophile, and eosinophile cells, is directly continuous with the nuclear network and that the clear space sometimes appearing to sepa- rate nucleus from body is artificial. (Gulland.) Centrosomcs have been demonstrated in the leucocytes of lower ver- tebrates by Flemming, Arnold, Van der Stricht, Hermann, Heiden- hain, and in human blood cells by Gulland. These bodies show no 110 THE LEUCOCYTES AND LEUCOCYTOSIS. special affinity for any dyes but are best demonstrated by basic stains. They have been stained by saffraniu and gentian-violet (Flemraing) and by methylene-bhie (Gullaud). In human blood they are ex- tremely minute and their demonstration appears to be a very uncertain undertaking. Small circular meshes sometimes appear in the nuclei of lymphocytes, as depicted by Ehrlich, but while these appearances resemble nucleoli it does not appear certain that their significance is really such. The only colorless cells in the blood in which the writer has been able to find distinct nucleoli are endothelia. Significance of Cell-granules. — The theory of the higher significance of cell granules was first clearly stated by Altmaun, although it was suggested by Ehrlich in his early studies of the granules of leucocytes. Instead of being degenerative or excretory products, it is now api^arent that they are a secre- tory product, and represent the center of the specific function of the cell. Strong evidence in favor of this view was furnished in the demonstrations by Heidenhain, Arnold, GuUand and others, that the granules are not para- plastic, that is, loose particles separate from the cytoreticulum, like fat, melanin, etc., but are plasmatic, i. e., integral parts of the cytoreticulum. The specific quality of the granules is also indicated by their chemical and tinctorial reactions. The diflfereutial chemical characters of the neutrophile and eosinophile granules will be considered later. (See Chapter V.) Al- though the granules of every cell of the same type may not all stain ex- actly alike, yet when submitted to fine chemical tests, one cell is found to contain only one kind of granule. The studies of Knoll, Hirschfeld, Kurloff, and others, upon the leucocytes of lower animals have shown that the wander- ing cells of vertebrates and invertebrates possess granules which are peculiar to each animal and are strictly limited to separate types of cells. Ehrlich regards the opposite behavior of different cells under different chemotactic influences as evidence of the specific nature and function of each variety of leucocyte and of its peculiar granule. While the above considerations point to the secretory nature of the granules, the exact function of the secretion has not been demonstrated. Altmann believes that they furnish oxygen for the metabolic processes of the cell. Ehrlich once (1891) regarded them as a sort of reserve material destined for use within the cell, but has not expressed himself recently on this point. Hankiu,^ Kanthack, Hardy and Keng, apply the term alexines to the secretoiy granules of leucocytes, and believe that they are destined to be thrown out into plasma or lymph and to exert a bactericidal or antitoxic influence. Sherrington and most observers cannot find sufficient ground for the acceptance of this view. Metchnikoff i refers to the granules as reserve material, but does not point out their exact destination. Numbers of Leucocytes. The first estimates of the number of leucocytes in the cubic milli- meter of blood were made by Samuel and Welcker and were generally accepted for some years. Malassez, with his improved method, con- siderably reduced these older figures. With the introduction of Thoma's instrument still more accurate estimates were secured which remain undisturbed at the present time, and agree fully with the more recent results obtained by Rieder, Reinert, and Limbeck. The following table is compiled from the chief contributions on this point : PROPORTIONS OF VARIOUS FORMS OF LEUCOCYTES. HI Observations on Normal Numbers of Leucocytes. Min. Samuel Welcker Moleschott Mahissez ^ Duperie Hayeni Bouchout, Dubrisay Grancher Hirt Thoma, Lyon.... Halla Tumas Sorensen Patrigeon Eeinecke Von Jakscli Limbeck -o- J f adults ... Kieder { , •, i t children Reinert j 6 A. M t4P. M. Average. Max. 14000 13432 14925 14000 7692 4545 6000 6116 9000 f 8388 15464 10590 7066 7533 10106 6200 9600 10 10000 7351 7482 8-9000 7680 9600 9960 12400 5125 8262 12857 12605 3750 3000 6784 4430 4960 4800 2000 4200 7200 In estimating the leucocytes in both health and disease one is con- fronted by a great variety of disturbing factors, including the causes of distinct physiological leucocytosis, the individual peculiarities of the subject, and the ordinary variations in the local condition of the part from which the blood is taken. All that has been said regarding ac- cidental variations in red cells applies equally to leucocytes, and one must carefully consider the effects of vaso-motor phenomena, of changes in the volume of plasma, and of the presence of inflammation or edema. It should be remembered that while the leucocytes remain nearly uni- form in the great vessels, their proportions in the capillary circulation may change more rapidly thau those of the red cells, owing probably to chemotactic influences. The most common sources of error may be avoided by taking specimens about 4 hours after a meal, and at the same hour each day. Proportions of Various Forms of Leucocytes. — The proportions of the different forms of leucocytes in normal blood are even less fixed than their numbers. Ehrlich's figures may well serve as a standard for healthy adults. Lymphocytes, 22-25 percent. Large mononuclear and transitional leucocytes, 2-4 per- cent. POLYNUCLEAR NEUTROPHILE LEUCOCYTES, 70-72 percent. Eosinophils cells, 2-4 percent. Mast-cells, .5 percent. The chief variations from these limits which deserve mention are the maximum percentages given by Rieder for lymphocytes (30 per- cent) and by Limbeck for polynuclear leucocytes (80 percent). In childhood the proportion of lymphocytes is usually much increased 112 THE LEUCOCYTES AND LEUCOCYTOSIS. (55-66 percent), and that of polynuclear cells correspondingly dimin- ished (28-40 percent). (Gundobin, Rieder.) Degenerative Changes in Leucocytes. While degenerative changes in leucocytes have always been recog- nized as necessary processes in degenerating tissues in which the leu- cocytes were lodged, references to acute degenerative changes in leu- cocytes of the circulating blood have been so scanty as to lead some writers to deny their existence. It therefore becomes necessary to consider briefly the general evidence at hand concerning the occurrence and significance of such changes in circulating leucocytes. It was perhaps Lowit's claim that leucocytolysis necessarily precedes leucocytosis, which called special attention in recent years to the de- struction of leucocytes in the circulation. Although Lowit's claim cannot be fully supported, the fact that the majority of leucocytoses are preceded or accompanied by the destruction of many leucocytes in the circulation has been amply proven by the work of Uskoff, Botkin, Gabritschewsky,'Engel, Klein, Goldscheiderand Jacob, and many others. E. Botkin followed up the subject in a series of experimental and clinical studies and showed that in 1 -percent solution of peptone and in the fresh plasma of various infectious diseases, all leucocytes dissolve and disappear with varying rapidity. The lymphocytes proved most resistant. The changes described bv Botkin were of an extensive variety but the principal features included the formation of coarse granules in mononuclear cells ; extrusion and loss of granules from polynuclear cells ; formation of vacuoles and clefts ; fragmentation and solution of cell-bodies ; formation of blood plates ; swelling, fading, shrinkage, subdivision and final disappearance of nuclei. Comparing these experimental changes with those observed by himself or reported by others in infectious diseases and in leukemia Botkin declared that no examination of the blood can he considered complete which leaves out of account the degenerative changes in leucocytes. The principal contributions which seemed then (1896) to justify Botkin's claim were those of Hankin, who found in the extrusion and loss of granules of the rabbit's leucocytes the source of bactericidal alexines, of Haetaguroff, who observed various degenerative changes in the leucocytes of typhoid fever, of Engel, who described a variety of changes in the leucocytes of an anemic child, and of Gumprecht, who called special attention to the pathological changes in the leucocytes of leukemia. In 1895 the writer described degenerative changes in the leucocytes of diphtheria. Similar lesions have since been reported by File, who regarded many of the altered cells as dead, and by many others. In the absence of more extended observation it seems best to merely describe rather than attempt to accurately classify the pathological changes in leucocytes, yet a tentative classification may be employed for convenience sake. ACUTE DEGENERATION OF LEUCOCYTES. 113 Acute Degeneration of Leucocytes. — This process is best demon- strated in acute infectious diseases, especially in severe cases. As seen in diphtheria the alterations aifect both body and nucleus. In diphtheria the writer believed he could detect an increased acido- phile Htain'ing tendenci/ in the neutrophile granules, a conclusion which he is disposed to retain, seeing that Kanthack had previously found that the pseudo-eosinophile granules of the rabbit may be made to show increased acidophile tendency and higher refractive quality by the injection of minimum immunizing doses of microbic poisons. This ob- servation, however, has not been further confirmed. That this change when it occurs belongs in the degenerative series is not clear, but it may be seen in cells which show other distinct signs of degeneration. Diminution in the number of neutrophile granules is commonly ob- served in the polynuclear cells in acute leucocytosis. This abnormality may progress till very few granules are left, but their complete absence is seen principally in chronic leukemia. It is usually associated with marked nuclear changes. Eosinophile cells may be similarly altered. Swelling and fragmentation of the bodies of leucocytes is commonly observed in acute leucocytoses. It is probable that many of the torn and distorted cells seen in dry preparations have undergone degenera- tive changes which have favored their complete destruction in smear- ing. These cells have been described as " leucocyte shadows " (Klein), a term well suited to recall their torn bodies, scattered granules, and faded nuclei. The reticulum of large and small mononuclear cells in leukemia and in severe acute toxemia may grow coarser and at times granular. Bot- kin has described the appearance of granules in the dissolving lympho- cytes of shed blood. Some of the lymphocytes of leukemia appear to be distinctly granular. Ehrlich depicts the separation of small periph- eral fragments of the protoplasm of lymphocytes. Globular projec- tions from such cells are frequently encountered but it does not appear certain that they represent degenerative plasmoschisis. Nuclear changes appear to be the most significant of the lesions of acute degeneration of leucocytes. In fresh or dry specimens the nuclei stain less densely with basic dyes, their outlines are irregular and the lobes shrunken. The degeneration may follow the type of karyolysis, with swelling and loss of chromatin, or of karyorhexis with hyper- chromatosis and subdivision of lobes. (Gumprecht.) In acute leuco- cytosis the former type is more usual, but in leukemia, the latter form is abundantly seen. (See Plate IX.) While the lobes of the normal polynuclear leucocytes are almost invariably connected with a thread of chromatin, many of the cells in severe acute leucocytosis show com- plete subdivision into 3-6 separate segments. Chronic Degeneration of Leucocytes. — In diseases like leukemia in which peculiar chronic changes in white cells are prominent, many of the alterations commonly seen in acute degeneration are also present. Thus, in myelogenous leukemia one finds leucocytes showing loss of granules, presence of vacuoles, granules of glycogen, faded irregular 114 THE LEUCOCYTES AND LEUCOCYTOSIS. nuclei, and distortion and fragmentation of cell-bodies. Some of the cells in leukemic blood, however, exhibit changes which at least in their abundance are characteristic of the disease and which seem cer- tainly to be the result of chronic processes. Complete loss of neutrophile granules may be observed in many cells, both polynuclear and mononuclear, in leukemia, less frequently in other conditions. The cell body remains slightly acidophile, while the nucleus commonly shows degenerative changes. (Plate IX.) Complete subdivision of nuclei of polynuclear cells into 6-1 hyper- chromatic segments is a rather characteristic appearance in leukemic blood, and is usually associated with complete loss of granules. (Plate IX.) Hydropic Degeneration. — In fresh blood many leucocytes may be found whose nuclei are indistinct or invisible and whose bodies contain many large and small vacuoles bounded by homogeneous or granular reticulum. These disintegrating cells are specially abundant in ma- laria and leukemia. In malarial blood when such cells are pigmented, the prominent vacuoles, opaque granules, and streaming movements common in dying protoplasm, strongly suggest the appearance of vacuolated parasites. In stained specimens they are found to be com- posed of pale staining remnants of nuclei. (See Plate VIII., Fig. 3.) Hydropic degeneration of nuclei is practically limited to some peculiar leucocytes, both mononuclear and polynuclear, of leukemia. (Plate IX.) Fatty degeneration of leucocytes has often been referred to as a part of the changes in pathological blood. The writer has seen globules in leucocytes blackening with osmic acid in sections of blood in the ves- sels of fatty viscera, very rarely in circulating blood. The presence of these globules in leucocytes is not a certain indication that they have resulted from the altered albuminous constituents of the cell nor does it necessarily indicate a degenerative process, so that in the ab- sence of further evidence, the writer is unable to include this form of degeneration among those affecting white blood cells. Even when fat is abundant in the plasma the leucocytes are commonly free from it. (Gumprecht.) Perinuclear Basophilia. — In 1894 Neusser (and later, Kolisch) described the appearance of basic staining granules about the nuclei of polynuclear and other leucocytes, which he termed as above and claimed to be a somewhat pathognomonic sign of the uric acid diathesis. At the time this observation was announced Lowit stated that perinuclear basophilia was an artifact which he had kept in mind for some ten years. Nevertheless Neusser's perinuclear basophilia brought consid- erable notoriety to the Vienna school during several years. Its lack of clinical or pathological significance has been fully demonstrated, among others, by Futcher, and by Simon. Blood-dust. Hemokonia.— The older histologists have rather uniformly described in normal plasma minute spheroidal or spindle-shaped bodies of SIGNIFICANCE OF INFLAMMATORY LEUCOCYTOSIS. 115 undetermined origin. (Bizzozero, Landois, Hayem.) Eecently Miiller 2 has drawn special attention to their occurrence, describing them as si:)heroidal, or dumb-bell, or spindle-shaped granules \ to 1 u. in diameter, and exhibit- ing active Brownian movements, but no locomotion. They are not fat, being insoluble in alcohol and ether and not stainable with osmic acid, and they appear to have no connection with the process of fibrin-formation. Their identity with the granules of leucocytes was considered by Miiller, and is claimed by Stokes and Wegefarth, who found them to increase in number as the specimen stood, and observed that in different animals their size va- ried with that of the granules of the leucocytes. Nicholls identifies them with leucocyte-granules on micro-chemical reactions. Miiller found them unusually abundant in a case of Addison's disease, and scanty in starvation and cachexia, but no clinical significance has been demonstrated for them. If not derived from leucocytes, they may be in part the extruded contents of fragmenting red cells, or precipitated albuminous particles from the plasma. LEUCOCYTOSIS. Nomenclature. — An increase of leucocytes distinctly above normal limits is called leiicocytosis. When the increase affects principally the polynuclear leucocytes, the condition is termed polynudear leucocytosis or simply leucocytosis. An increase of lymphocytes alone is called lymphocytosis. Sometimes, several varieties of cells are found in in- creased numbers, when the term mixed leucocytosis is used. Eosino- philia is perhaps an admissible word to designate an increase of eosino- phile cells. A distinct diminution of leucocytes is either so phrased, or termed hypoleuGocytosis, in contrast to hyperleucocytosis which signifies an ex- cess of cells. Classification of Leucocytoses. — Rieder first employed the follow- ing natural divisions of the subject : I. Physiological Leucocytoses. Leucocytosis of digestion. Leucocytosis of pregnancy. Leucocytosis of the new-born. II. Pathological Leucocytoses. Post-hemorrhagic leucocytosis. Cachectic leucocytosis. Ante-mortem leucocytosis. Inflammatory leucocytosis. Each variety of leucocytosis has a somewhat particular significance. That of the inflammatory type will first be considered. Significance of Inflammatory Leucocytosis. — It was early seen that the most typical occurrence of transient leucocytosis is limited to certain acute infectious diseases and here its significance has been the subject of much study and discussion, the extent of which has shown this subject to be one of the broadest of biological problems. A very simple illustration of a principle believed to exert prominent influ- ence in the behavior of leucocytes is found in the actions of certain myxo- mycetes in the presence of chemicals. Stahl found that Ethalium septicum placed on a moistened surface close to a drop of infusion of oak bark moved 116 THE LEUCOCYTES AND LEUCOCYTOSIS. actively toward and into the infusion, but moved actively away from a solution (5-percent) of glucose'. He also noted that the plasmocUum of Fuligo, which at first moves away from a 2- percent solution of salt, will later, if needing water, advance and enter the solution. Pfeifter called this attractive and repulsive influence of chemicals upon protozoa chemotaxis, and the above observations may serve to illustrate positive and negative chemotaxis and the transformation of a negative into a positive chemotaxis. PfeifFer also extended these observations to bacteria, finding that nearly all nutritive substances exert a positive attraction for bacteria, which is determined not by movements of diffusion of the liquids, but by the specific nature of the chemical substances. A chemotactic influence of bacteria upon leucocytes apart from the phagocytic tendencies of MetchnikolF was pointed out by Peckelharing, among the first, who inserted beneath the skin of frogs pieces of parchment containing anthrax, and found them after some hours surrounded by leuco- cytes. Gabritschewsky 2 showed that the attractive agent was dissolved in the fluids of these cultures. An extensive mass of observations has accumu- lated, showing that there is a wide variety of bacteria and bacterial and chemical products which exert a specific positive attraction upon leucocytes. In some instances a preliminary repulsion is succeeded by a positive chemo- taxis, but the examples of distinct negative chemotaxis of bacteria upon leucocytes are so uncertain as to lead Kanthack to conclude that probably no bacteria actually repel leucocytes, though some very rapidly destroy the attracted cells, so that the surrounding area is comparatively free from them. An apparently clear example of negative chemotaxis is found in the action of the bacillus of chicken-cholera in some susceptible animals, while there are a great many examples of failure of positive chemotaxis {Bacillus tuber- culosis, tyjyhosus). A distinctly new element Avas added to the knowledge of chemotaxis when it was first clearly shown, by Leber, 1 that chemotactic action of bac- teria is exerted upon leucocytes at a distance. On placing within the anterior chamber of the eye of the rabbit a fine tube filled with a substance extracted from cultures of Staphylococcus aureus, Leber found that the leucocytes would promptly overcome the action of gravity and force a way for a considerable distance through lymph channels, in order to enter the tube. The subject was again considerably enlarged in the experimental produc- tion of leucocytosis by Limbeck, who pointed out that along with the local afflux of leucocytes following subcutaneous injections of bacterial cultures, there is a marked increase in the general circulation, which in fact precedes the local changes. This action upon distant blood producing organs causing an outpour of leucocytes has been variously interpreted. It has been referred by most authors to the direct chemotactic action of the virus distributed through the circulation. Limbeck regarded the intravascular leucocytosis as somewhat in the nature of a forerunner and integral part of the purulent exudate which gathers at the point of inoculation. Weiss carries this idea still further, actively maintaining that leucocytes in inflammatory conditions are formed in various tissues whence they pass into the blood. Leucocytosis is therefore not of uniform significance and origin, but is the expi'ession of various underlying tissue changes in disease. Lowit believed that the increase of leucocytes merely represents a regenerative effort on the part of the marrow to replace destroyed leucocytes, his observations supporting Virchow's original view of the origin of leucocytosis. Buchner and Roraer also refer leucocytosis solely to the increased production of cells under the stimulus of bacterial proteins. That an increased flow of lymph occurs in many cases of in- flammatory leucocytosis, especially in that in which the lymphocytes are increased, is indicated by the demonstration of a special lymphogogic property in bacterial proteins. (Buchner, Gartner and Romer.) Schulz, Rieder, and Goldscheider and Jacob, have held more or less SKiSIFICANCE OF INFLAMMATORY LEUCOCYTOSIS. 117 strictly to the opinion that the leucocytes are not greatly multiplied during leucocytosis, but on the contrary are merely overabundant in peripheral vessels or to a moderate extent are drawn into the circulation from the depots of ready-formed cells in the marrow. It is impossible here to consider in detail the basis of these various hypotheses. Of the hist mentioned it must be said that there is abun- dant evidence both of a new production of leucocytes in the viscera and of an increased outpour of these cells during leucocytosis ; that Schulz's unconditional statement was erroneous, and that in so far as later ob- servers have held his views they too remain in error. Although the evidences of new formation of leucocytes in the cir- culating blood, lymph nodes, and marrow, found by Schmidt, Lowit, and Rieder, seemed to them inadequate to account for a great increase in the number of these cells, this result may be referred to the uncer- tainty regarding the origin of leucocytes, to the mistaken opinion that polynuclear leucocytes are deriv^ed from lymphocytes, and to the diffi- culty of demonstrating early proliferative changes in such widely distrib- uted fields as the blood and marrow. Yet Lowit" could find consider- ably increased numbers of mitotic and amitotic nuclei in circulating leucocytes during leucocytosis. Recent studies of the marrow during experimental leucocytoses have fully established the existence of greatly increased cellular prolifera- tion beginning in the early stages of leucocytosis, and, in prolonged cases, completely transforming the histological appearance of this tissue. The writer in 1895 compared the marrow of typhoid fever with that of pneumonia and other exudative diseases, in a series of cases, and found a very striking difference in the pulp cords in the two condi- tions, leucocytosis being accompanied by marked cellular hyperplasia, affecting principally the neutrophile myelocytes. Rogers and Josue found, 48 hours after the injection of Staphylococ- cus aureus into rabbits, marked congestion and cellular hyperplasia of the marrow cords affecting principally the nucleated red cells and eosinophiles, but by the third day these cells disappeared and neutro- phile myelocytes with many giant cells formed the bulk of the now greatly hypertrophied cords. The sole ground that remains to those who deny that inflammatory leucocytosis is essentially a new formation of leucocytes, is the proba- bility that many leucocytes drawn into the circulation during the first hours of leucocytosis are ready-formed in the marrow. Of the other hypotheses it still appears that each contributes a por- tion of the truth. Probably the combined influences of chemotaxis, and of the new formation of leucocytes stimulated by bacterial proteins, etc., are chiefly responsible for the general afflux of white cells. The destruction of leucocytes has repeatedly been shown to be a subordi- nate fiictor and entirely too slight to stimulate the production of new cells through ordinary channels, as suggested by Lowit. Limbeck's early belief in a close connection between local exudate and general leucocytosis is at least partially true, and serves to empha- 118 THE LEUCOCYTES AND LEUCOCYTOSIS. size as Grawitz points out, that many tissues, to an extent, and not merely the blood-forming organs, furnish leucocytes to the circulation. In so far as chemotactic influences are concerned in leucocytosis it should be remembered that the essential nature of the attractive force remains entirely unknown. In studies upon the visceral changes in leucocytosis the writer be- came convinced that some phenomena now attributed to an obscure chemotactic influence may be fully explained on purely mechanical principles. This statement applies especially to the sifting of cohesive leucocytes by swollen capillary endothelium, and to a less extent to the outpour of new-formed leucocytes from the marrow, but in the phe- nomena of local exudates other factors seem undoubtedly to be con- cerned. The Course of Leucocytosis. — The experimental study of leucocytosis has greatly increased the knowledge of the order and significance of the phenomena connected therewith. It was first observed by Werigo that the inoculation of animals with bacterial cultures is followed within 5-10 minutes by a pronounced diminution of circulating leu- cocytes {hypoleucocytosis), by the deposit of the bacteria in the capil- laries of the lungs, liver, and other viscera, and by their complete dis- appearance from the blood. Approaching the subject from another standpoint Goldscheider and Jacob, and the writer, found that along with the deposit of bacteria, the leucocytes are also sifted out of the circulation and permanently lodged in visceral capillaries, principally in the lungs and liver. The writer also partially agreed with Lowit that actual destruction of leu- cocytes follows such intravascular inoculations, especially with large doses, but did not find, contrary to Lowit, that leucocytolysis in dis- ease plays more than a very subordinate part in ridding the blood of white cells. The initial hypoleucocytosis has been observed very con- stantly to precede the appearance of leucocytes in increased numbers, and the same phenomenon has been observed in the course of many natural infections in the human subject. In experimental studies it has been found that many animals perish during the stage of hypoleucocytosis, especially when an active virus is introduced into a susceptible animal. Thus the inoculation of guinea- pigs with virulent cultures of diphtheria is followed by death within 24- 48 honrs, duriiig 2^ersistent hypoleucocytosis, while in the less susceptible rabbit the same treatment usually kills the animal but not before leu- cocytosis has become well established. In general, the increase of leu- cocytes follows the initial diminution more rapidly in those animals in which the course of the infection is more favorable. The same rule is found to hold in the human subject, initial hypoleucocytosis, sometimes persistent, having been observed in malignant cases of pneumonia (Kikodse et al.), diphtheria (Billings), septicemia, etc., and its unfavor- able import in these diseases has been fully demonstrated. In infections of ordinary virulence hypoleucocytosis is established within one-half to two hours after the initial diminution, and Gold- SIGNIFICANCE OF INFLAMMATORY LEUCOCYTOSIS. 119 scheider and Jacob succeeded in exciting pronounced leucocytosis with- out any demonstrable initial decrease in cells. During the course of febrile reaction to exudative processes, and sometimes during a con- siderable period thereafter, hyperleucocytosis persists. The grade of leiiGocytosis has been found to vary often with the height of the temperature, more closely with the extent of the exudate, but as might be expected from its recognized significance, measures more exactly the reaction of the system, to the infectious agent. Throughout the stages of inflammatory leucocytosis it is the polynu- clear leucocytes which are chiefly or exclusively concerned with the process. During hypoleucocytosis these cells, and as the writer finds, the large mononuclear cells, are greatly reduced, having been caught in the capillaries, while the lymphocytes remain in large proportion. Very soon, with the increase in total numbers, the poly nuclear cells are found in excessive proportions (80-95 percent), and even when the total increase is not great the excess of polynuclear cells is so constant as to furnish at times significant information regarding the character of the infection. The hyperemia and excitation of the marrow is indicated by the presence of normoblasts (Timof jewsky) and of a few myelocytes which have now been found in many severe leucocytoses, especially in diph- theria (Engel). An absolute increase of lymphocytes, sometimes of extreme grade, as observed by the writer in diphtheria, is often noted in the course of inflammatory leucocytosis and is to be referred to special involvement of lymphoid tissues. Ehrlich ^ holds that while polynuclear leucocytosis is referable to chemotactic influences acting from a distance, lymphocytosis is always the result of local irritation acting mechanically. Relative or absolute lymphocytosis is seen also in digestion leucocytosis (Rieder), in typhoid fever, with tumors of bone, and, in general, is a frequent feature of in- flammatory leucocytosis in children. During the subsidence of acute leucocytosis the neutrophile are sometimes replaced by a considerable proportion of eosinophile cells, as in pneumonia, septic processes, etc. The significance of this phe- nomenon is entirely unknown, but it has been shown to be of favor- able import. The behavior of the different cells in leucocytosis is largely in accord with their known characteristics elsewhere. The actively ameboid and phagocytic cells are chiefly affected, while the non-ameboid lympho- cytes are much less and only mechanically disturbed. An exception to this rule is found in the actively ameboid eosinophile cells, but these are rarely found to show phagocytic qualities. The bearing of these facts upon the significance of leucocytosis will be considered later. Relation of Leucocytosis and Phagocytosis to Immunity. — The studies of Metchnikoff upon phagocytosis in lower animals have not only es- tablished the great importance of this process as a protective measure 120 THE LEUCOCYTES AND LEUCOCYTOSIS. in the protozoa but have gone far to explain the significance of leiico- cytosis in the higher raetazoa. There is no room to doubt that monads, for example, englobe, digest and destroy living threads of kptothrix, and the steps through which Metchnikoff has traced the process seem to show that in animals as high as frogs, the leucocytes alone are re- sponsible for the englobement and destruction of anthrax bacilli. In the higher vertebrates, however, other factors may very well super- vene, and before accepting the phagocytic doctrine of leucocytosis in the human subject it is necessary to consider in some detail the extent and limitations of the process in higher vertebrates. The process of phagocytosis, as described by Metchnikoff, begins with the cheraotactic attraction of leucocytes toward bacteria, the englobement of the germs, often in the living virulent condition, and their digestion in stages which may be followed b^' changes in form and staining quality. The power to approach and englobe bacteria is obviously found in the ameboid activity of leucocytes. The power to digest bacteria is indicated by the morpho- logical and chemical changes above mentioned, which include the breaking up of a bacillus into granules and its gradual loss of staining capacity. Moreover a digestive ferment has been demonstrated in leucocytes by Ross- bach, and Leber, 2 and many investigators, especially Buchner,^ have dem- onstrated very high bactericidal power in exudates rich in pus. That leu- cocytes actually exert these bactericidal powers in infectious diseases is strongly indicated by the different behavior of these cells in refractory and susceptible animals. It has been repeatedly shown that the leucocytes of susceptible animals which do not naturally attack certain bacteria may be made to do so by the immunization of the animal, as in the infection of mice and guinea-pigs with anthrax, and pigeons and rabbits with chicken cholera. That the englobed bacteria are alive and active has been shown by their movements within the surrounding vacuoles of the leucocytes, and by the subsequent development of the englobed bacteria when the leucocytes are placed in bouillon. (Anthrax in pigeons, Vib. Metch. in immune guinea-pigs.) Indeed, an increase in virulence has been found in cultures obtained from some bacteria passed through the leucocytes of refractory animals. More- over Lubarsch has shown that frogs' leucocytes are more actively attracted by living bacteria than by the same germ when dead. It should be noted that drugs which inhibit the activities of leucocytes have been found to greatly accelerate the spread of certain infections. (Opium in cholera, Cantacuzeue.) The histological study of the reactive process at the seat of inoculation of a virus under various conditions has, also, furnished very strong support of the belief in the essential importance of phagocytosis. It has repeatedly been shown that in susceptible animals injections of highly virulent bacteria are, in general, followed by serous and bloody exudates which fail to limit the infection which becomes general and kills the animal, while the same treatment of immune animals is followed by highly purulent exudates, the limitation of the infection, and recovery. (Gabritschewsky.^) Finally, the results of minute study of the behavior of various phagocytes should be considered. The two actively phagocytic leucocytes, the polynu- clear neutrophile and large hyaline cells, do not attack all bacteria in the same manner. The former are less selective, englobing nearly all forms of bacteria. Yet the bacillus of leprosy is found almost exclusively witliin the large hyaline cells which do not attack streptococci or gonococci. Kanthack and Hardy have made important observations in this field. Examining a hanging drop of frog's lymph inoculated with a few anthrax bacilli, they found that the eosinophile cells were first attracted to the bacilli, that they discharged their granules upon coming into contact witli the germs, and that STONIFICANCE OF INFLAMMATORY LEUCOCYTOSIS. 121 de.s:enerative changes soon appeared in the bacteria. Later the large hyaline ceils surrounded the mass, the ox_yphiles dropped away, and the altered bacteria were englobed and digested by the hyaline leucocytes. These ob- servers found the same division of labor among leucocytes in various animals up to man, the hyaline cells acting as phagocytes of the celomic and lym- phatic systems, and the finely granular oxyphile cells as phagocytes in the hemal system. They ascribe important bactericidal functions to the gran- ules of eosinophile leucocytes in prepai'ing germs for englobemeut by the ameboid phagocytes. Hankin also came to identical conclusions regarding the importance of eosinoiihile cells in phagocytosis in the rabbit. Based upon these and many other similar observations, verified by many but actively combated by others, Metchnikoff concluded that phagocytosis is the essential feature of inflammation, and the chief mecha- nism in immunity, and gave exclusively to the leucocytes and the endothelial and other phagocytic tissue cells a function of vast importance. It has been necessary, however, to modify the original claims of Metchnikoff in an important particular, that an extracellular influence is exerted, upon bacteria which is sometimes a necessary preliminary to phagocytosis. The steps by which this fact has been demonstrated may be briefly indicated. Traube, Fodor, and others, recognized that the destruction of bacteria in the living blood is often extremely active, and it was suggested that the fluid plasma, and not the leucocytes and eudothelia, is the chief agent in freeing the animal organism from infectious germs. Grohman then demon- strated the bactericidal action of shed blood. Nnttall, next, found that during the process of inflammation, about the seat of inoculation in rabbits infected with attenuatad anthrax, many bacteria which are not englobed nevertheless show degenei'ative changes, and he also found that blood serum and lymph possess very active bactericidal powers. These observations were verified and greatly extended by Nissen, Hankin, Behring, Buchner, and Vaughan, who have fully established the presence both of bactericidal and of antitoxic principles. Some of these observers have gone so far as to deny any great importance in phagocytosis, and they may be said to represent the " humoral theory" of immunity. It soon became apparent, however, that there exists a very close connec- tion between these "defensive proteids " of blood and exudates, and the leucocytes. Hankin, Buchner, Kossel, and Vaughan have shown that a nucleo proteid derived probably from the nuclei of leucocytes is the bacteri- cidal agent in serum, exudates, and lymph. This bactericidal proteid belongs to the class of enzymes, is extremely unstable, is destroyed by heat- ing to 55-65° C. , and has been called by Buchner " alexine." The existence of a distinct compound of this character has not been accepted by all phj'si- ological chemists. According to Hankin, blood serum ordinarily contains little or none of this agent, but after the second day of leucocytosis it is secreted by the eosinophile cells of rabbit's blood or is discharged from these cells during defibrination. Hahn and Van de Velde have shown rather conclusively that the alexines are a secretory product of the living leucocyte, and are not produced by the destruction of these cells, although they may be set free thereby. Hankin, Buchner, 3 Hahn, Bordet, Shattenfroh, Bail, and others have established the essential relation between the bactericidal power of exudates and their content in leucocytes, and have placed this relation be- yond doubt by such demonstrations as that of Denys and Havel, that the blood and exudates of dogs lose their bactericidal power when freed by fil- *?: 1 122 THE LEUCOCYTES AND LEUCOCYTOSIS. tration or the ceatrifu2:e from intact leucocytes, but regain this power when the leucocytes are replaced. The controversy regarding the so-called ^ ^ phenomenon of Pfeiffer^^ has also ended in a way to favor the importance of leucocytes in the defense of the organism. PfeifFer found that when cholera cultures are injected into the peritoneum of highly immunized guinea-pigs, the germs are broken up into granules and to some extent dissolved by the peritoneal fluid before leucocytosis and phagocytosis have become fully established, but not, as some citations of this work would lead one to suppose, before the leucocytes present are actively engaged in phagocytosis. Metchnikoff 2 promptly repeated this experiment, finding that the cholera spirillum flourishes in a hanging drop of the same peritoneal fluid which in the body is bactericidal, that the peritoneal fluid is really very rich in leucocytes, that if a great afflux of leucocytes is previously excited by injec- tion of bouillon the phenomenon of PfeitFer is suppressed, but the germs are englobed and destroyed just as quickly by the leucocytes, and he found, in general, that the phenomenon of Pfeiffer occurs only in animals whose leucocytes possess very active bactericidal and phagocytic qualities, and only in those situations where leucocytes are or have been abundant. The discovery of the "phenomenon of Pfeiffer" placed beyond question that bacteria may be destroyed by entirely extra-cellular processes, a fact now not infrequently illustrated in the dissolution of bactei ia in Widal's test. While Metchnikoff" accepted this fact and altered his views as to the inva- riable necessity of phagocytosis, the application of his doctrine was ex- tended rather than narrowed, for it still remains true that the bactericidal principles are elaborated principally in the leucocytes. The foregoing review of the steps through which the present con- ception of phagocytosis and immunity has been reached, while defec- tive in important and highly interesting details has been rendered neces- sary in order to set in its proper light the significance of inflammatory leucocytosis. It is now apparent that leucocytosis represents Nature^s attempt to rid the blood and the system, by means of leucocytes and their products, of the bacterial and toxic causes of disease. It is now possible, also, to understand the varying significance of hypoleucocytosis, and to place correct interpretations on the various grades of hyperleucocytosis in infectious diseases ; to trace the relation between local and intravascular leucocytosis, and explain many of the phenomena of exudative inflammation ; to properly interpret the pre- dominance of mononuclear, or of polynuclear, and to some extent, of eosinophile cells, and to recognize the specific quality of the diiferent varieties of leucocytes ; Avhile the application of all these facts to ques- tions in general pathology has been most extensive. Clinical Types of Leucocytosis. Digestion Leucocytosis. — Although digestion leucocytosis was one of the earliest of observed phenomena regarding leucocytes (Nasse, Yirchow), owing to the uncertain methods employed by some later ob- servers, and to the subsequent discovery of other marked physiological variations in the numbers of the cells, the occurrence of a leucocytosis referable to digestion was not fully accepted until Pohl, in dogs, and DIGESTION LEUCOGYTOSIS. 123 Limbeck and R. Muller, and Rieder, in man, investigated the subject with the necessary precautions against possible disturbing factors. It is now evident that in order to demonstrate digestion leucocytosis it is necessary to know the normal percentage for each individual, and not to rely upon general averages, as the effect of digestion is not great, and the leucocytes at periods well distant from meals vary con- siderably according to the health, age, state of circulation, nutrition, and digestion, of the individual. With these precautions it has been found that a full meal raises the number of leucocytes in the average healthy subject about 33 percent (Rieder), beginning certainly within 1 hour, reaching a maximum in 3—4 hours, and thereafter gradually declining. The exact figure reached cannot have much significance, for the reasons above stated, but in adults the maximum is not often above 15,000-16,000. In healthy fasting subjects the increase is rarely over 3,000 cells, but in invalids the variations are slightly greater. A failure of digestion leucocytosis in healthy adults is sometimes ob- served and may be referred to a prolongation of the process, or to chronic torpidity of the intestines, or to accidental causes. The quality of the food has sometimes a distinct effect on the grade of leucocytosis. Highly albuminous meats, readily digested and ab- sorbed in considerable quantity, have much more influence upon leuco- cytes than a meal of vegetables and fats. Indeed, although Duperie has reported digestion leucocytosis after exclusively vegetable diet, it is usually not seen in any vegetarian animal, probably because of slower digestion and absorption. Of the disorders in which digestion leucocytosis may be expected to fail the mildest are those marked by torpidity of stomach and bowels, which lengthens the time, and diminishes the completeness of digestion and absorption. (Rieder.) R. Muller found it difficult to excite leucocytosis in anemic subjects, except by very large meals, until the general condition of the patient was much improved, and Rieder also observed no increase but even a diminution of white cells in simi- lar cases. Among various hospital cases confined to bed from chronic complaints the leucocytosis sometimes exceeded the normal limits, sometimes was absent. When from any other cause the leucocytes are distinctly increased, as in pregnancy, inflammatory diseases, etc., digestion usually fails to pro- duce a further demonstrable increase. In carcinoma of the stomach, R. Muller first called attention to the absence of distinct digestion leucocytosis which marks this condition, finding that without regard to the presence of HCl or pepsin, these cases do not show a pronounced increase of leucocytes after meals. From later reports it appears that about 90 percent of cases of gastric cancer fail to show any distinct digestion leucocytosis (2,000-3,000) although there are some clear exceptions to the rule, and more examples of a slight increase. This failure of leucocytosis is apparently inde- pendent of the presence of HCl or pepsin, occurs when stenosis does 124 THE LEUCOCYTES AND LEUCOCYTOSIS. not exist, and is referred by Schneyer to involvement of the neiglibor- ing lymphatics, and to complicating gastric catarrh. In very advanced cases it is perhaps invariably absent. (Hartnng.) In various other conditions sometimes simulating gastric cancer, as benign stenosis, ulcer of stomach, chronic gastric catarrh (Schneyer, Capps, Cabot), as well as in carcinoma of other viscera (Hartnng), well-marked di- gestion leucocytosis is the rule. As a dlagnost'iG test, the absence of digestion leucocytosis has not been found to give as reliable evidence as was first hoped, because, as recently shown by Sailer and others, it is too frequently absent in other conditions. As evidence against cancer, the presence of diges- tion leucocytosis is, however, of considerably greater value, though by no means positive. Nevertheless the examination of the blood de- serves a place in the diagnosis of this often obscure disease. In children, up to 15 years of age, most observers have found more marked digestion leucocytosis than in adults, especially after a meat diet. (v. Jaksch, Rieder.) The first digestive activity in the new- born infant excites a considerable leucocytosis, Schiff" finding an in- crease to 19,000-27,000 one hour after the first nursing, and 36,- 000 at the end of 48 hours, after several nursings. Between the tenth and fifteenth years Rieder found leucocytoses of 3,000-9,000, the average increase being greater than in adults. In the aged it is some- what less than in adults. (Pohl.) Origin of Digestion" Leucocytosis. — The essential fiictor in di- gestion leucocytosis is, without doubt, the absorption of a considerable quantity of albuminous principles. This absorptive process has been shown by Hoffmeister to excite a considerable poliferation of mononu- clear cells in the adenoid tissue throughout the gastro-intestinal tract. It would seem that sucli increase of mononuclear cells must necessarily cause an increased number of these cells to reach the circulation, and Pohl in fact, reported finding tlie mesenteric veins during digestion in dogs, much richer in leucocytes both mononuclear and polynuclear than the arteries. Rieder, however, could not find such a dispropor- tion between veins and arteries, and in the absence of other experi- mental data it becomes necessary to suppose that the marked increase of lymphocytes which characterizes digestion leucocytosis is referable to an increased outpour of lymph from the thoracic duct, the occur- rence of which during digestion there are obvious reasons for accepting. Since the proportions of mononuclear and polynuclear cells are not greatly disturbed in digestion leucocytosis, the marrow must furnish a considerable number of neutrophile leucocytes, and the chemotactic properties of the absorbed albumens must be regarded as the chief cause of their outpour. Regarding the exact proportion of different cells in digestion leuco- cytosis data are not numerous but sufficient to show that this is a *' mixed leucocytosis," both lymphocytes and polynuclear cells being increased, more especially the lymphocytes. The eosinophiles are usually reduced. (Rieder.) LEUCOCYTOSIS OF PREGNANCY. 125 Leucocytosis of Pregnancy. — The occurrence of moderate leucocy- tosis in the latter niontlis of pregnancy, though previously known, was first fully studied by Halla who found a marked increase in 10, a slight increase in 6 (12,000-13,000), and no change in 3, out of 19 cases examined. Rieder eliminated any possible effects of digestion, allowing his patients to fast 16 hours and found a leucocytosis, averaging 13,000 cells, in only 21 of 31 cases, all the exceptions being multiparte. Of 17 multiparse leucocytosis was absent in 10. Limbeck found 11,000 -13,000 cells in each of four cases examined, and Cabot reports un- usually high counts (25,000-37,000) in 3, and the ordinary increase in 9 cases. Hubbard and White report polynuclear leucocytosis in 80 percent of their 55 cases, most marked and constant in young primiparte. The averages 24 hours before labor, were, primiparge 15,000, multipara 11,700. Before the end of the third month no leucocytosis is to be expected (Rieder), but the exact period of its ap- pearance has not been determined. Rieder made the interesting ob- servation in six cases that in the ninth mouth of first or subsequent pregnancies, whether leucocytosis exists or not, digestion causes a diminution instead of an increase of white cells. After parturition, the leucocytes gradually diminish, reaching the normal usually in 4-14 days, but this diminution is frequently inter- rupted by slight disturbances referable to the repair of marked erosions or tears of the genital tract, to mild disorders of the breasts, and to the loss of blood. During and immediately after childbirth a considerable increase of leucocytes has been observed in a moderate num- ber of reported cases. (Kosina and Eckert, Malassez, Fouassier, Rieder.) The leucocytosis of pregnancy has been referred to various causes, none of which seem to meet all the requirements. Virchow found a parallel between the increase of leucocytes and the widening of the uterine and abdominal lymphatic vessels and nodes and the increase of metabolism in the uterus and its contents. Mochnatschetf 's compari- sons of the numbers of leucocytes in the finger-blood and in the cervix uteri do not seem calculated to demonstrate an increased supply of any variety of cell from the latter region. Limbeck suggests that the changes in the breasts are the chief factor, recalling the round cell in- filtration which many such glands show, as an indication of very active cellular processes. He, with others, have supposed that this leucocytosis, like that of the new-born, may represent a continuous eifect of digestion, an opinion which cannot be accepted since digestion leucocytosis is usually suppressed in late pregnancy. Considering the behavior of leucocytes in general, it seems hardly a matter of surprise that the active cellular processes in the breasts, uterus, vascular sys- tem, and fetus, and the associated increase of metabolism should, when instituted for the first time, find a sympathetic excitement in the blood- producing organs. Being a " mixed leucocytosis," with all but eosinophiles normally represented, this leucocytosis is probably not either inflammatory or toxic. 126 THE LEUCOCYTES AND LEUCOCYTOSIS. The Leucocytosis of the New-born. — That the blood of new-born infants invariably shows a well-marked leucocytosis was noted by many early observers, in greatest detail by Hayem, who found an average at birth of 16,000-18,000, a rapid decrease to 7,000 toward the third to fourth day, when the initial loss of weight becomes most marked, and a subsequent increase to 9,000-11,000, with beginning increase in body weight. In children at 8 months he found 14,000- 21,000 white cells, while by the fifteenth month the numbers commonly dropped to 10,000, and he noted also that the increase was largely in the nature of a lymphocytosis, the small lymphocytes being 4-5 times as numerous as in adults. These main features of the leucocytosis of the new-born appear again in Rieder's observations, who found slightly greater numbers, 14,200-27,400, in a few subjects, in whom the initial decrease on the third to fifth day was also well marked and apparently unaffected by the first demands on digestion. Rieder found the leucocytosis of the child uniformly greater than that of the mother, and, except for the preliminary decrease of the first week, much more persistent. His differential counts show that the lymphocytes are not in excessive pro- portion at birth, but that this excess becomes established later. (So also Gundobin.) His cases showed at birth a considerable excess, also, of eosinophile cells. Schiff examined 11 infants twice daily for two weeks, his results showing excessive variations in the numbers of leucocytes in the first few days (10,000-36,000), referable apparently to overfeeding and diarrhea. The preliminary decrease beginning on the third to fifth days is quite noticeable in his tables. At the end of two weeks the leucocytes num- bered 10,000-15,000. Gundobin's observations show that the leucocytes remain high dur- ing the first year of infancy (11,000-14,000), that the excess of lymphocytes is established before the tenth day and persists through- out the first two years (50-65 percent), when they begin to decline, reaching the proportions of the adult about the eighth to tenth year. Observations on the Leucocytosis of the New-born. Hayem average for 48 hours 18,000 " third to fourth day 7,000 " after fifth day 9,000-11,000 Otto ten to twenty-five hours.... 23, 000-25, 000 Schiff 48 hours 10,000-32,800 " fourth to eighteenth day... 12, 000-13,000 Woino-Oransky.. at birth 16,980 " " ...14 hours 20,980 " " ...second day 31,680 Gundobin last day, fetus 8,053 " at birth 19,600 " 24 hours 23,000 " 48 hours 17,500 Kruger one to three days 18,000 " after third day 15,000 POST-HEMORRHAGIC LEUCOCYTOSIS. 127 Rieder at birth 14,200-27,400 " second to fourth day 8,700-12,400 " after fourth day 12,400-14,800 From the above conflicting reports it is difficult to give precise de- tails regarding the leucocytosis of the new-born. Rieder's observa- tions seem to the writer the most reliable, in which a moderately high proportion of white cells is found at birth, a distinct decrease on the second to fourth day with the diminution of weight, followed by a less marked but persistent leucocytosis. Further very careful studies are needed in the field. Regarding the origin of this form of leucocytosis opinions are at variance. That the initial increase is referable to concentration of the blood and venous stasis seems to be true both of red and of white cells. The preliminary decrease the writer believes to be due to im- provement in the venous circulation with relief of cyanosis, and the absorption of fluids. The subsequent increase and permanent leucocy- tosis of infancy, being largely a lymphocytosis probably represents to a considerable extent continuous digestion leucocytosis. That it is not entirely referable to digestion is indicated by the low counts found by Rieder on the second to fourth days, after frequent nursing. Post-hemorrhagic Leucocytosis. — A well-marked leucocytosis fol- lowing acute hemorrhage was described by the older writers (Virchow, Nasse, etc.), but the detailed study of this condition was reserved until Lyon and Huhnerfauth investigated the effects of hemorrhage by the newer methods. Huhnerfauth found in 9 dogs, losing about 4 percent of their body weight by acute hemorrhage, that the leucocytes were often slightly diminished immediately after the operation, much increased (maximum 45,000) on the following day, and that this increase persisted in some degree for two to three weeks. Lyon, in similar experiments, also found an initial decrease a few minutes after the hemorrhage, but soon a very marked increase of the white cells, reaching its highest point (62,- 000) within a few (6-8) hours, declining rather rapidly after three to four days, but persisting thereafter in moderate degree for days or weeks. In the human subject after a surgical operation attended with danger- ous hemorrhage, Lyon found, after one hour, 41,625 leucocytes, after five days 14,300, followed by a slight increase for one week. In a case of leukemia, splenectomy caused the leucocytes to jump from 463,- 000 to 850,000 several hours before death from hemorrhage and shock. Rieder repeated the experiments on dogs finding the usual changes, as a rule, but in one instance a large hemorrhage failed to cause leuco- cytosis. In differential counts he observed that most of the leucocytes were poly nuclear, finding as high as 97 percent of neutrophile cells immediately after the hemorrhage. In the human subject he was un- able to find very pure examples of post-hemorrhagic leucocytosis, but re- ported 15,000 cells after pulmonary hemorrhage in phthisis ; 32,600 (1,300,000 red) after hemorrhage from cancer of the uterus; 26,500 (1,985,000 red) in ulcer of stomach. 128 THE LEUCOCYTES AND LEUCOGYTOSIS. Ill general, the leueooytosis following hemorrhage is in proportion to the extent and rapidity of the loss of blood, hut it usualli/ disappears or greatly diminishes long before the red cells are restored. The infusion of fluids has been found to considerably increase the leucocytosis. Tliis result has been repeatedly observed in cases of splenectomy (q. v.). The not infrequent failure of post-hemorrhagic leucocytosis as re- ported by Rieder, Stengel, Cabot, and others, has not been explained. In such cases the remaining cells may be largely lymphocytes. (Stengel.) Erb, however, reported marked absolute lymphocytosis after hemorrhage. It is generally agreed that the chief factor in the production of post-hemorrhagic leucocytosis is the outpour of lymph which restores the lost bulk of blood. Were this the only factor, however, the new cells ought to be principally lymphocytes, and the high proportion of neutrophile cells commonly seen points to a special draining of the marrow or possibly to increased cellular activity in this tissue. Against this view are the facts that normoblasts are not numerously present till the second to the third day, while the leucocytes are most numer- ous after a few hours. The old opinion of Samuel, Virchow, and others, that the more cohesive neutrophile cells are retained in the vessels is perhaps still worthy of consideration. After the infusion of salt solution the high percentage of polynuclear cells can be partly referred to the chemotactic influence of this fluid upon the marrow. Cachectic Leucocytosis. — Of the varieties of leucocytosis classified by Kieder that which is frequently associated with severe anemia and cachexia is the least homogeneous. In its causation figure many of the conditions already shown to frequently excite leucocytosis, espe- cially loss of blood and inflammatory processes. Yet there exists a considerable group of cases marked by continuous leucocytosis, in which other influences combine to increase the production of white cells, and to develop a somewhat peculiar form of leucocytosis which deserves a special description. White, Cruveilhier, Paget, and others before them, who first de- scribed cachectic leucocytosis, regarded the excess of cells as fragments of the neoplasm. Andral and Gavaret, Chaillou, and Vidal, recognized these cells as leucocytes. This correct interpretation was soon followed by the more detailed study of the condition by Lucke, Sappey, Nepveu, and especially by Hayem, Alexander, Schneider, Pee, Rein bach, and Rieder, each of whom have contributed largely to the present knowledge. Some of the clearest examples of this sort of leucocytosis are seen in the cachexia of malignant tumors, where the combination of hydremia, local inflammatory processes, and specific toxemia, are the principal factors concerned. Chronic anemia may lead to increase of leucocytes in the circulation through the watery condition of the blood, the increased activity of red marrow, here often moderately hyperplastic, and the lowered blood pressure. Illustrations of the working of these influences are seen in CACHECTIC LEUCOCYTOSIS. 129 the increase of post-he morrhagic leucocytosis after iafusion, and in the periodical appearance of normoblasts and mixed leucocytosis in ad- vanced chlorotic anemia. Yet chronic hyperemia and extreme hyper- plasia of the red marrow alone are not necessarily accompanied by leu- cocytosis ; but often exist with diminished leucocytes, as is usually seen in uncomplicated primary pernicious anemia. So that it is impossible to ascribe to chronic hydremia more than a predisposing influence in cachectic leucocytosis. Horbacewski and Lowit^ ascribe both the hydremia and the leuco- cytosis of cachexia to the lymphogogic and chemotactic influence of nu- cleins which in increased quantity are set free by the excessive destruc- tion of tissue elements, but Lowit claims from his experimental study of the subject that increased fiow of lymph alone is incapable of produc- ing leucocytosis. Local inflammations are probably the most frequent causes of marked cachectic leucocytosis. These may be traced in the growing edges or ulcerating surfaces of malignant tumors, in various internal suppurative or necrotic processes in syphilis and tuberculosis, and in exacerba- tions of underlying chronic inflammations. The special involvement of the blood-forming organs by neoplasms and inflammations has been found not only to increase the leucocytes in the circulation but at times to give a special character to the leuco- cytosis which may be of diagnostic service, e. g., eosinophilia and lymphocytosis with sarcoma of bone or lymph nodes. While sarcoma is usually attended with lymphocytosis, carcinoma on the other hand usually excites a polynuclear leucocytosis, a difference which may per- haps be referred to the freedom of the lymph paths in the former and their occlusion in the latter case. Finally there is to be considered as the essential element in cachec- tic leucocytosis a chronic toxemia, which, in several conditions, is of more or less specific character. Some importance here attaches to the observations of Hayem who noted the disappearance of persistent leucocytoses (10,000-21,000) in 4 cases after removal of cancers of the breast, while a recurrence of the tumor he claims to have predicted from a slight persistent increase of leucocytes. Grawitz has shown that injections of extract of carci- nomatous tissue and of tuberculin have a marked lymphogogic effect. The distinct cachexia which often foreshadows the appearance of a malignant neoplasm is abundant clinical evidence that these new growths elaborate powerful toxines but it has never been conclu- sively shown that these toxines have any particular influence upon leucocytosis, which in the majority of cases seems to arise from other factors. The same conclusion must be drawn from the absence of any uniform relation between cachectic leucocytosis and the extent of miliary tubercles and guramata, or the extent, location and histological struc- ture of malignant tumors. General miliary tuberculosis, acute or chronic, usually fails to excite leucocytosis. The writer has seen severe anemia without leucocytosis, with extensive gummata in liver, 9 130 THE LEUCOCYTES AND LEUCOCYTOSIS. spleen, and thoracic and abdominal lymph nodes, and in general car- cinomatosis. Indeed, in a review of many reported cases and in the writer's own experience, in the great majority of cases of tertiary syphilis, tuberculosis, nephritis, in a large proportion of carcinoinata, and in a rather smaller proportion of sarcomata, cachexia is unaccompanied by leucocytosis un- less there is distinct local inflammation, necrosis, or hemorrhage. Whence it may be stated as a general rule that marked leucocytosis in the course of cachexia, suggests a search for one of these complications. Ante-mortem Leucocytosis. — In 1883 Litten reported from sev- eral hundred examinations of the blood of 40 moribund cases in whom no leucocytosis had previously existed, that he had found in many a more or less pronounced increase of white cells, beginning at a varia- ble period before death. When dissolution was rapid no leucocytosis was found, but when prolonged, the increase of white cells was very marked (1-5 of red). Litten was unable to offer a plausible explanation of the phenomenon and failed to give exact numerical estimates and important clinical details of his cases, so that his interesting observations attracted but little notice. In 1886, Gottlieb found 30,000 leucocytes 6 hours before death in a case of pernicious anemia which had previously shown no increase. Rieder found well-marked ante-mortem leucocytosis in three of four cases in one of which 89.5 percent of 60,000 cells were polynuclear. The case not showing leucocytosis was complicated with aspiration- pneumonia. Many later reports of isolated observations on this variety of leu- cocytosis have proven the frequency of its occurrence, but have not added greatly to the knowledge of its special peculiarities or its sig- nificance. It appears most probable that a considerable variety of factors is concerned in the ante-mortem increase of leucocytes. Cohnheim's theory, that diminishing blood pressure causes an increased outpour of lymph is supported by the occurrence of post-hemorrhagic leucocytosis, but not by the high percentage of polynuclear cells and of nucleated red cells usually present. Litten's suggestion that there is an unequal distribution of leucocytes in favor of the peripheral capillaries is in accord with the known effects of stasis. Limbeck's opinion that ante- mortem leucocytosis probably results in most cases from terminal in- fections may certainly apply to many instances marked by excess of polynuclear cells, but not to all, especially to those rare cases showing lymphocytosis. The character of ante-mortem leucocytosis seems to depend largely upon the precedent condition. In leukemia and inflammatory diseases the cells principally increased are the polynuclear forms of the mar- row, and along with these many nucleated red cells and some mye- locytes are commonly drawn into the circulation. In the terminal stages of lymphatic leukemia tlie lymphocytes may be greatly increased. EXPERIMENTAL LEUCOCYTOSIS. 131 Pronounced ante-mortem lymphocytosis has been seen in pernicious anemia by Cabot and by the writer, in diphtheria and other conditions in children, in typhoid fever, and to a less extent in malaria. It would seem that the effects of terminal inflammations, ante- mortem dissemination of intestinal and other bacteria, ante-mortem hyperpyrexia, vaso-motor paresis, serous exudates, diarrhea, lack of fluids ordinarily ingested, etc., have not received sufficient attention in this important field, and that the entire subject needs investigation on a much more elaborate scale than has yet been attempted. Experimental Leucocytosis. 1. By Drugs and Chemicals. — The earliest observations on experimental leucocytosis were those of Hirt who, in 1856, reported an increase (100-300 percent) of white cells following the administration of Tr. ferri pom., Tr. myrrh., etc. The same effect was noted by Meyer from turpentine, camphor, and oil of peppermint. Later Pohl, in the endeavor to explain the origin of digestion leucocytosis tested the effects of a large number of drugs, finding that the aromatic extracts and oils, vegetable bitters, certain alkaloids (piperin, strychnia, and others) caused a rather distinct (-10— 120 percent) leucocytosis, when administered to fasting dogs. Alco- hols, salts of various alkalis, lead acetate, cupric sulphate, calomel, caffeine, and quinine, he found to be inert, and oxide of iron usually so. In general these leucocytoses appeared within one-half hour and disappeared in two hours, and were therefore less marked and per- sistent than those excited by the absorption of digested albumens. Repeating these experiments on man, Binz and Limbeck induced leucocytoses by means of camphor and oils of cinnamon, peppermint, and anise, but Rieder found it impossible to secure much effect from extract of gentian or tincture of myrrh. Even such a bland fluid as solution of common salt in subcutaneous injection has been shown to excite transitory leucocytosis. (Lowit, Rieder.) Bernard, and Meyer and Seegen found the leucocytes doubled after the administration of 20 gt. of ether. Horbaczewski " found a moderate diminution of leucocytes in man after administration of quinine and of atropine, and a considerable in- crease from antipyrine, antifebrine, and pilocarpine. The writer has failed entirely to produce distinct leucocytosis in rabbits by use of pilocarpine. Winternitz studied the effects of a great variety of drugs with special reference to the relation between the grade of local inflamma- tion and the degree of intravascular leucocytosis. He divided these agents into two classes, one of which, including salts and simple irri- tants such as free acids and alkalies, by subcutaneous injection, induced slight local disturbance, and moderate leucocytosis and fever, while the other, including vesicants, sapotoxin, digitoxin, silver nitrate, copper sulphate, mercurials, and antimonials, produced aseptic suppuration and higher leucocytosis. He therefore established for these agents the same relation between local reaction and general leucocytosis that Lim- beck has shown for bacterial cultures. r 132 THE LEUCOCYTES AND LEUCOCYTOSIS. 2. By Bacterial Cultures. — Limbeck injected cultures of bac- teria into the knee-joints of fasting dogs, and found the maximum leucocytosis at the end of 6-24 hours, 2-3 times the normal numbers being commonly reached, and 88-93 percent of the cells being poly- nuclear. The pyogenic staphylococci were most active, increasing the leucocytes 6-7 fold ; streptococcus pyogenes came next, and Fried- lander's pneumobacillus third, in effectiveness. Rieder repeated these experiments using cultures from solid media only. He showed that such material was very much less active in producing leucocytosis than are the fluid cultures, that the increase is preceded by a transient decrease of cells, and that the injections may be followed by persistent hypoleucocytosis and death of the animal. A sufficient number of observations have since shown that practically all pathogenic bacteria in subcutaneous or intravenous injection exert positive chemotactic influence upon leucocytes. The duration of the period of diminution and the degree of subsequent increase of leuco- cytes vary considerably with different species, cultures, and individuals. 3. By Bacterial Proteins, etc. — After Limbeck's demonstra- tion of the chemotactic effects of bacterial cultures, it was soon shown that various bacterial and other extracts are equally powerful excitants of leucocytosis, as cadaverin, putrescin, ptomaines of decomposing flesh, extracts of sterilized cultures of staphylococci, and phlogosin, a crystalline alkaloid obtained from similar cultures. (Scheurlen, Grawitz,^ Behring,^ Arloing, Leber.* ) The more complete separation of the active chemotactic principles of bacterial cultures was accomplished by Buchner following Nencki's method. Buchner found that the leucocytes are influenced solely by the albuminous principles of bacterial cultures and these he isolated from a considerable number of species, in the form of bacterial " pro- teins." From Bacillus pyocyaneus he secured a protein which, after 4 daily doses of 2 grams, increased the leucocytes seven-fold. Testing the effects of the decomposition-products of animal tissue, peptone, alkali-albumen, and calcium, he found these agents comparatively inactive and concluded that the principles derived from inflamed tissue are unimportant adjuncts to the bacterial proteins in exciting inflam- matory leucocytosis. Koch's tuberculin was early shown to cause at the height of the reaction a moderate and rather transitory increase of polynuclear leuco- cytes and a simultaneous diminution of eosins. (Uskow, Tschisto- vitch, Zappert.) Botkin found it impossible to increase the effects by repeated daily doses. A marked and persistent increase of eosinophile cells, reaching in one case 85 percent, and in another continuing for ten weeks, following the febrile reaction and frequently associated with cutaneous eruptions, has been noted by Grawitz,' Neusser,^ Canon, Botkin,^ and Zappert. A recent extensive study by Liebmann indicates that tuberculin is positively chemotactic to both leucocytes and the tubercle bacillus, drawing each into the circulation, in both man and animals. Lieb- EXPERIMENTAL LEUCOCYTOSIS. 133 mann could find no great uniformity in the behavior of eosinophile cells, but in rabbits and guinea-pigs tuberculin caused a marked in- crease of neutrophile cells, a moderate increase of mononuclear leuco- cytes, and the appearance of a moderate number of mast-cells. That leucocytosis follows the injection of a great variety of bacterial filtrates and extracts has been shown also by a large number of later experiments. The grade of leucocytosis varies considerably, depending on the quantity and virulence of the toxine, and the susceptibility of the animal. Rieder found the leucocytes increased twelve-fold after three daily injections of pyocyanin. Occasionally one encounters persistent hypo- leucocytosis, as did Romer after using large injections. 4. By Various Organic Principles. — The Dorpat school first studied the effect upon leucocytes of another group of agents, including several animal principles. They found hypoleucocytosis followed by hyperleucocytosis in marked degree after the injection of fibrin ferment, pepsin, peptone, hemoglobin, decomposing albuminous fluids, pus, and crushed lymphoid tissue. (Hirt, Bojanus, Hoffman, Himmelstjerna, Heyl, Groth.) Lowit studied the effects of various albuminous and organic prin- ciples, injected into animals, finding hypoleucocytosis followed by vary- ing grades of hyperleucocytosis from hemi-albumose, peptone, pepsin, nucleinic acid, nuclein, urea, sodium urate, curare, and pyocyanin and tuberculin. Vegetable albumens were first used to excite local purulent exudates by Buchner, who obtained aseptic pus in large quantities by the injec- tion into the pleura of ground wheat, pus, fossil earth, and gluten- casein. Rieder and Hahn also secured a high grade of leucocytosis by the same agents. Goldscheider and Jacob proceeded still further in the study of the chemotactic properties of animal extracts finding powerful effects from the injection of glycerine extracts of spleen, thymus, and marrow, but none from extracts of thyroid, liver, kidney, and pancreas. Applications of Experimental Leucocytosis. — In 1894 Paw- lowsky reported that he had saved animals from fatal doses of anthrax by injections of papayotin, abrin, and ricin, which are very active ex- citants of leucocytosis, and he recommended that anthrax in man be treated by local injection of these agents. In guinea-pigs and rabbits in which leucocytoses had been excited by papayotin, or better by abrin, injections of tubercle bacilli were less frequently followed by generalization of the tuberculous process, sometimes by greater ten- dency to fibrous growth in the tubercles, or even by total disappearance of tubercles already existing. In 1895 Lowy and Richter reported that rabbits in which they had established marked leucocytosis by repeated injections of " spermin " withstood injections of pneumococcus cultures 3—4 times as large as were required to kill control animals. The therapeutic effect was much less marked when the bacterial injections followed those of spermin within 24 hours or less. 134 THE LEUCOCYTES AND LEUCOCYTOSIS. Goldscheider and Mnller then took occasion to recall some unsuc- cessful experiments of their own performed on guinea-pigs, in which leucocvtosis had been excited by injections of splenic pulp extract. This procedure they had found to yield no therapeutic influence on in- fections by Bacillus diphtherim, tetani, proteus, or pneumococcus. Goldscheider and Jacob had also previously attempted, without success, to favorably influence by artificial leucocytosis the course of typhoid fever and puerperal sepsis. In animals which had received injections of albumoses Jacob found that subsequent infection with virulent bacterial cultures was better withstood if the infection occurred Avhen the leucocytes were on the increase from the previous injection of albumose, but when the in- fection occurred in the stage of hypoleucocytosis, it was more rapidly fatal than in the control animals. Considerable therapeutic effect was apparently obtained when the administration of albumose occurred during the increase of leucocytes resulting from the bacterial injection. Hahn was able to retard the progress of anthrax infection in rabbits by previously exciting leucocytosis by means of albumoses, but with a few favorable results were a larger number of failures. He went on to show that the bactericidal action of blood both in man and animals is increased during the stage of hyperleucocytosis. From the foregoing studies it is evident that there is a rational basis for the employment of artificial leucocytosis in the treatment of some infectious processes, since both the phagocytic and bactericidal powers of the blood are thereby increased. Yet the results in man are as yet not very favorable, since there is no known method of ex- citing with impunity continuous leucocytosis in the human subject. The limitation of puerperal septicemia by exciting localized abscesses by injections of vesicants, the extrusion of inoperable malignant tumors by erysipelas, and the treatment of typhoid fever with pyocyanin, il- lustrate some of the practical difficulties in the way of this branch of therapeutics. Bibliography. Leucocytes. Alexander. Th^se de Paris, 1887. Altmann. Die Elenientarorganismen. Andral, Gavaret. Traite de Hematoclinique, Paris, 1843. Arnold. Virchow's Archiv, Bd. 132, p. 502. Arloinrj. Compt. Rend. Acad. Sci., T. 106, p. 13G5. Bail. Berl. klin. Woch., 1897, No. 41. Behring. ' Zeit. f. 4Iyg., 1890, p. 424. ^Deut. med. Woch., 1888, p. 653. Bernard. Cited by Limbeck. Billings. IS. Y. Med. Record, Vol. 49, p. 577. Binz. Archiv f. exper. Path., Bd. 5, p. 109. Bordet. Annal. Institiit Pasteur, 1895, p. 398. Botkin. Virchow's Archiv, Bd. 137, p. 476 ; Bd. 141, p. 238 ; Bd. 145, p. 369. 2Deut. med. Woch., 1892, p. 321. Bouchut, JDubrisay. Gaz. med. de Paris, 1878, p. 168. Buchner. 'Berl." klin. Woch., 1890, No. 47. « Munch, med. Woch., 1894, p. 469. 3 Arch. f. Hvg., 1890, Bd. 17, p. 112. Canon. Deut. med. AVoch., 1892, p. 206. Cantucuzene. These de Paris, 1894. BIBLIOOBAPHY. 135 Capps. Boston Med. Surg. Jour., 1897, Nov. 4. Chaillou. These de Paris, 1865, p. 46. Cruveilhier. Compend. de Med., T. 2, p. 63. Denys, Havel. La Cellule, 1894. Duperie. Kevue de Med., 1878. EhrUch. lArchiv f. PlivsioL, 1879, pp. 166, 571. ^Qesammte Mittheil. Berlin, 1891. 3 Die Anaemie, Th. I., 1899. Einhorn. Diss. Berlin, 1884. Engel. Virchow's Archiv, Bd. 135. Ewing. N. Y. Med. Jour., 1895, Vol. 61, p. 257 ; Vol. 62, p. 161. File. Lo Sperimentale, 1896. Flemming. Archiv f. micr. Anat., 1891, Bd. 37. Fodor. Deut. med. Woch., 1886, No. 36. Fouassier. These de Paris, 1876. Futcher. Johns Hopkins Bull., 1897. Gabritscheu'shy. ' Morphology of the Blood, cited bv Botkin. ^ Annal. Institut Pasteur, 1890, p. 346. ^Ibid., 1894. Gartner, Eomer. Wien. med. Blatt., 1891, No. 42. Wien. klin. Woch., 1892, No. 2, Goldscheider, Jacob. Zeit. f. klin. Med., Bd. 25. Goldscheider, Muller. Fort. d. Med., 1895, p. 351. Gottlieb. Wien. med. Blatt., 1886, No. 17. Grohman. Diss. Dorpat., 1884. Grancher. Gaz. med. de Paris, 1876, p. 321. Grawitz. 'Deut. med. Woch., 1893, p. 468. 2 Virchow's Archiv, Bd. 110, p. 1. 3 Charity Annal., Bd. 16, p. 291. Gulland. Jour, of Physiol., 1896, p. 385. Gumpreeht. Deut. Archiv klin. Med., Bd. 57. Gundobin. Jahrb. f. Kinderheilk., 1893, p. 187. Haetaguroff. Inaug. Diss. (Russian), cited by Botkin. Hahn. Archiv f. Hvg., Bd. 28, p. 312. Halla. Prag. Zeit. f. Heilk., 1883, p. 198. Hankin. Cent. f. Bact., 1892, p. 782. Ibid., 1898, p. 692. Hartung, Wien. klin. Woch., 1895, p. 697. Hayem. Compt. Eend. Soc. Biol., 1887, p. 270, Heidenhain. Archiv f. micr. Anat., Bd. 43, p. 423. Hermann. Archiv f. micr. Anat., Bd. 37. Hirschfeld. Inaug. Diss. Berlin, 1897. Hirt. Muller's Archiv, 1856, p. 174. HoffmeiMer. Archiv f. exper. Path., Bd. 22, p. 306. Horbaczew.^ky. iMonatsch. f. Chemie, 1891, XII., p. 221. «Sitz. Wien. Acad. Wissen., 1891, Bd. C, Abt. III., p. 101. Hubbard, White. _ Jour, of Exper. Med., 1898, p. 639. Huhnerfauth. Virchow's Archiv, Bd. 76, p. 310. Jacob. Zeit. f. klin. Med., Bd. 30, p. 447. Kanthack, Hardy. Jour, of Physiol., Vol. 17, p. 91. Keng. Jour, of Physiol., Vol. 15, p. 361. Kikodse. (Eef.) Cent. f. Path., 1891, No. 3. Klein. Volkmann's Vortriige, 1893, No. 87. Knoll. U. d. Blutkorp. bei wirbells. Thieren, Sitzungsb. kais. Acad. Wien., 1893, Bd. 102. Kolisch. W^ien. klin. Woch., 1894, p. 797. Kosina, Eckert. Cited by Rieder. Kossel. Zeit. f. Hyg., Bd. 16. Kruger. Diss. Doi'pat, 1886. Kurloff. Abstracted by Ehrlich, Die Anaemie, I., p. 56. Leber. 1 Fort. d. Med., 1888, p. 460. 2 Die Entstehung d. Entzundung, p. 508. Licbman. Virchow's Archiv Suppl., Bd. 144, jj. 123. Limbeck. Zeit. f. Heilk., 1889, p. 392. Litten. Berl. klin. Woch., 1883, p. 405. Loicif. 'Sitz. kais. Acad. Wissen. Wien., 1885, Bd. 92, Abt. III. 2 Archiv f. micr. Anat., 1892, p. 524. ^Stud. u. Blut. u. Lymph., 1892, p. 72. Loewy, Richter. Deut. med. Woch., 1895, p. 240. Lubarsch. Fort. d. Med., 1888, No. 4. Lucke. Das Carcinom, Erlangen, 1867. Lyon. Virchow's Archiv, Bd. 84, p. 207. Malassez. 'Gaz. med. de Paris, 1876, p. 297. 2 Archiv de Physiol., 1874, p. 32. 136 THE LEUCOCYTES AND LEUCOCYTOSIS. Moleschott. Wien. med. Woch., 1854, p. 114. Metchnikoff. • Comparative Path, of Inflammation, Transl., 1893. ^ Annal. Institut Pasteur, 1895, p. 369. Meyer. Schmidt's Jahrb., Bd. 180, p. 121. Meyer, Seegen. Cited by Limbeck. Mocnatscheff. Archiv f. Gynaecol., Bd. 36, p. 277. MuUer. Zeit. f. Heilk., 1890, p. 213. ^Cent. f. Path., 1896, p. 529. Nasse. Untersuch. z. Physiol, u. Path., 1839, II. Nepveu. Memoires de Chirurg. Paris, 1880. Neusser. ^Wien. klin. Woch., 1894, p. 71. «Wien. med. Presse, 1892, Nos. 3-5. Nicholls. Phila. Med. Jour., 1898, Feb. 26. Nissen. Zeit. f. Hyg., 1889, p. 487. JVuttall. Zeit. f. Hyg., 1888, p. 353. Otto. Pfliiger's Archiv, Bd. 36, p. 12. Paget. Lect. on Surg. Path., 1870, p. 784. Patrigeon. Th^se de Paris, 1877. Pawlowsky. XI. Intemat. Cong., IV., p. 260. Pee, Rdnbach. Archiv de klin. Chir., Bd. 46, p. 486. Pekelharing. Semaine Med., 1889, p. 184. Pfeiffer. IJntersuch. a. d. botan. Institut Tubingen, I., II. Zeit. f. Hyg., Bd. 16, 268 ; Bd. 18, p. 1. Pohl. Archiv f. exper. Path., Bd. 22, p. 306. Rieder. Beitrage zur Kennt. d. Leucocytose, 1892. Bomer. Berl. klin. Woch., 1891, No. 36. Virchow's Archiv, Bd. 128, p. 98. Rogers, Josue. Presse Med., 1897, V., 113. Rossbach. Deut. med. Woch., 1890, p. 389. Sailer, Taylor. Intemat. med. Mag., 1897. Samuel. Eulenberg's Kealencyc, 2 Aufl., Bd. 3, p. 200. Sappey. Des. elem. fig. du Sang., Paris, 1881. Schattenfroh. Miinch. med. Woch., 1897, Nos. 1, 16. Scheurlen. Archiv f. klin. Chir., Bd. 36, p. 925. Schiff. Zeit. f. Heilk., Bd. 11, p. 30. Ibid., 1890, p. 1. Schmidt. Ziegler's Beitrage, 1892, p. 199. Schneider. Wien. med. Woch., 1890, No. 36. Schneyer. Zeit. f. klin. Med., Bd. 27, p. 475. Schultze. Archiv f. micr. Anat., Bd. 1, p. 1. Schulz. Deut. Archiv klin. Med., Bd. 51, p. 234. Sherrington. Proc. Royal Soc, 1894, Vol. 55, p. 161. Simon. Amer. Jour. Med. Sci., Vol. 117, p. 139. Sorensen, Hoffman. Schwalbe's Bericht, 1876, Abt. III. Stahl. Botan. Zeit., 1884, Nos. 10-12. Stengel. 20th Cent. Practice. Stokes, Wegefarth. Johns Hopkins Bull., 1897. V. d. Stricht. Archiv de Biol., 1892, T. XII. Thoma, Lyon. Virchow's Archiv, Bd. 87, p. 201. Timofjewsky. Cent. f. Path., Bd. 6, p. 108. Traube. Cited by Buchner, Arch. f. Hygiene, 1890. Tschistowitch. Berl. klin. Woch., 1891, p. 835. Tumas. Deut. Archiv klin. Med., Bd. 41, p. 323. Uskoff. Cited by Botkin. V. d. Velde. La Cellule, X., 1894. Vidal. Archiv gen. , 1890, p. 333. Vaughan. Med. News, 1893, Dec. 23. Weiss. Jahrb. f. Kinderheilk., 1893, Bd. 35, p. 146. Welcker. Cited by Rieder. Werigo. Annal. Institut Pasteur, 1892. Winternitz. Archiv f. exper. Path., Bd. 35, p. 77. Woino-Oransky. Cited by Gundobin. Zappert. Zeit. f. klin. Med., Bd. 23, p. 296. PATHOLOGICAL VARLiTIONS. 137 THE OCCURRENCE OF EOSINS, MAST-CELLS, MYELOCYTES, AND LYMPHOCYTES. EOSINOPHILIA. Physiological Variations. — The number of eosinophile cells in the blood of healthy adults varies according to Zappert, between 55-784 per cubic millimeter (.67-11 percent), but in the majority of indi- viduals the limits are between 1-4 percent. Zappert regards 50-100 per cmm. as a low normal count, 100-200 as intermediate, 200-300 as high normal, and over 300 as pathological. In children a relatively high proportion of eosinophile cells is the rule, and the variations in number are slightly greater. The average appears to be about 1-2 percent greater than in adults. There are no uniform changes referable to old age, sex, pregnancy, menstruation, or digestion. A slight increase after coitus has been ob- served. Pathological Variations. — From the fact that eosinophile cells may be multiplied many times without passing the limits established as physiological, it is obvious that none but considerable changes in their absolute numbers can have any distinct pathological significance. Yet the observations on eosinophilia have referred quite as much to per- sistently high or low averages within normal limits, as to distinct in- crease or decrease ; and a certain pathological significance has been shown to go with these lesser variations. Diseases of the Blood. — In leukemia the presence of large numbers of eosins was first regarded by Ehrlich as the pathognomonic sign of myelogenous leukemia, but it has since been shown that eosinophilia may be very marked in other conditions, and that the proportion of these cells in leukemia almost invariably lies within normal limits (1—7 percent). Their total numbers however are very greatly increased, Rieder's highest estimate reaching 29,000 per cmm. In lymphatic leukemia they are commonly absent. The presence of eosinophile myelocytes was regarded by Muller and Rieder as pathognomonic of leukemia, but these cells are not infre- quently found in other conditions. The writer finds that very large darkly-staining eosinophile granules in mononuclear cells are practi- cally limited to leukemic blood, but isolated examples have been seen in malaria. (Bignami.) In chlorosis, the eosins may be moderately increased, normal, or much diminished. Zappert was unable to verify Neusser's belief that absence of eosins in chlorosis is of unfavorable import. In secondary anemia the occurrence of eosinophile cells depends largely upon the underlying condition, the anemia itself not effecting any increase but rather a decrease. There are, however, many examples of secondary anemia with marked eosinophilia, e. g., anchylostomiasis. Diseases of Lungs. — The occurrence of large numbers of eosinophile cells in the blood and sputum of asthmatics was first noted by Gollasch 138 THE LEUCOCYTES AND LEUCOCYTOSIS. and has been uniformly confirmed by later observers (Leyden, Zap- pert, Mandybur, Weiss, Gabritschewski), who have found them to vary in the blood of such cases from 9-22 percent. In simple emphysema they are not markedly increased, and in pulmonary tuberculosis they are usually much diminished. Where tuberculosis complicates emphy- sema both blood and sputum may, however, show an excess of eosino- phile cells. (Weiss, Aronson and Philip.) In tuberculosis of lungs or other tissues, an absence of eosinophile cells from the blood is often observed, and this fact has been of con- siderable value in differential diagnosis between this and other con- ditions in which normal or increased numbers of these cells are present. Yet when tuberculosis is accompanied by cachexia and irregular suppuration, eosinophile cells may reappear in moderate numbers. Neusser believes that when eosins persist in tuberculosis, the outcome is usually favorable, since gouty subjects are comparatively resistant to this infection. Yet tuberculosis in emphysematous lungs is not in- frequently seen at autopsy. The eosinophilia following injections of tuberculin has already been considered. In febrile diseases of liver and gastro-intestinal tract the occurrence of eosinophile cells in the blood shows no uniformity. They have been found moderately increased in cirrhosis of liver. In chronic nephritis the eosins are apt to reach the higher physiolog- ical limits. Neusser believes that uremic seizures are regularly ac- companied by increase of eosins in the blood, and Zappert and Castellino report such cases. In nervous diseases, functional and organic, Zappert found, as a rule, a moderate increase of eosinophile cells, without being able to substantiate Neusser's classification of neuroses into those with and those without eosinophilia. Tlie latter he preferred to regard as ex- ceptions to the very general rule. In a distinct sympathetic neurosis, the eosins were not increased. In general insanity no uniform in- crease of eosins has been demonstrated. (Zappert, Krypiakiewicz.) Jelliffe found a moderate increase in one, and a total absence of eosins in five of twenty cases of general paresis. In Basedow's disease the eosins are usually increased. (Neusser.) Neusser finds an increase of eosins in a variety of nervous disor- ders, the psychoses of menstruation, puerperal mania, epilepsy, tetany, Basedow's disease, hemicrania. In acute mania not connected with pregnancy, and in melancholia, Somers found a well-marked and uni- form eosinophilia. Tumors. — From observations on 24 cases Zappert concluded that a very uncertain eosinophilia may occur with new growths, especially when cachexia is not advanced. He found 17.76 percent of eosins in one of several cases of lymphosarcoma, and a distinct increase in isolated cases of carcinoma. Neusser claims that eosinophilia arising in the course of a tumor growth indicates metastasis in the marrow, and Reinbach found 60,000 eosins in a case of lymphosarcoma with metastasis in the marrow. PATHOLOGICAL VARIATIONS. 139 Skin Diseases. — Various cutaneous lesions have furnished some of the most marked and interesting examples of eosinophilia. While the highest numbers of these cells have been found in pemphigus (4,800), it has been shown by Canon and verified by Zappert that the eosins are affected not so much by special forms of cutaneous lesions as by the extent, intensity, and lack of healing tendency, on the part of the lesion. Moreover it appears that local afflux of eosinophile cells oc- curs in the early and in the less active stages of many cutaneous lesions, and is often insufficient to cause a general increase in the blood. That many lesions are caused by local toxic agents having special chemotactic influence over eosinophile cells is indicated by the fact that in pemphigus artificial blisters do not necessarily contain the larger numbers of eosins that are found in the spontaneous lesions. (Kreibich.) Among the diseases showing high eosinophilia may be mentioned : pemphigus, eczema, scleroderma, psoriasis, pellagra, lupus, if widespread, urticaria (60 percent Lazarus). (Zappert, Neusser, Canon, Tscheleneff", Laredde.) In the blood of leprosy, Gaucher and Bensaude found 8.48—28 per- cent, Darier as high as 61 percent, and Bettman 7-18.4 percent. While the vesicles of herpes zoster may contain a high proportion of eosins, the blood usually shows no increase of these cells. (Bettman.) Post-febrile Eosinophilia. — In nearly all forms of acute polynuclear leucocytosis an absence of eosinophile cells has been noted. Scarlet fever, acute rheumatism, and malaria, sometimes furnish notable ex- ceptions to this rule. Early in the decline of the febrile process, and before the neutrophile leucocytes have greatly decreased, eosinophile cells begin to reappear, and may shortly be found in more than normal numbers. The reappearance of eosinophile cells in pneumonia and septic processes, etc., has often been found to herald a favorable turn in the disease, but usually other favorable clinical symptoms are present at the same time. In malaria, eosinophile cells usually persist in small numbers during the paroxysm, while in the afebrile interval, and in chronic cases, a slight eosinophilia exists and may be a useful diagnostic feature of the blood. In acute rheumatism, the eosins are usually present in moderate num- bers during the fever, increase slightly during convalescence, and in one of Zappert's cases a relapse was attended with a further increase. Acute Exanthemata. — In scarlatina there is a remarkable exception to the usual rule that eosins disappear during febrile leucocytoses, for in this disease the eosinophile cells usually persist and may be markedly increased. This peculiarity was first noted by Kotschetkoif, who found that the oxyphile cells steadily increase during the fever, reach a maximum in the second or third weeks (8-15 percent) and fall to normal in six weeks. In his very severe cases however they fell more rapidly. Rille found marked eosinophilia in severe cases and Zappert, Silvester, and Felsenthal, have added confirmatory reports. 140 THE LEUCOCYTES AND LEUCOCYTOSIS. In measles the eosins are usually normal or diminished. (Zappert, Felsenthal, Cabot.) In some other acute simple erythemata Zappert found distinct eosin- ophilia. Gonorrhea. — According to Bettman eosins appear in gonorrheal pus during the first few days of the disease when the exudate is largely serous. Here as elsewhere it seems that when the stage of exudation of neutrophile cells is reached the eosins largely or entirely disappear, to return again during the decline of the discharge. This latest con- tributor also agrees that the involvement of the posterior urethra and prostate is usually attended with a marked increase of oxyphile cells in the discharge, and he finds further a similar influence from the involve- ment of the epididymis. Most observers are agreed that the eosinophile cells of the blood are increased when they are abundant in the discharge, but not in any pro- portionate degree. (Posner, Finger, Pezzoli, Janowski, Epstein, Tsche- leneff, Vorbach.) Bettman found 25 percent in a case complicated by epididymitis. Sypliilis. — A very uniform increase of eosins accompanying cuta- neous syphilis was noted by Rille and by Loos, but the observations of Muller and Rieder and of Zappert, though less numerous, did not bear out these claims. In a case of congenital syphilis Muller and Rieder found a well-marked increase (12 percent). Of osteomalacia Neusser finds two types, with, and without, eosino- philia. The latter are usually advanced cases with inflammatory changes in the bones and marrow and myelocytes in the blood. The former include those cases in which castration has been followed by re- covery. Gout. — Neusser first called attention to the high proportion of eosinophile cells in the blood of gouty subjects. Although his opinion was not based on extensive observations or reports, it will generally be found that in the gouty diathesis the eosins reach Zappert's intermediate or high normal limits. Yet in testing this rule the writer soon found that this feature of the blood is of little value in the diagnosis of acute gout, meeting with cases in which the blood contained few oxyphile cells. In the irregular manifestations of the diathesis, eosinophilia ap- pears to be more uniform but still subject to many variations. Intestinal Parasites. — Infection with various common intestinal para- sites is accompanied by extreme eosinophilia, as first noted by Buckler in 1894. In cases infected by Oxyuris, Buckler found 1 9 percent of eosins ; by Asearis, 16 percent; Anchylostoma, 72 percent; Tenia mediocanel- lata, 34 percent. In most of these cases eosinophile cells and Charcot- Leyden crystals are abundant in the feces. Anchylostomiasis, however, is not always accompanied by eosino- philia or crystals in feces. (Ehrlich.) In the anemia from Bothrioceph- alus latus Schumann found few eosins in the blood. Trichinosis has lately been added to the list of diseases marked by extreme eosinophilia. ORIGIN AND SIGNIFICANCE OF EOSINOPHILIA. 141 Origin and Significance of Eosinophilia. — The very diverse opinions that have prevailed regarding the origin of eosinophiha can only be finally reconciled by holding strictly to the view that eosino- phile cells are derived solely from the eosinophile cells of the bone marrow. That a local multiplication of these cells may occur under many circumstances is however strongly suggested by their abundance in certain foci, especially in the skin and mucous membranes. Thus, in the absence of general eosinophilia, rich deposits of eosinophile cells are often found in normal tissues, in cutaneous vesicles and pustules, in chronically inflamed mucous membranes, in nasal polyps and other tumors, and in leprous, tuberculous, and syphilitic lesions. Lately they have been found in very large numbers in bloody exudates in the pleura. (Weiss.') Their abundance in the mucosa and exudate of the inflamed bronchi in asthma, led Neusser, and on parallel grounds many others, to actively maintain their local origin apart from the bone marrow. In one case of pemphigus with marked local and general eosinophilia, Neusser * is reported to have found no increase of eosins in the bone marrow, but studies in this important field are yet entirely inadequate to figure in the discussion. In the marrow of cases of malaria showing eosinophilia the writer found an increased number of eosinophile cells, while in other cases without eosinophilia, these cells in the marrow also were deficient. Dominici found an excess of eosins in the marrow of a rabbit showing marked eosinophilia. The exponents of the local origin of eosinophile cells have been un- able to bring forward evidence to prove that these cells undergo mitotic division in the skin or that neutrophile granules are there transformed into eosinophile. The writer believes that all the phenomena connected with general and local eosinophilia can best be explained by the same chemotactic principles that are known to control neutrophile cells. From the ob- servations on eosins in gonorrheal pus, in cutaneous and serous exu- dates, and in the blood, it appears that inflammatory products attract eosins at one stage and neutrophile cells at another, more acute, stage. As purulent exudates of small extent may cause no general leucocy- tosis (e. g., furunculosis), so, many cutaneous lesions with exudates composed largely of eosins cause no general increase in the circulation, but when either reaches a certain grade the neutrophile or eosinophile cells are drawn from the marrow in sufficient numbers to cause a noticeable increase in the blood. Ehrlich finds reason to believe that substances which attract eosino- phile cells may be derived from the destruction of epithelial cells, as seen in the local eosinophilia about the ulcers of lupus after injection of tuberculin; from the mucin in nasal polyps; and from toxines of para- sites. Michaelis has shown that when lactation is interrupted in the guinea-pig many eosinophile cells collect in the breasts. But in the human breast after stagnation of milk Unger found many mast-cells. Weiss has recently maintained that local eosinophilia tends to occur where there is extensive extravasation of blood, the derivatives of Hb 142 THE LEUCOCYTES AND LEUCOCYTOSIS. ; being absorbed by the polynuclear cells and deposited in the form of eosinophil granules. He bases this opinion on the observation of a case of septicemia with hemorrhagic pleural effusion. In the blood there were 40 percent of eosinophile cells, while in the sediment of the exudate 76 percent of these cells were found. He refers to two other somewhat similar cases reported by Harmser. The most comprehensive view of the significance of eosinophilia is that of Neusser ^ and his pupils, who, from extensive observations have found evidence that the supphj of eosinophile cells in the blood is con- trolled by the sympathetic nervous system, and that eosinophilia is the ex- pression of sym,pathetic nervous irritation. This irritation they believe may proceed from the generative organs, ovaries, uterus, or prostate, disorders of which, with their related neuroses, are usually accom- panied by eosinophilia ; from the skin, diseases of which have furnished some of the best examples of eosinophilia ; from the intestines, from which the toxemia of parasites and that of gout, which Neusser regards as of intestinal autotoxic origin, give rise to marked eosinophilia, etc. Following out this idea Neusser has elaborated an extensive scheme of the altruistic relations of the viscera in disease, by which he is able to correlate nearly all the diseases showing eosinophilia. He finds a necessary sequence of events in a case which suffered from migraine in youth, pemphigus and asthma in old age, and died from cancer of the prostate with general metastasis in the bone marrow. Although the recent tendency has been to deny any great impor- tance in Neusser's conception of eosinophilia, it must be remembered that it is based on a well-known and far-reaching principle in path- ology. While in practice a great many exceptions have been found to the rules which he formulated, it will yet be wise to keep them in mind in drawing conclusions from eosinophilia. Value of EosinopMlia in Diagnosis. — In general, the presence of a moderate or increased number of eosins in the blood is often of value in the diagnosis between a condition in which these cells are known to persist as against those in which they are commonly absent. The writer has found situations of the sort most frequently in the diagnosis between gouty and tuberculous affections. Other situations have arisen in the diagnosis between active malarial and typhoid fevers, scarlet fever and measles, trichinosis and rheuma- tism. In prognosis the reappearance of eosins in suppurative processes is undoubtedly a fav^orable sign indicating the approach of defervescence. Many clever deductions have been drawn by Neusser and his pupils from the behavior of eosinophile cells in the blood, but others have not been so successful in this field. OCCURRENCE OF MAST-CELLS. The knowledge of mast-cells dates chiefly from the studies of West- phal, 1880, and of Unna, who found abundant collections of large OCCURRENCE OF MYELOCYTES. 143 mononuclear cells with large strongly basophile granules, in tumors, and in tuberculous, syphilitic, and other lesions of the skin. The occurrence of such cells in tissues has since been described by many writers and it is now accepted that their appearance is favored by a considerable va- riety of chronic disturbances of nutrition in tissues and that they are commonly associated with eosinophile cells. The majority of coarsely granular basophilic cells in tissues, are not, however, identical with the mast-cells of the blood. In the blood, Canon failed to find them in 9 of 22 healthy subjects, counting 500-1,000 leucocytes, while in the other 13 cases they yielded an average of .47 percent. In a variety of skin diseases (20 cases) they were slightly more numerous (.58 percent) and in general were slightly increased when eosinophile cells were abundant. Sher- rington found them distinctlv increased in the blood of moribund cholera patients. The writer constantly fails to find mast-cells in the better class of healthy subjects, but in hospital and dispensary cases with lesser ailments they appear to be more numerous. In cases of severe malaria he has seen them more abundant in the blood than in any condition except leukemia, while in the viscera of such cases large granular basophilic cells were unusually frequent. Neusser examined a gonorrheal exudate composed exclusively of mast-cells. {^Letter to Ehrlk-h, Die Anaemie, I., p. 111.) In myelogenous leukemia there is, in the majority of cases, mast-cell leucocytosis, these cells being here more abundant than in any other condition. They often outnumber the eosinophile cells (Ehrlich), and must be considered the sole isolated pathognomonic sign of this disease. The writer cannot support Ehrlich's statement (D/e Anaemie, p. 123), that they are invariably present in considerable numbers in myeloge- nous leukemia, having missed them entirely in some cases of acute leukemia and searched in vain over thousands of cells in chronic cases. Ehrlich believes that mast-cells are derived exclusively from the bone marrow, and respond to chemotactic influence from principles however which are very rarely present in the human organism. OCCURRENCE OF MYELOCYTES (EHRLICH'S). Although the eosinophile myelocytes and to a less extent CorniPs very large pale neutrophile myelocytes are found almost exclusively in myelogenous leukemia, the smaller variety of mononuclear neutro- phile cells have been shown to occur in the blood in a considerable variety of conditions. From the many recent reports of the occurrence of a few myelocytes in the blood it appears that such cells may be swept from the marrow by several distinct causes. 1. Accompanying the polynuclear leucocytosis of infec- tious DISEASES, a considerable number of myelocytes may be found. This fact first appeared in the studies of Turk, was emphasized through the discovery by Engel of unusually large numbers (12 percent) in 144 THE LEUCOCYTES AND LEUCOCYTOSIS. ; unfavorable cases of diphtheria, and has since been extended by others to most severe, acute, and subacute infections. Usually the percentage of these cells in polynuclear leucocytosis is much lower than in diphtheria, and not being associated with eosino- phile cells or mast-cells it can seldom raise a suspicion of leukemia. Nevertheless a difficulty arose with a case of acute leukemia, observed by Thomson and the writer, in which the first examination of the blood in a case resembling typhoid fever showed leucocytosis and 5 percent of myelocytes. A few days later there were 12 percent of myelocytes and at the autopsy the lesions of leukemia were demon- strated. Eosins and mast-cells were absent. The presence of myelocytes in polynuclear leucocytosis is readily ex- plained by the hyperplasia and hyperemia of the marrow. 2. In primary and secondary anemias of severe, or even of moderate grade, a few myelocytes have frequently been observed. (Klein, Krebs, Loos, Hammerschlag, Capps.) They are perhaps most frequently seen in pernicious anemia (Neusser), in the anemia of syphilis (Rille), with malignant tumors (Cabot), and in v. Jaksch's anemia. It has been supposed that tumors involving the marrow would yield myelocytes as well as eosins in the blood, but this suppo- sition has not been verified in fact, although they have been noted in considerable numbers in rachitis, osteomyelitis, and osteomalacia. In all the above conditions the appearance of myelocytes must be referred to hyperplasia of red marrow, or mechanical dislodgment of marrow cells resulting from structural changes in the marrow or hydremic states of the blood plasma. 3. After severe mechanical disturbances of the circu- lation a few myelocytes may be found in the blood. This rule is illustrated by their discovery in the blood of uremia, asphyxia, acute mania, etc. (Neusser.) The writer has seen a few myelocytes appear during ante-mortem leucocytosis. The occurrence of myelocytes in leukemia will be considered with that disease. LYMPHOCYTOSIS. A relative or absolute increase of lymphocytes in the blood is of frequent occurrence and has at times important significance. In speaking of lymphocytosis, as of eosinophilia, since each of the cells concerned represents an independent series, conceptions will be more accurate if the actual numbers found, as well as the relative propor- tions, are reported. In estimating lymphocytosis it is important to distinguish also between the large lymphocytes and large mononuclear leucocytes with faintly basophile protoplasm. Physiological variations in the proportions of lymphocytes are observed at different periods of life. The first embryonal leucocytes are all mononuclear basophile cells (Saxer), but as other varieties of leucocytes make their appearance, the proportion of lymphocytes diminishes till, at birth, the healthy infant shows 50-66 percent of LYMPHOCYTOSIS. 145 these cells. Throuj^hout healtliy iafancy, this proportion steadily diminishes till, at the tenth to fourteenth years, the usual percentage is 27-30 percent. Pathological Lymphocytosis, — Anything which interferes with the natural development of the infant retards tiie progress of this change in the proportion of lymphocytes, which is found to be rela- tively high in anemic or poorly developed children. Sometimes the development of lymphatic tissue reaches an excessive grade, and a condition known as ^' eonsfitutio li/mpJiatica'^ is established, marked by lymphocytosis, with simple hyperplasia of many lymph nodes and of the red marrow, by rachitis, and by many other developmental anom- alies. Since Ohlmacher has demonstrated that most cases of idio- pathic epilepsy are associated with the coiistitutio li/mphatica, it is reasonable to suppose that lymphocytosis is a frequent feature of the blood in such cases. The lymphocytosis of Basedoiv's disease should probably be placed in this connection. Rachitis is almost always attended with a well-marked increase in the number and proportion of lymphocytes in the blood. (Rieder, Monti, Berggrun.) This fact is plainly referable to the lymphoid hyperplasia and the hyperemia of the bone marrow in this disease. In some tumors Reinbach has reported an extreme diminution of lymphocytes, only .6 percent being present in one case of Ii/mpho-sar- coma colli. This result is referred by Ehrlich to closure of the lymph paths by sarcomatous growth. In many lymphomata, however, there is marked uniform lymphocytosis, and it would appear that these neo- plasms when leaving the lymph paths free induce lymphocytosis. As the more malignant tumors obliterate lymph paths, some inference re- garding prognosis may be derived from the presence or absence of lymphocytosis in these cases. Various other types of sarcoma are fre- quently associated with lymphocytosis and there are on record several cases of sarcoma associated with lymphatic leukemia. (See Leukemia.) In most splenic tumors there is relative or absolute lymphocytosis. (Muller and Rieder, Weiss.* ) On the other hand, after splenectomy in animals, prolonged lymphocytosis is usnally observed. (Kurloff.) In man, both mononuclear and polynuclear cells are usually much in- creased after this operation. The lymphocytosis of infectious di.seases is probably always associated with acute hyperplasia of the lymphatic structures. In typhoid fever the hyperplasia is most marked in the abdominal nodes and spleen, but may become more general. There is in this disease a relative lymphocytosis, which after the first week usually in- creases till there is a high percentage of lymphocytes. The writer has found a uniform relation between the lymphocytes in the blood and the grade of lymphatic hyperplasia found at autopsy. In one case the examination of the blood led to a strong suspicion of lymphatic leuke- mia, and at autopsy the mesenteric glands were of unusually large size, and the edges of the partly necrotic intestinal ulcers rose 1.5 cm. above the mucosa. 10 146 THE LEUCOCYTES AND LEUCOCYTOSIS. In the infectious diseases of children a well-marked lymphocytosis, together with increase of neutrophile cells, is so common that it should occasion no surprise, and points merely to a special involvement of the lymphoid tissues. In diphtheria a close relation between lymphocytosis and lymphatic hyperplasia was noted in some of the writer's cases. In others, how- ever, the lymphocytes, though increased, were not in relatively high proportion, while the cervical lymph nodes were much enlarged. In broncho-pneumonia Cabot reports a case with 94,600 leucocytes, 69 percent of which were lymphocytes. In whooping-cough during the convulsive period Meunier found the neutrophile cells doubled but the lymphocytes quadrupled in number. During the course of measles, at the close of scarlet fever or small- pox, and during prolonged lysis in pneumonia (Klein), a relatively high proportion or increased number of lymphocytes is usually ob- served. In the less acute diarrheas of infancy, lymphocytosis is the rule, and is referable to the hyperplasia and irritation of the intestinal lym- phoid structures. (Weiss.") Digestion Icucocj/tosis is also for the same reason largely a lymphocytosis. (Kieder.) In the initial or persistent hypoleucocytosis sometimes observed in infectious diseases (pneumonia, diphtheria), the remaining cells arc largely lymphocytes. Of chronic diseases affecting the lym})hatic tissues both tuberculosis and syphilis are frequent causes of lymphocytosis. In various forms of pulmonary and visceral tubereidosis the leuco- cytes are usually normal or diminished in number and the majority of those remaining are lymphocytes, a relation which is of course altered when from any cause polynuclear leucocytosis is established. The condition of the blood is here in accord witli tlie histological character of the visceral lesions in which the infiltration with " round cells " is a prominent feature. Lymphatic anemia is a term applied by Xeusser to a rather charac- teristic group of cases in which there is chlorotic anemia, lymphocy- tosis, and, usually, evidence of tuberculosis. Kieder could not agree with Neusser that most cases of chlorotic anemia with increase of lymphocytes are associated with tuberculosis and are of less favorable prognosis. The writer finds that the chlorotic anemia of tuberculosis as a rule shows absence of eosins as well as increase of lymphocytes. Injections of tuberculin are usually followed by fever and lymphocy- tosis, which may be referred to irritation of inflamed tuberculous lymph nodes. (Ehrlich.) In congenital and in secondfiry acquired syphilis, it has been shown by Bieganski, Rille, Anc, and others, that there is uniform and con- siderable lymphocytosis, which must here again be referred to the in- volvement of lymph nodes. The presence of an increased proportion of lymphocytes in some cases of scurvy and hemophilia, has been observed by Neusser, and the same feature is recognized by Ehrlich in many seve7'e anemias. In * LARGE MONONUCLEAR LEUCOCYTES IN LYMPHOCYTOSIS. 147 anemia infantum pseudo-leukemica the very numerous leucocytes are principally large mononuclear cells. In pernicious anemia, a consider- able leucocytosis may consist largely of lymphocytes. ExPERiMEXTAL LYMPHOCYTOSIS has been produced by injection of tuberculin and of extract of carcinomatous tumors by Grawitz who refers this eifect to the lymphogogic action of these agents. Wald- stein reports lymphocytosis in rabbits from pilocarpine, which the writer finds to induce in rabbits only a relative increase of lympho- cytes by diminution of polynuclear cells. In the origin of lymphocytosis Ehrlich finds that mechanical rather than chemotactic influences are chiefly or wholly concerned, since the lymphocytes are not sensible to chemotactic attraction. The histology of tuberculous and syphilitic lesions, especially of the " lymphoid tubercle " and chancre, indicate that there are some conditions which gather large numbers of these cells from the blood and lymph nodes. Janowsky's experiments in which large collections of lymphoid cells were shown to gather at the site of injection of turpentine, etc., point to the same conclusion. In all of these situa- tions, however, the lymjjhocytes, like eosinophile cells, seem to follow entirely different rules from those governing neutrophile cells, a fact which emphasizes the importance of Kanthack's classification of lym- phocytes and eosins as celomic, the neutrophile, as hemic cells. In general, lymphocytosis seems to result from mechanical discharge of cells from lymphoid structures, aided by diminished blood pressure and increased flow of lymph. Relation of Lymphocytosis and Eosinophilia. — It is of interest to note that a peculiar relation seems to exist between the appearance in blood and tissues of lymphocytes and eosinophile cells. Neither are apt to appear when neutrophile cells are prevailing in blood or exu- date, but either may be found in the less acute exudative processes. The sequence of lymphocytosis followed by eosinophilia is especially clear after the injection of tuberculin (Grawitz) and after splenectomy in animals. (Kurloff.) Many other less patent examples of this rela- tion may be noted by comparing the conditions showing lymphocytosis or eosinophilia, as above detailed. Large Mononuclear Leucocytes in Lymphocytosis. — In the ma- jority of cases reported as showing lymphocytosis no distinction has been made between lymphocytes and large mononuclear leucocytes. Yet there is considerable evidence to show, as Ehrlich believes, that these cells represent separate series, and in some conditions it is the large mononuclear leucocytes and not the lymphocytes which are in- creased in number. This fact has been noted in v. JaJcsch's anemia, especially by Hock and Schlesinger ; in rachitis by Kieder ; in syphilis by Rille ; and after smallpox and scarlet fever, by Felsenthal.^ It is sometimes distinctly seen after splenectomy. In the cases of acute lymphemia described by Frankel and others, the majority of cells are of large size. The significance of an increase of large mononuclear cells apart from lymphocytes cannot at present be stated. 148 THE LEUCOCYTES AND LEUCOYCTOSIS. Bibliography. EosiNOPHiLiA, Lymphocytosis, etc. Anc. Diss. Wratsch., 1895, No. 5. Aronson, Phillip. Deut. med. Woch., 1892, No. 3. Bieganski. Arcliiv f. Derm. u. Syph., 1892, I. Bettmann. Archiv f. Derm. u. Svph., Bd. 39, p. 227. Miincli. med. AVoch., 1898, p. 1229. Buckler. Miinch. med. Woch., 1894, Nos. 2-3. Cabot. Boston Med. Surg. Jour., Vol. 134, p. 4. Canon. Deut. med. Woch., 1892, p. 206. Cajjps. Amer. Jour. Med. Sci., July, 1896. Castellino. Gaz. d. ospedali, 1891, p. 476. Darier. Annal. de Dermat., 1896, p. 842. Dominici. Compt. Rend. Soc. Biol., 1900, p. 73. Engel. Deut. med. Woch., 1899, pp. 118, 137. Epstein. 65th Versl. d. Naturfors., Nurnberg, 1893. Ewing. N. Y. Med. Jour., Vol. ('>6, p. 37. Fehenthal. i Archiv f. Kinderlieilk., 1892, p. 86. ^ibid., 1893, p. 78. Finger. Blenorrhoea d. Sexual organe. Frankel. Deut. med. Woch., 1895, p. 623. Oabritschewitky. Archiv f. exper. Patli., Bd. 28. GoUasch. Fort. d. Med., 1889, p. 3(;i. Gaucher, Bensaude. Annal. de Dermat., 189(), p. 204. Hammarschlag. Berl. klin. Wocli., 1894, p. 793. Hock, Schlesinger. Beitriige z. Kinderlieilk., 1892. Janowsky. Cent. f. Path., 1892. Ziegler's Beitriige, Bd. 15, p. 128. Jelliffe. N. Y. State Hospitals Bull., II., p. 397. Klein. Volkmann's Vortriige, 1893, No. 87. Kotschetkoff. Cent. f. Path., 1892. Krebs. Inaug. Diss. Berlin, 1S92. Kreibieh. Archiv f. Derm. u. Syph., Bd. 50, p. 225. Krypiakiewicz. Wien. med. AVoch., 1892, No. 25. Kurloff. Abstract bv P^hrlich, Die Anaemic, I. Laredde. Annal. de Dermat., 18i)8, p. 1016. Leyden. Deut. med. Woch., 1891, No. 38. Loos. Wien. klin. Wocli., 1892, p. 261. Mandybur. Wien. med. Woch., 1892, Nos. 7-9. Meunier. Compt. Rend. Soc. Biol., 1898, p. 103. Michaelis. Cited bv Ehrlich, Die Anaemic, I., p. 112. Muller, Rieder. Deut. Archiv klin. Med., Bd. 48, j). 105. Neusser. 'Wien. klin. Woch., 1892, No. 4. '^ Cited bv Kreibieh. Ohlmacher. Bull. Ohio State Hospital, 1898. PezzolL Archiv f. Derm. u. Syph., Bd. 35. Posner. Diagno.stik d. Hautkrankii. Reinbach. La ngen beck's Archiv, Bd. 46. Rille. Archiv f. Derm. n. Svph., Bd. 24, p. 1028. Saxer. Anatomische Hefte, "l896, Bd. \'I. Schaumann. Zur Kenntniss d. Bothriocephalus Anaemie, Berlin, 1894. Sherrington. Proc. Royal Soc, 1894, Vol. 55, p. 161. Silvester. St. Bartholomew's Hospital Rep., Vol. 32. Somers. N. Y. State Hospital IJull., Vol. I., p. 75. Thomson, Ewing. N. Y. Med. Record, Vol. 53, p. 333. Tscheleneff. Annal. de Dermat., 1898, p. 1152. Turk. Blutuntersuch. bei Infectionskrankh., 1898. linger. Cited by Elirlich, Die Anaemie, I., p. 112. Unna. Monatsh. f. prak. Dermat., Bd. 18. Vorbach. Diss. Wurzburg, 1895. Weiss. 'Cent. f. inn. Med., 1899, p. 97. 2 Wien. med. Presse, 1891, p. 1538. Westphal. P^hrlich's Gesam. Mitt. Zappert. Zeit. f. klin. Med., Bd. 23, p. 227. Fig. 19. CHAPTER V. DEVELOPMENT OF BLOOD CELLS. AVhile the mode of origin of the first blood cells of the embryo has been fully determined, the alterations in the process which supervene durins: late fetal and adult life still remain an obscure and difficult problem. It becomes necessary, therefore, to consider separately the formation of blood cells at different periods of intra- and extra-uterine life. ERYTHROCYTES. Formation of Red Cells in the Embryo. — In early embryonal life the formation of red cells in isolated groups of mesodermal cells in the "vascular area " of the chick was first noted by Pander, who named these foci " blood islands." The process was later studied more closely by Hiss, Remak, Kolliker,^ "VVissosky, Klein, Strieker, and others. From their researches it appears that the first blood cells of vertebrates are formed by the ap- pearance of Hb in some of the cells of the mesodermal cords which go to form the first capillaries. Upon the formation of the vessel these cells lie free in the lumen as nu- cleated red blood cells. Both the vessel wall and the primitive ery- throcytes are thus derived from the same group of cells. Theory of Endoglobular Formation of Red Cells. — In many mammals all the early red cells are nu- cleated, but in some (rat, guinea- pig), many non-nucleated red cells are abundant at an early stage, al- though in the human embryo they appear certainly only after the fourth week and form only 25 percent of the red cells at the fourth month. The formation of the first non-nucleated red cells has been studied by Schafer, Ranvier, Hayem, and others, who describe in the rat and guinea-pig, the appearance in certain large connective tissue cells (angioblasts, hematoblasts, cellules vaso-formatives) of discs and globules Theory of eudoglobiilai- foriuatiou of red cells (SCHAFEK.) 150 DEVELOPMENT OF BLOOD CELLS. of hemoglobin, which, by subdivision of the body of the cell, become non-nucleated red blood cells. These red cells soon come to lie free in a central cavity within the cell, which is itself transformed into a young capillary. According to this theory, the red discs are not cells but fragments of cells, of endoglobular origin, ichile nucleated red cells have a different signifiGance, being formed by mitotic division of other nucleated red cells. With the development of the lymphatic system and liver, the intracellular formation of red cells is said to cease, and few or no traces of it have been described in most mammals at birth. Kuborn and Malassez, however, describe the formation of red cells in the late embryonal liver and marrow, through gemmation of the giant cells of these organs. The theory of intracellular formation of blood discs has not been supported by more recent investigations. Spuler and Saxer find no traces of such a process, regarding the vaso-formative cells of Ranvier as endothelial cells which have become separated from the delicate young vessels and have carried with them some of the red cells of the vessel. In the adult marrow such cells have not been fully identified. Recently, however, Francois has attempted to substantiate the theory of endogenous red cell formation in the rabbit. In later embryonal periods the process of development of red cells is very similar to that in extra-uterine life. Mode of Origin of Red Cells in Late Embryonal and Extra-uterine Life. — That the nucleated red cells of the nutrrow multiply by indirect di- vision of other nucleated red cells was first stated by liizzozero,' and has been accepted by most histologists. Bizzozero, however, claimed that the large nucleated red cell of the adult marrow is the most primitive cell in the body capable of producing erythrocytes, believing that the intermediate stages between the embryonal wandering cell and the colored erythroblast have been lost in later embryonal life. But it has since been shown to the satisfaction of most authorities, that the series of mitoses leading to the production of- red discs begins in color- less cells antecedent to the large nucleated red cells, and which are only very slightly different from the original mesoblastic cell of the embryo. It thus appears that the formation of nucleated red cells in the adult is practically the same as in the embryo and that at all periods of life the red cell is the product of several series of mitoses of a colorless meso- blastic cell. The difficulty of tracing this series from the large nu- cleated red cell to the colorless mesoblastic " mother cell " in the mar- row, has given rise to the diverse opinions now held regarding the ultimate development of red corpuscles. As to the exact morphology of the original erythroblasts of adult mar- row, opinions are more or less at variance, l)ut it is agreed by Neumann,^ RindHeisch, Obrastzow,' Malassez, Howell, Lowit,^ Muller, and many others, that it is a large mononuclear cell, larger than the ordinary nucleated red cell, with pale nucleus and without hemoglobin, and it is usually described as lacking a nucleolus. From these " embryonal " cells are derived, by mitotic division, two or more series of cells which ERYTHROCYTES. 151 gradually approach the type of the nucleated red cell which is rich in hemoglobin, and whose nucleus is compact and pyknomorphous. Among dissenting opinions may be mentioned the conclusions of Foa, according to whom the red cells are derived from the giant cells of the marrow, spleen, and liver, and the theory of Hayem,2 partially accepted by Afanissiew and Pouchet, that the red discs are derived from the blood plates. Regarding important details in the origin and relation of the formative erytkroblasts, opinions are also widely at variance. Lowit describes in the marrow, spleen and lymph nodes, two separate series of colorless cells which produce, one, leucocytes, the other, red cells. The cells of these two series differ in the structure of the nucleus, in the fact that one type (leucoblasts) exhibits ameboid and phagocytic activity, and in their method of division, the erythroblasts multiplying by indirect division, the leucoblasts by a modified form of karyokinesis described as divisio indirecfa per granula. According to Lowit the red cells are derived from the erythroblasts after these have reached the circulation from the lymphoid tissues. Bizzozero 2 describes the primitive erythroblast as separate from the leucoblast, both in morphology and especially in its relation to the blood vessels of the marrow. While numerous nucleated red cells in mitotic division are found, accoi'ding to Bizzozero, in the masses of cells lying within the capillaries of the marrow of birds, intervascular tissue is entirely free from such cells. The capillary network of the marrow is, therefore, an en- dovascular blood-gland. This observation has been verified by Torre, Salvioli, and Denys, for nucleated hemoglobin-holding cells, but other authors trace the erythroblasts back to colorless cells, which are not in- variably intravascular. In man, however, while the erythroblasts lie closer to the lumen of the capillary than do the leucoblasts, the walls of the sinuses are incomplete and the islands of nucleated red cells are found in the midst of masses of leucoblasts. (Muir.) H. F. Muller describes a common form of mother cell, for both red and white blood cells. By their indirect division are produced (1) cells which resemble the original, (2) mononuclear leucocytes, and (3) nucleated cells which develop hemoglobin and by the gradual disappearance of their nuclei become the non-nucleated red blood discs. Denys believes, with Lowit, that there are two separate developmental series of leucoblasts and erythroblasts, each of which multiplies by mitosis, and is originally coloi-less. Both are ameboid, but the erythroblasts are very slightly so, Denys, like Bizzozero, finds that the leucoblasts lie in the cords outside the vessels of the marrow, while the erythroblasts are found within the vessels. Hayem believes the red blood discs to be derived from the blood plates of Bizzozero, which he therefore calls ^' hematoblasts.''^ He claims to have seen all transition forms between blood plates and red blood cells and finds that when blood formation is unusually active the blood plates and the transition forms are more abundant. In regard to many details of the origin of blood plates and their transformation into red cells, Hayem was unable to reach satisfactory conclusions. To summarize the work in this field, it may be said that we do not know certainly whether any common cell of origin of red and white blood corpuscles exists in the late embryo or adult, or whether these corpuscles are derived from completely separate series. The late con- 152 DEVELOPMENT OF BLOOD CELLS. tributions favor the existence of a common mother cell for both groups, persisting at least into late embryonal life. The Transformation of Nucleated Red Cells into Blood Discs. — To that minority of observers who believe that the red blood discs are derived from other sources than the nucleated red cells the task of explaining the disappearance of all trace of the nucleus of the parent cell presents no difficulties. Among the majority of writers who accept the origin from nucleated cells, the theories regarding the disappearance of the nucleus are various but none has been fully demonstrated. Kolliker/ Bizzozero,'^ Neumann,' Lowit, Foa, and many recent ob- servers believe that the nucleus gradually disappears in the cell. The difficulty in accepting this opinion lies in the fact that Avhile red cells with rather small nuclei are abundant in red marrow, the. Jinal stac/ex of the disappearance of the nucleus are traced only with difficulty. Yet Schmidt, Spuler, Israel and Pappenheim, and Masslow, claim to have found these final stages in abundance. Eiirlich believes that the nucleus of the normoblast is extruded, while that of the megaloblast fades within the cell. Rindfleish first, and later Howell, described fully the shrinkajje and final extrusion of the compact nuclei of red cells in man, mammals, and amphibia. Van der Stricht, Kostianecki,i Saxer, Albrecht, Disse, Muir, and others find that after considerable shrinkage the nucleus may become compact and homogeneous or may be fragmented, but it is eventually extruded and either breaks up in the plasma or is euglobed by leucocytes, giant cells, or endo- thelia. The significance of this somewhat remarkable phenomenon has not been fully demonstrated. The opponents of this theory claim that free nuclei are seen in sufficient numbers to account for the production of many red cells, only in smears of the marrow, but scantily in sections. The writer finds this objection valid, unless the destruction of extruded nuclei is more rapid than Ave have any good reason to believe. As a modification of the "extrusion" theory, Malassez, Fellner, and Duval believe that the nucleated red cell extrudes a portion of its substance in the form of a red disc entirely free from nuclear material, but the histology of the marrow offers insufficient ground for belief in such a process. Engeli holds that mcgaloblasts produce megalocytes by gemmation, leav- ing normoblasts, from which the nuclei are extruded. Mondino and Sala find evidence that the nuclear material finally diffuses along the periphery of the cell, where it undergoes a chemical change which causes it to lose its affinity for basic dyes. Obrastzow ~ also believes that the nucleus persists in diffuse form in the disc, and that it may reappear as a result of post-mortem processes. Finally, Buttcher and Brandt claim that the nucleus persists and may be demonstrated in the normal red veils by certain staining methods, while Wooldridge, and Botazzi and Capelli. by chemical analy.sis, are always able to demonstrate nuclear material in the red cells. While the weight of opinion and evidence seems to favor the grad- ual fading of the nucleus within the cell, and while it can hardly be denied that extrusion frequently occur.s, there is abundant evidence to show that the red cell is not entirely devoid of nuclear material. The chemical analysis of the stroma of these cells decidedly favors the presence of nuclear elements (Lilienfeld, Wooldridge) ; the appearance of degenerating cells strongly suggests a nuclear remnant, and the em- ERYTHROCYTES. 153 ploymeut of recent stains, neutral red, polychrome blue, frequently yields the specific reaction of chromatin in the centers of slightly altered red cells. The recent studies of Maximow indicate that while the bulk of the nucleus is extruded, a portion remains in finely granu- lar form and imparts a basic staining quality to the centers, specially of young red cells. Seats of Formation of Red Cells. — («) In the early embryo the pro- duction of red cells follows the development of capillary blood vessels in many tissues. Very soon the rudimentary liver and lymph nodes make their appearance and in them the function of red cell formation is gradually concentrated. (6) In the fully developed embryo, the majority of investigators locate the chief formation of red cells in the liver, to a less extent in the spleen, while the bone marrow, which is at first less prominent in this process, gradually absorbs the function, and in the human fetus at birth represents its chief seat. In most infants at birth, however, the liver, and less constantly the spleen, are found to contain many nucleated red cells, while in some cases the proportion of these cells in the liver is very large and nu- merous small collections of them may be found scattered along the portal capillaries. (c) In Extra-uterine Life. — The fact that the red marrow is probably the exclusive depot of formation of red cells in the adult was discov- ered almost simultaneously in 1868 by Neumann^ and by Bizzozero.^ This original claim has been uniformly substantiated by those later observers who recognize as erythroblasts only those cells which contain a trace of Hb. Others, as Lowit and Muller, Gibson, Foa, Saxer, who trace the red cells back to colorless cells, regard all lymphoid tissues as probable sources of erythroblasts. In pathological conditions it has been shown that the spleen (Bizzo- zero,^ Neumann,^ Howell) in the human adult may resume its embry- onal function of red cell formation, as indicated by the presence of many nucleated red cells. Under similar conditions it has been shown that the limits of red marrow, normally confined to the flat bones, ribs, vertebrse, and upper and lower thirds of the long bones, may be ex- tended throughout nearly the entire cavities of all bones except those of the feet. (JSTeumann, Litten and Orth, Bizzozero and Salvioli.) In infancy, and up to the sixteenth year, the cavities of all bones contain almost entirely red marrow, but there are considerable vari- ations in the proportions of lymphoid and fat tissue in the shafts of the long bones. It is probable that in most healthy infants the formation of red cells is already limited to the lymphoid marrow, but considerable numbers of nucleated red cells, often grouped in islands, are rather frequently to be found in the spleen and liver of apparently healthy asphyxiated infants. In the acute and chronic anemias of infancy and childhood, the facility with which the spleen and liver resume their former role of red cell production partly explains the special involve- ment of these organs in such diseases. 154 DEVELOPMENT OF BLOOD CELLS. LEUCOCYTES. Development of Leucocytes. — The earliest indications of the for- mation of leucocytes are seen in the presence of primary wandering cells^ of mesodermal origin, which are found principally in the loose connec- tive tissues of the early embryo. The specific quality of these wander- ing cells must stand as one of the fundamental facts connected with the development of blood cells. Though of mesodermal origin they are from the first quite distinct in morphology and apparently in function from the capillary endothelium and fixed connective tissue cells. The primary wandering cells are of large size, 8-9 li (Saxer), with a single large nucleus containing one or more nucleoli and with a moderate quantity of finely granular slightly acidophile protoplasm. Their de- velopment has been traced by H. E. Ziegler to masses of mesodermal cells surrounding the cords from which the capillaries are formed. It thus appears that the parent leucocytes lie originally outside the vessek^ into which they make their way by virtue of ameboid poAvers. The identity of the primary leucocytes with the primary wandering- cells has been generally accepted, but Schmidt, Bonnet, and others, claim that leucocytes are derived from the capillary endothelium. The weight of evidence, however, is much against this view, which inserts the highly specialized endothelial cell as a transition stage between the primary mesodermal cell and the leucocytes. An isolated position also is assumed by v. Davidoff, Maurer, and others, who believe the leuco- cytes to be direct derivatives of the epithelial cells, principally of the intestine. This view is based upon the appearance of lymphocytes in and between the epithelial cells lining many mucous surfaces, which has been repeatedly and fully explained as an infiltration by leucocytes, not a production of new lymphocytes from these cells. (Saxer.) The white blood cells of vertebrates make their ai)pearance in the circulation only after the development of an extensive vascular system and long after the red cell formation has been fully established. In the lower vertebrates their absence has been noted nine weeks (Ziegler') after the appearance of red cells, but Hay em found them in frogs' blood 34 days after the appearance of red cells. The time of their ap- pearance in the blood of the human embryo has not been accurately fixed. Gulland finds leucocytes in lymph vessels in the human em- bryo of one and one-half inches in length. In form the earliest leuco- cytes of the embryonal circulation are the small and medium-sized basophilic lymphocytes. Process of Development of Leucocytes from Primary Wander- ing Cells. — Most observers find that the primary wandering cells pro- duce by mitotic division one or more generations of colorless cells which gradually approach in morphology the early basophilic leuco- cytes of the circulation. Most of the obscurity which still surrounds the early development of red and white blood cells is connected with this stage of the process, and the numberless conflicting opinions have resulted from the different interpretations placed upon the DEVELOPMENT OF LEUCOCYTES FROM WANDERING CELLS. 155 intermediate stages between primary wandering cells and circulating blood cells. Denys, Lowit/ Ziegler,^ v. d. Stricht, and others claim that red cells and leucocytes develop from separate series of cells which have become differentiated from the primary mesodermal cells with the first appear- ance of blood and blood vessels. Kostianecki," Muller, Schmidt, Saxer, and others, believe that the primary wandering cell persists in the blood-forming organs as the parent of both red and white cells. It is not impossible to partially reconcile these opposing views. While the exhaustive studies of Saxer seem to leave no room to doubt the common origin of embri/onal red and white cells, it by no means follows that the blood cells of the adult may be traced to a common origin. Several generations (3-4, Saxer), each multiplying by indirect division, are required before the primary wandering cell is transformed Fig. 20. Primary wanderinfj Mt^^ Enihryoiidl Liver ftc. cell ^H^^^ Dis(ippe<(rs in late foetal life Saxer'a fl orc/er in order into red cells or leucocytes. It is quite possible that the first mem- bers of the series, including the primary mesodermal cells, long persist in the embryo, but disappear in the adult, in whom, therefore, no com- mon cell of origin of erythrocytes and leucocytes can be demonstrated. A review of the principal minute studies in this field seems to the writer to warrant the above general conclusion in regard to the relation of red and white blood cells. That all blood cells of the em- bryo can be traced to a common origin seems to be fully proven. Although the exhaustive studies of Denys, v. d. Stricht, and Lowit, were largely upon embryonal tissues, the morphological criteria on which they separate primary leucoblasts from erythroblasts do not ap- 156 DEVELOPMENT OF BLOOD CELLS. pear sufficiently distinct to convince other observers, while the clenion- stration by Saxer and others of earlier mitotic cells in the same em- bryonal tissues seems to show that the common cell of origin persists in the fetal liver. Muller's studies of leukemic blood in Avhich he seems to have traced leucocytes and red cells to a common origin in the adult indicates that this pathological condition involves a reversion to the embryonal type of blood formation, but does not prove that a common mother cell of both red and white blood corpuscles persists in the normal adult marrow. (See Fig. 20.) Cellular Divisiox in the Development of Leucocytes. — Since the full demonstration by Flemming of numerous mitotic figures in the cells of the lymph follicles, karyokinesis has been accepted as the chief method, and many claim it to be the only method, of multi- plication of leucocytes. This view has been actively combated by Lowit, who finds that while erythroblasts multiply by mitosis, leucoblasts follow a modified form of the process which he calls divisio indireda per granula. This author has very ably and successfully defended the importance of direct division in mammals and lower vertebrates, and subsequent authors have admitted that this method is of "very widespread'' (Saxer) occurrence among leucocytes. Direct division appears to be of special importance in the development of the giant cells of the embryonal liver and adult marrow, which play an important but somewhat obscure part in blood formation. Indirect fragmentation is a term applied to a somewhat peculiar modification of the usual process of nuclear division described by many authors, especially by Arnold. 1 While the formation of leucocytes occurs principally in the blood- forming organs, many have found in the circulating blood, and espe- cially in the lymph, evidences of active multiplication by mitosis and more frequently by amitosis. (Spronk and Prins, Lowit,' AVertheim.) Sites of Origin of Embryonal Leucocytes. — Before the leucocytes begin to appear in the circulation mitotic figures are abundantly seen in the primary wandering cells in various situations. These are gath- ered in groups, first in the loose connective tissues of various regions where lymph nodes subsequently develop, but the chief seat of the pro- duction of leucocytes, as of red cells, is found in the embryonal liver. In both situations, the wandering cells are found ii\ the lymph and blood capillaries, in the interstices of the connective tissues, and be- tween the liver cells. In later embryonal life the formation of leuco- cytes is gradually transferred from the liver to the lymphoid and adenoid tissues, as indicated by the development of the lymph nodes, spleen, marrow, and thymus. While the majority of observers be- lieve that these tissues are at all times developed by the proliferation of extra-vascular wandering cells and leucoblasts, Stohr and Gulland believe that these cells multiply principally tcitldn the vessels and sub- sequently wander out at certain localities where they collect to form lymph nodes. Formation of Leucocytes in Adult Life.— Under normal condi- THE NEUTROPHILE LEUCOCYTES. 157 tions the reproduction of leucocytes in the adult is limited to the lym- phoid structures, including the lymph nodes, spleen, and marrow. Appearance of Different Varieties of Leucocytes. — In the later months of fetal life, the blood begins to show the presence of other varieties of leucocytes besides the lymphocytes. The minute mor- phology of these early cells and the exact time of their appearance is not known, but it is certain that all varieties seen in the adult are present at birth, and there is much evidence to show that long before this period the development of several distinct types of leucocytes is well established and their formation at least partially limited to certain viscera. From the considerations of the previous part of this section it is evident that all forms of leucocytes must have a common origin in the primary wandering cell, but it does not by any means follow tJiat the various forms of leucocytes in the adult are developinental forms of one series, although this view has been generally accepted as correct. The Basophile Leucocytes. — These cells are the direct descend- ants of the first embryonal leucocytes, and are uudoubtedly produced, in the adult as in the embryo, by the mitotic division of cells lying in the proliferation zones of lymph follicles, and diffusely in all adenoid tis- sues. Virchow regarded the small basophile cells (lymphocytes) as originating in the lymph nodes and the large basophile cells (spleno- cytes) as derivatives of the spleen ; but this distinction has not been supported. It seems more probable, though by no means certain, that most or all of the larger basophilic leucocytes of normal blood are de- veloped from lymphocytes. Kanthaok and Hardy endeavored to sub- divide this class into basophile and hyaline cells, the former being " tissue cells " and rarely appearing in the blood, the latter identical with the ordinary small and large lymphocytes. Kanthack's coarsely granular basophile cell is the " mast-cell,^' which has no connection with the lymphocytes, but his finely granular basophile cell has not been recognized as distinct from the hyaline cells whose protoplasm is not strictly hyaline but reticulated, and which undoubtedly originates in the lymphoid organs. The MAST-CELLS, which contain large, strongly basophilic granules, are classed by Kanthack as celomic (tissue) cells. They are quite dif- ferent in appearance from the coarsely granular basophilic cells of chronically inflamed tissues and their exact origin has not been ex- plained. (Cf. " Mast-cells," Section on Leukemia.) The Neutrophile Leucocytes. — The time of appearance of these cells in the human embryo is not known. Saxer finds in the sheep's embryo of 4.5 cm., in connective tissues and scantily in the lymph, polynuclear leucocytes identical in appearance with those of the adult animal. If the polynuclear leucocytes are developed from lymphocytes, the seat of their production in the infant at birth is of course coextensive with that of lymphocytes, but if polynuclear neutrophile cells are de- rived as a separate series from the neutrophile myelocytes, their forma- tion is limited, at birth and thereafter, to the red marrow. Whichever 158 DEVELOPMENT OF BLOOD CELLS. view may be correct, it is certain that the neiitrophile myelocytes are largely concerned in the formation of these cells, and that the marrow is the chief seat of their production. The opinion widely held that neutrophile leucocytes are developmental forms of the mononuclear basophile cells still lacks any definite support. That they are certainly produced by mitotic division of neutrophile myelo- cytes has, however, been placed beyond question, by the constant appear- ance of increased numbers of mitotic myelocytes in leukemia and leucocytosis. The larger size, distinctly vesicular nucleus, basophilic protoplasm, and in- variable absence of neutrophile granules in the most typical examples of the so called "transitional leucocytes" cannot be reconciled with the view that these cells are ever transformed into neutrophile leucocytes. Some have re- ferred to myelocytes deficient in neutrophile granules as transitional forms, but in leukemia where such cells are common, a deficiency of neutrophile granules is very frequent in both myelocytes and polynuclear leucocytes. Again, if basophile leucocytes form neutrophile cells it is difficult to see why numerous transition forms are not found in lymphatic leukemia where basophile cells are excessively numerous, but neutrophile are abnormally scarce. Lowit here resorts to the suggestion that the further development of these cells is inhibited by changes in the plasma. Yet in myelogenous leukemia no such changes in the plasma prevent the transformation of myelocytes into polynuclear leucocytes which here often outnumber the myelocytes. Out of the constantly accumulating evidence which goes to establish the specific qualities of different leucocytes may be mentioned the studies of Zenoni, who found no change in the proportion of neutrophile cells during very marked variations, experimentally induced, of the basophile. The Eosinophile Leucocytes. — These cells very early become differentiated from the primary leucocytes, having been found in the chick's blood on the fifth day of incubation (Engel ^) and in the human thymus and lymph nodes before the appearance of bone marrow (Schaf- fer, Gulland). Though it was early found tliat they are specially abund- ant in the marrow, their very wide distribution in the tissues shows that the conditions essential to their development are not limited to any single organ or tissue. The discovery of mitotic division of eosinophile leuco- cytes, first reported by Muller and subsequently verified by many others, would seem sufficient proof that these cells constitute a distinct self-per- petuating series. Nevertheless the majority of current writers either tacitly accept or actively argue that the eosinophile are derived from the neutrophile granules of polynuclear leucocytes. Demonstration of the truth of this view is at present entirely lacking. Only Gulland attempts to describe transitional stages between neutrophile and eosinophile gran- ules in embryonal blood cells, but his claims have not only not been veri- fied but are distinctly contradicted by the results of the vast majority of observers, who have failed to find any such transition forms in the adult. On the other hand there is abundant physiological, morphological, and microchemical evidence to show that eosinophile leucocytes are not derived from neutrophile. Physiologically, the eosinophile cells are celomic tissue cells, finding their natural habitat in the tissues and not in the blood, whereas the neutrophile leucocytes are chiefly hemic cells occurring almost exclu- THE BLOOD PLATES. 159 sively in the blood. Moreover, since these cells certainly multiply by mitosis, it is unnecessary and contrary to analogy to suppose that any other method of development or formation exists. The morphological evidence includes the fact that transitional granules have not been demonstrated ; that the nuclei of the two cells have rather distinctive characters, and that the granules are integral parts of the cytoreticulum (Heidenhain, Gulland), and not, as has been supposed, excretory prod- ucts in the cell. The microchemical evidence shows that the eosino- phile granules yield Lilienfeld's and Monti's reaction for phosphorus (Sherrington), contain iron (Barker), and give the vanillin and alde- hyde reactions of Weiss, none of which characters have been demon- strated in ueutrophile granules. THE BLOOD PLATES. The blood plates were first described by Donne, in 1842, who found them in blood upon the addition of water and regarded them as parti- cles of globulin derived from the red cells. In 1847 Zimmermann found in defibriuated blood certain " elementary granules," at first about 1 /^ in diameter, but gradually increasing in size, and, as he be- lieved, eventually becoming red cells. The first complete descriptions were given by Schulze, Kolliker,^ Ranvier, and Bizzozero,* and later by many authors. They are circular or ovoid, colorless, homogeneous, or granular bodies, about 1-3 /j. in diameter, usually showing no nuclear portion, and staining lightly by both basic and acid dyes. The common im- pression that they disappear promptly after shedding is true only of the spindle cells of the blood of lower animals, while human blood- plates are scanty in fresh blood and increase in number upon standing. They are extremely cohesive and apparently of high gravity, and since they collect in masses from which the fibrinous threads of co- agulating blood commonly radiate, they are supposed to take an im- portant part in the formation of fibrin. On account of their uncertain morphology and origin, attempts to enumerate these bodies have not been very successful. In normal blood they may be said to vary between 1 80,000 (Fusari) and 500,- 000 (Pruss). Estimates of their numbers based upon the general appearance of the blood indicate that they are specially abundant in afebrile anemias, leukemia, hemorrhages, and that they are deficient in febrile diseases, malaria, and after the administration of various poisons. (Afanissiew, Limbeck, Fusari, Pizziui.) The significance of the so-called " third corpuscles " of human blood, the blood plates, appears to have been greatly obscured by the discovery in the blood of frogs by Recklinghausen, and of birds by Hayem, and Bizzozero and Torre, of small spindle-shaped elements, which differ from red cells in the absence of hemoglobin and from leucocytes in their simple oval nucleus and non-ameboid protoplasm, and which have been reofarded as the homologue of the human blood 160 DEVELOPMENT OF BLOOD CELLS. plates. Lowit" classes these spindle cells with leucocytes. Muller believes them to be peculiar elements without relation to either red or white cells. Rauvier regarded them as loosened vascular endothelia. The conclusion of Recklinghausen, Gobulew, Schlarewsky, Bizzozero, Hayera, Vulpian, Eberth and Schinimelbusch, that these cells are analogues of the blood plates is the chief ground for the belief in the existence of a third corpuscle in human blood. When the evidence derived from the examination of human blood is considered, it becomes clear that the existence of a third cellular body is without proof. Bizzozero's convincing demonstration of the pres- ence of extremely fragile bodies in the blood of the frog's mesentery has naturally never been repeated in human blood, and the dissimi- larity between human blood plates and the spindle cells of the frog requires much more cogent evidence of the relation of these two bodies than has yet been furnished. Consequently later investigators have with apparent success endeavored to show that the blood plates are not cells but detritus of cells or of plasma. Howell, Gibson, and Hlava, re- gard the blood plates as fragments of the nuclei of disintegrated leu- cocytes. Lowit believes many of them to be precipitated globulin particles, and others to be fragments of degenerating leucocytes. On account of their chemical composition Lilienfeld holds that they con- sist of nuclein, and that they are derived principally from the nuclei of disintegrated leucocytes. Czermak finds them abundantly in the proliferation zones of lymphoid organs, and wdth Mondino and Sala believes them to be fragments of the nuclei of colorless cells. The studies of Klebs, Eugel, Bremer, Wlassow, Arnold,- Maximow, and others, have placed beyond doubt the opinion that the chief source of blood plates is by the extrusion from red cells of masses or chains of globular material, which give many of the reactions of the nucleo- proteids. While this process may be observed in almost every dry specimen stained by methylene-blue (Plate II., Fig. 2), it appears to be favored by proximity to leucocytes, may be accelerated by many reagents, and is probably especially frequent in young red cells. Bibliography. Development of Blood. Afanissiew. Deut. Arch. klin. Med., 1884, Bd. 35, p. 217. Albrecht. Sitznngsb. d. Ges. f. Morph. u. Phvs. Miinchen, 1895, Bd. II., H. 1. Arnold. ' Vircliow's Arcliiv, Bd. 140. ^Cent. f. Path., Bd. 8. Barker. Johns Hopkins Hospital Bull., 1894. Bizzozero. iQaz. med. Ital., Lombard, 1868. 2 Archiv Ital. de Biol., 1891, p. 293. 3 Virchow's Archiv, Bd. 95. *Virchow's Archiv, Bd. 90. Bizzozero, Salvioli. Centralblatt f. med. Wissensch., 1880, p. 737. Bonnet. Grundriss der Entwickelungsgeschichte. Botazzi, Capelli. Archiv Ital. de Biol., 1899, p. 151. Bottcher. Archiv f. micr. Anat., Bd. 14. Brandt. Archiv f. micr. Anat., Bd. XIII. Bremer. Archiv f. micr. Anat., Bd. 45. Davidoff. Archiv f. micr. Anat., 1887, Bd. 29. Dekhuyzen. Anat. Anzeiger, 1891, VI., p. 220. Denys. La Cellule, 1887, I., IV. Disse. Merkel, Bonnet. Ergebenisse, Anat., V. Donne. Compt. Rend. Soc. Biol., 1842. BIBLIOGRAPHY. 161 Duval. Precis d' Histologie, Paris, 1897. Eberth, Schimmelbusch. Virchow's Arcliiv, Bd. 103-105. Ehrlich. Farb. Untersuch. Berlin, 1891, p. 96. Engd. 1 Archiv f. micr. Anat., Bd. 42. 2 Ibid., 1894, Bd. 54. Fellner. Wien. med. Jahrbucher, 1880. Flemming. Merkel u. Bonnet's Ergebniss., Vol. II. Foa. Archiv Ital. de Biol., 1882, p. 436. Francois. Archiv Ital. de Biol., Bd. 13. Fusari. Cent. f. klin. Med., 1887, p. 45. Gibson. Jour. Anat. and Phys., Vol. XX., 1886. Globulew. Sitzungsber. d. kais. Acad. Wissen., Wien., 1868. Gulland. Jour, of Path, and Bact., 1894. Hayem. ' Du Sang, Paris, 1889. ^Compt. Eend. Soc. Biol., 1877, I., p. 86. His. Zeit. f. wissenscli. Zool., Bd. X.-Xl. Hlava. Arch. f. exper. Path. u. Pharm., Bd. XVII. HoweU. Jour, of Morph., 1891, Vol. 4, p. 57. Israel, Pappenheim. Virchow's Archiv, 153. Kanthach, Hardy. Jour, of Physiol., 1894, Vol. 17. Klcbs. Ziegler's Beitriige, Bd. IV. Klein. Sitzungsb. kais. Acad. Wissen., Wien., 1871, Bd. 63. KoUiker. ' Histologie, 1846. ^ Gewebelehre, 1867, p. 629. Kuborn. Anat. Anzeiger, 1890, p. 277. Lilimfeld. Zeit. f. phys. chemie, Bd. 20, 1895, p. 155. Lilienfeld, Monti. Zeit. f. phvsiol. Chem., Bd. 17, p. 410. Litten. Orth. Berl. klin. Woch., 1877, No. 51. Lmvit. • Archiv f. micr. Anat., 1891, Bd. 38, p. 598. ^Port. d. Med., 1888. Vir- chow's Archiv, Bd. 117. Malassez. Archiv f. physiol. norm, et path., 1882, p. 1. Maurer. Morph. Jahrbuch., Bd. 16, p. 203. Massloiv. Archiv f. micr. Anat., Bd. 51. Maximow. Archiv f. Anat. u. Pliysiol. An. Abt., 1899, p. 33. Mondino, Sala. Archiv Ital. de Biol., XII. Muir. Jour, of Anat. and Physiol., 1893, V., 38. Midler. Sitzungsber. d. kais. Acad. d. Wissen., Wien., 1889, Bd. 98. Deut. Archiv f. klin. Med., 1891, Bd. 48. Neumann. 'Zeit. f. klin. Med., Bd. 3. ^ Archiv f. micr. Anat., 1876, Bd. 12. 3 Med. Centralblatt, 1868. Archiv f. micr. Anat., 1876, Bd. 12. Virchow's Archiv, 1890, Bd. 119, p. 386. Obrastzow. 'Cent. f. d. med. Wissen., 1880, p. 433. ^ virchow's Archiv, Bd. 84. Pizzini. Rif. med., X., Nos. 107-108. Pouehet. Quart. Jour. Micr. Sci., 1880, Vol. 20. Pruss. Cent. f. klin. Med., 1887, p. 469. Ranvier. Traite d' Histologie. Recklinghausen. Archiv f. micr. Anat., Bd. II. Remak. Canstatt's Jahresber., 1841. Rindfleisch. Archiv f. micr. Anat., 1879, Bd. 17. Saxer. Anat. Hefte, Bd. VI., 1896. Schajer. Quain's Anat. Hist., 1891, p. 218. Schaffer. Cent. f. d. med. Wissen., 1891. M. Schmidt. Ziegler's Beitriige, 1891, Bd. XI. Schlaretvsky. Cent. f. med. Wissen., 1867. Schulze. Arch. f. micr. Anat., Bd. 1865. Sherrington. Proc. Royal Soc, 1894, Vol. 55, p. 161. Spronk, Prins. Cited by Saxer. Spuler. Archiv f. micr. Anat., Bd. 40. Stohr. Text-book of Histology. V. d. Stricht. Archiv de Biologic, XII., 1892, p. 199. Strieker. Gewebelehre. Vulpian. Compt. Rend. Soc. Biol., 1877. Wertheim. Zeit. f. Heilk., 1891, Bd. XII. Wissosky. Archiv f. micr. Anat., 1877, Bd. 13. Wlassnw. Ziegler's Beitriige, Bd. 15. Wooldridge. Chemistry of Blood, London, 1893. Ziegler. ' Ber. d. naturforscher Gesell. Freiburg, 1889, Bd. IV., 171. ^Archiv f. micr. Anat., Bd. 30. Zivimennan. Rust's Mag. f. d. gesam. Heilkunde, 1846, Bd. 66, H. 2. 11 i PART 11. SPECIAL PATHOLOGY OF THE BLOOD. CHAPTER VI. CHLOROSIS. Chlorosis is a primary anemia occurring almost exclusively in young women, rarely in young men, resulting from defective hemato- genesis and affecting principally the hemoglobin but secondarily, also the number of red cells. Etiology. — The studies of the etiology of chlorosis having estab- lished a great variety of predisposing and exciting causes, and rendered it evident that the disease results under very many diverse conditions, have left the essential factor in the evolution of the disease undeter- mined. On this account some are inclined to regard chlorosis as a symptomatic anemia, holding the peculiar clinical characters of the disease to result from the age and constitution of the individuals in whom it occurs. (Kruger.) Regarding the general causes a number of theories, supported by many important facts, have been brought to light. Hypoplasia of the arterial system is an anatomical ground- work shown by Rokitansky and Virchow to exist in a certain group of cases. This hypoplasia may affect only the heart and larger arteries or the genital organs as well, or is frequently associated with the other developmental anomalies of the conditutlo lymphatica. More re- cent studies (Paltauf, Ortner, Fraentzel) indicate that the congenitally small heart and aorta are more closely associated with the " constitutio li/inphatica'^ and with cardiac disease, than with chlorosis. This hypo- plasia has been found in subjects never suifering from chlorosis, and the majority of chlorotics give no other evidence of such vascular anomalies and make complete and permanent recovery. Intestinal autointoxication was first suggested as the essential cause of chlorosis by Hoffman, and this theory has been supported on various clinical and pathological grounds by Duclos, Clark, Nothnagel, Bouchard, and many others. Bunge's authority has lent some force to the autotoxic theory as this investigator holds that the blood obtains iron solely from the nucleo-albumens, the iron from which is combined with sulphur during intestinal putrefaction and thus becomes nonab- sorbable. In chlorosis the iron administered by mouth combines with the H^S and permits the normal absorption of the nucleo-albumens. 164 SPECIAL PATHOLOGY OF THE BLOOD. This theory has been completely set aside by the demonstration by V. Noorden/ Rethers, Moruer, and Lipraan and Wulf, that in chlorosis increased intestinal putrefaction is not commonly present, nor is there increased excretion of the derivatives of Hb. Stockman showed, moreover, that sulphate of iron will cure the disease, while bismuth and some other absorbents of H^S will not. Intestinal intoxication can, therefore, figure in only a limited group of cases. V. Hosslin, finding marked increase of iron in the feces of some chlorotic patients, concluded that the blood changes result from small multiple intestinal hemorrhages. Luton accepts this theory. Functional disturbances of the nervous system have from the first been actively maintained as prominent or exclusive etiological factors in chlorosis. The train of events here involved is variously explained. Murri supposes that there is a vasomotor disturbance arising from the generative organs, and causing changes first in the rate of flow, later in the chemistry of the blood. This theory represents the culmination of opinion as handed down by the older physicians. (Trousseau, cf. Lloyd Jones.) According to Meinert the nervous in- fluence arises from irritation of the abdominal sympathetic and results from gastropfosis. This observer has brought together very strong evi- dence to show that in one group of cases the above condition must be an important factor in the coincident anemia. Kruger believes that hypertrophy of the spleen commonly associated with chlorosis (Fuhrer, Scharlau, Chvostek, Clement, Grawitz, Rummo, and Dorri) is an indi- cation of disturbance of function in this organ which results in an in- creased destruction of Hb. This disturbance of function is believed to arise through irritation of the sympathetic, in the genital organs, stomach, intestines, etc. Predisposing Causes. — The almost exclusive occurrence of the disease near the establishment of menstruation must prove any theory of etiology inadequate which fails to consider the aye of puberty, the female sc.r, and the function of )iienstn(ation. There is abundant evi- dence to show that heredity also is an important factor, as the disease is often observed in the daughters of chlorotic mothers, while a tuber- culous family history has been remarkably frequent in some series of cases. (Trousseau, Jolly, Bramwell.) Great importance undoubtedly attaches to general lack of hygiene, especially to the combination of poor food and mental depression. Stockman's analyses indicate that the diet of chlorotic girls contains as a rule too little iron. Considering the evidence thus briefly reviewed, the writer concludes that chlorosis is a specific form of essential anemia which results from the association of several factors, of which the two most important and invariably present are (1) a predisposition to the disease, and (2) de- fective absorption of iron from the food. The predisposing factors are most pronounced in young girls at the establishment of menstruation when the unaccustomed losses of blood and the peculiar changes in the nervous system, favor the ex- PLATE [II. Mild Chlorosis. (Eosin and Methylene Blue.) Fig. I. Slightly deformed red cells, with enlarged central clear areas, indicating loss of Hb. Fig. 2. Poikilocyte. Figs. 3. Normal rouleaux. Fig. 4. Medium-sized lymphocyte. Fig. 5. Polynuclear neutrophile leucocyte. CHANGES IN THE BLOOD. 165 pression of hereditary tendencies and exaggerate the effects of bad hygiene, and when additional elements are thrown in the balance, such as improper food, constipation, gastroptosis, mental depression, etc., the blood begins to suifer in a peculiar way and clinical symptoms are established which are not seen in secondary anemias occurring in the same subjects. The essential importance of the absorption of iron is indicated by its specific action, but it seems unnecessary to conclude, as Stockman has done, that the food must be deficient in iron, provided digestion and absorption are defective. It seems to be an invariable rule that chlorotics do not recover until the appetite improves. The exact nature of the pathogenic process in the disease remains, however, untouched by these considerations, and it still remains un- determined just how and where the elaboration of Hb is disturbed, whether in the intestinal mucosa (Garrod, Forcheimer), or in the spleen (Kruger), or in the bone marrow. At present the weight of evidence appears to be almost conclusive that chlorosis results from a functional insufficiency of the bone marrow, brought about in congenitally pre- disposed subjects by a series of conditions most often combined in young women at puberty. Changes in the Blood. Specific Gravity. — The specific gravity of the blood in chlorosis is reduced in very uniform ratio with the loss in Hb. From the observations of Devoto, Hammarschlag, Schmaltz, Siegel, Hock and Schlesinger, Menicanti, Stinzing, and Biernacki, it has been shown that this constant relation between specific gravity and Hb-content is somewhat peculiar to chlorosis. It is most uniform when there is no great reduction in cells, but in the severer types of the disease the ratio is less constant. Jones found the specific gravity of the blood between 1.032-1.045 in 87 cases, while that of the serum remained nearly normal, 1.0259 -1.029. These results were verified by Hammarschlag in 30 cases, in which the blood averaged 1.045, the serum 1.030. The reduction in gravity of chlorosis corresponds, according to Jones, to a physio- logical diminution occurring in girls at puberty, but not in boys. The bulk of blood is probably not altered in uncomplicated chlorosis, although Lloyd Jones considers an increase in the volume of plasma to be a very constant feature, and Stintzing and Gumprecht believe that oligemia exists in some peculiar cases. The exuded drop is pale in proportion to the anemia, and abnormally fluid. The Hemoglobin. — The chief alteration in the blood is the loss of Hb which is so far out of proportion to the reduction in red cells that a low Hb-index is one of the diagnostic features of the disease. A contrary opinion is held by Eichhorst and especially by Biernacki who, finding, as did also Zumpf, a normal or increased percentage of iron in the dried residue of the blood in some cases, believe that a loss of other albumens and not especially of Hb is the essential change in the disease. The basis of this dissenting opinion will be further consid- ered later. The percentage of Hb obtained by modern instruments has been 166 SPECIAL PATHOLOGY OF THE BLOOD. reported as low as 10 (Bramwell) ; but when the Hb falls below 20 percent there should always be a suspicion of some complicated or secondary anemia. In cases of average severity the Hb ranges between 35-45 percent, and the Hb-index about .5. Yet some authors record severe cases with high Hb-index (.80-.95, Bramwell), although from the details of these reports it is not clear to what this variation from the rule is re- ferable. These cases, in the writer's experience, fall in an intermedi- ate group between chlorosis and pernicious anemia and are more refrac- tory than the others. Occasionally the Hb-index is very low (.20). (Bramwell.) v. Noorden,' from extensive observations, finds that the Hb-index is apt to be somewhat higher in recurrent than in initial attacks, but Romberg could not verify this statement in a group of relapsing cases. The writer has seen illustrations of v. Noorden's rule and, since in- crease in size of cells is an indication of chronicity in anemia, believes that the observation is, in general, well founded, yet as Romberg states, the character of the blood changes is controlled by individual peculiarities in blood formation and by numerous unhygienic condi- tions. The red cells are in most cases slightly reduced, although almost invariably much less affected than the Hb. Yet not infrequently a considerable grade of anemia is associated with a normal proportion of red cells. A high count of red cells is a favorable prognostic sign. In cases of average severity the red cells run from 3.5 to four mil- lions, while in severe cases the number falls below three millions. Ex- ceptional examples in which less than two million cells were found are recorded by Bramwell, Cabot, and many others, while Limbeck records one case with 1,190,000, and Hayem a minimum of 937,360. In such cases the Hb-index is often comparatively high. In morphology the red cells present characteristic changes. In mild cases and those of average severity, the sole alteration is a loss of Hb, which causes a progressive widening of the central clear area to be ob- served in cells rather thickly spread and slowly dried. In all well- marked cases the appearance of such blood in stained smears is charac- teristic of simple anemia. It is always possible, however, to find a few very small cells and some extremely deficient in Hb, but in the less severe cases these features are not pronounced. (Plate III.) When chlorosis is of long standing or of very severe grade, with hydremic plasma, and showing a tendency to relapse, the red cells begin to exhibit marked variations in size and to some extent also in shape. In most cases in which the response to iron is fairly prompt these changes do not affect any considerable proportion of cells. In others in which the response to iron is less certain there is a moderate number of large cells with abundant Hb. (Plate IV.) In no case do these changes reach the grade seen in pernicious anemia, but in a group of cases which do not respond well to iron and are apt to relapse the variations in size of the cells is considerable, and CHANGES IN THE BLOOD. 167 many show an abundance of Hb. On the other hand the writer has seen persistently relapsing chlorosis in which all the cells, varying con- siderably in size, were very deficient in Hb. PoikUocytosis may be seen in all severe cases of chlorosis but is seldom very marked. Other changes occurring with moderate frequency in the red cells of chlorosis are the polychromatic degeneration of Maragliano and that of Gabritschewsky. Granular degeneration is seen to a slight extent in well-marked cases. A qualitative change in the Hb in chlorosis is suggested by the re- sults of Henrique's experiments, by which he finds that oxyhemoglo- bin is much more slowly reduced in chlorosis than in health or in other forms of anemia. Nucleated red cells are seen in the severer cases of chlorosis under various conditions. It is said that when present they are sub- ject to periodical variations constituting the " blood crises." (Neu- dorfer.) Most writers have found them too scanty even in severe cases to permit a full verification that " blood crises " are of frequent occurrence, and this common experience is in accord with the theory that functional insufficiency of the marrow is essentially connected with the pathogenesis of the disease. Their appearance is somewhat favored by rest in bed and large doses of iron, which, as Hoffman has shown, induce hyperemia and cellular hyperplasia of the red marrow. The writer has seen an increased number during the leucocytosis of an inter- current pneumonia. The nucleated red cells of chlorosis are of the normoblastic type, and while Neudorfer and Hammarschlag report finding megaloblasts in severe relapsing cases, the exact nature of these cases is not entirely clear. Their observations require confirmation. Chemistry. — The alkalinity of chlorotic blood has been found by the majority of investigators to vary within normal limits (Kraus, Steindler, Limbeck) or in some cases to be slightly increased (Graeber, Peiper, Rumpf ). v. Jaksch, however, found it diminished. The ISOTONIC TENSION of the red cells was found by Limbeck in two cases to be very low, .38-.4 percent NaCl, their resistance there- fore being correspondingly diminished. The COAGULABILITY of chlorotic blood is notably greater than that of other forms of anemia of equal grade, a rule which accords with the fact of the preservation of the albumens of the plasma. (Cf. Biernacki.) The PROPORTION OF WATER in chlorotic blood increases from the normal (77 percent) to 80-90 percent, and the dry residue falls from 22-36 percent to 19 percent, with 60-70 percent Hb, or as low as 10.64 percent of residue with 25 percent Hb. In this residue the chlorides, as in all hydremic states, are in most cases considerably in- creased. Phosphorus is moderately reduced, and Fe is deficient in the well-marked cases only. (Biernacki.) When the Hb is below 70 percent the reduction in Fe is rather uniformly proportionate with the loss in Hb. 168 SPECIAL PATHOLOGY OF THE BLOOD. Albumens. — Since the specific gravity of the serum is but slightly altered in the average case of chlorosis, the loss of substance must af- fect principally the red cells. It is very generally agreed that, as shown by Becquerel and Rodier, the principle chiefly affected is the Hb, but Biernacki found a normal proportion of iron in the dry resi- due of some milder cases of chlorosis and has concluded that other albumenous constituents may suifer more than the Hb. In Biernacki's analyses of more severe cases, however, a marked and uniform reduc- tion of Fe is recorded. In all the severer cases the serum is poor in albumens, but markedly so only in very severe cases. The hydremia of the serum in chlorosis is relatively far less than in simple anemia, e. g., post-hemorrhagic. (Grawitz.) This rule is found by Stintzing and Gumprecht to hold only when the Hb is between 50-80 percent, as illustrated in the fol- lowing table : Hb5«. 21-30 31-40 41-50 51-60 61-70 71-80 81-85 Dry Simple Anemia 13.4 % 12.0 13.3 14.5 16.1 18.3 18.5 Residue Chlorosis 12.5 13.4 15.8 18.8 19. Regeneration of the Blood in Chlorosis. Red Cells and Hb. — Hayem divided the process of regeneration of the blood in chlorosis into two periods, one in which the red cells are increased in number by the appearance of many small pale and deformed cells, followed by a second in which these new cells gradually acquire normal characters. Similar observations have been made by the majority of recent ob- servers who have regarded the small cells seen in relative excess in well-marked cases as being recently formed. On the other hand, Laache, Stintzing and Gumprecht, Reinert, and Graeber, find that the increase in Hb usually outstrips that of the red cells. More recent studies in this field have thrown some light on the difficulty, Romberg showing that with a slight decrease of cells the Hb increases the more rapidly, while in cases with marked decrease in cells these increase first and more rapidly but less uniformly than the Hb. In some severe cases Romberg noted a slight initial diminution in cells, following the use of iron. Schaumann and Willebrand have recently fully demonstrated that in moderately severe cases, under iron, there is a progressive increase in the average diameter of the red cells, resulting from the disappearance of undersized corpuscles, while the number of large cells becomes con- siderable. The authors feel justified in concluding that the progress of regeneration is not marked by the appearance of many small pale cells, which they believe are degenerated forms. In Romberg's 100 cases, in all of which the cells were below 4 millions, the average duration of treatment was 26.5 days, the Hb increasing 9.9 percent and the red cells 430,645 every ten days. The best results followed the use of ferri carb. sac, but they did not use Bland's pills. PLATE IV Severe Chlorosis. (Eosiii and Methylene Blue. g- 3 g-4 g-5 g-6 g- 7 I. Short rouleau of cells very deficient in Hb. Medium-sized megalocyte deficient in Hb. Microcytes deficient in Hb. Large red cell with central droplet of Hb. Red cell with central basophilic nuclear remnant. (Polychromasia of Maragliano.) Normoblast. Polynulcear neutrophile leucocyte, over-stained with methylene blue and showing basophilic cytoreticulum. g. 8. Blood plates. VARIETIES OF CHLOROSIS. 169 Distinct polycythemia has in many instances been found to follow re- covery from chlorosis. Schaumann and Willebrand report 7.4 million cells with 67 percent Hb, but found that the polycythemia usually diminishes as the Hb reaches a normal proportion. In some severe cases, on the other hand, it seems to be difficult to secure complete re- generation of the blood and the anemia relapses. An increase (12 pounds) in the patient's weight without correspond- ing increase in the Hb of the blood cells, but with increase in the solids of the serum, has been reported by Grawitz, as an interesting feature of the regeneration of the blood in chlorosis. A moderate in- crease in the number of nucleated red cells and of leucocytes is some- times observed in cases rapidly recovering. Leucocytes. — The leucocytes in the average case of chlorosis show no abnormal variation in numbers or proportions. (Graeber, Grawitz.) The average in 239 cases reported by Thayer, Cabot, and Romberg, was 8,013. Romberg's tables show a progressive diminution of leu- cocytes as the anemia grows more severe, and he refers this rule not to any absolute decrease in the number of cells but to increase in the volume of plasma. Neusser holds that some less favorable types of chlorosis, without regard to their severity, show an excess of lymphocytes while in others the presence of eosins indicates a more favorable type of the disease. In the latter group of cases Neusser's pupils found, during treatment with iron, a slight mixed leucocytosis in which the proportions of various cells are normal. The occurrence of myelocytes, as reported by Hammarschlag and Neudorfer is very unusual. Varieties of Chlorosis. — (a) The typical examples of the disease observed in young w^omen form a central group of cases which has furnished the classical and accepted picture of chlorosis. The blood in these cases shows a low Hb-index, with or without marked loss of cells. They run an acute or chronic course but respond well to iron. On chemical analysis the dry residue and the Fe are much diminished and the symptoms are in proportion to the grade of anemia. Of these cases one may distinguish three groups of somewhat dif- ferent prognosis. 1 . The red cells are slightly or not at all reduced (about 4 millions) ; the Hb-index is low ; there are no changes in the size and shape of the cells. Such patients commonly recover very promptly and the anemia does not relapse. 2. The red cells are below 4 millions, but there are no marked changes in the shape and size of the cells ; the Hb-index is very low, the patients are prostrated but with iron usually recover promptly, and the anemia does not often relapse. 3. The red cells are markedly reduced (below 3-3.5 millions) ; there are considerable changes in shape and size, the Hb-index varies, but is usually very low, the patients respond more slowly to iron, and the anemia often relapses. 170 SPECIAL PATHOLOGY OF THE BLOOD. (b) A group of cases of idiopathic anemia common in young women has long been recognized, in which the symptoms resemble those of re- fractory chlorosis, but which in some other respects are quite different. Delafield states that they are chiefly distinguished by their failure to react to iron. A chronic course with relapses is commonly observed, but a change to distinct pernicious anemia rarely occurs. The blood of such subjects shows a marked reduction in red cells, commonly between two and three millions, and marked variation in size of the cells, a considerable number of which have increased Hb. The Hb- index may be very low, but is apt to be higher than in the average case of chlorosis. It is possible that some of these cases are much more anemic than the ordinary blood-examination indicates, because complicated by oligo- plasmia, as suggested by Lloyd Jones and Biernacki. (c) Pseudo-chlorosis is a term employed by various recent writers to denote a group of cases in which the external symptoms of chlorosis are present but in which, from the examination of the blood, it is found that the Hb, number of red cells, percentage of dry residue, and usu- ally the proportion of iron, are nearly normal. Several explanations nave been offered regarding these obscure cases. Lloyd Jones believes that they are cases of anemia with oUgoplas- mia, in which the real deficiency of Hb and cells is masked by dimi- nution in the quantity of plasma. He offers no chemical analyses to support his view. Biernacki has studied this condition in detail and as a result of clinical observations and chemical analyses, he concludes that in cases of anemia there is a condition of oligoplasmia with hydremia of both cells and plasma, so that while the blood shows normal Hb, yet the swollen cells are deficient in this principle, and both cells and plasma are deficient in albumens. In short, Biernacki claims to have demonstrated that it is possible to have hydremic blood which yet gives a normal percentage of Hb and a normal number of red cells. Biernacki points out that two of the eight original cases ex- amined by Becquerel and Rodier, and two cases reported by Stintzing and Gumprecht showed slight hydremia, a nearly normal percentage of Fe, but well-marked symptoms of chlorosis. He protests against calling such cases pseudo-chlorosis, believing that they are genuine examples of the disease, and claims that the essential element in chlo- rosis is not the loss of Hb and Fe, but that the disease is a neurosis with hydremia, and with or tvithout loss of Fe. If Biernacki's results obtain confirmation, which they have not yet secured, much needed light may be thrown upon a large group of cases in which the symptoms of anemia are associated with apparently normal blood. Such cases occur usually in young women, are often encountered in gynecological clinics because of amenorrhea, and have been fully described especially by Biernacki, and Romberg. It is possible, also, that a similar explanation may apply to the cases of pseudo-chlorosis described by Stintzing and Gumprecht and others, and very commonly encountered in tuberculous subjects. BIBLIOGRAPHY. 171 Bibliography. Chloeosis. Biernacki. Zeit. f. klin. Med., Bd. 24, p. 500. Bihler. Deut. Archiv klin. Med., Bd. 52, p. 281. Bouchard. Lecons sur 1' autointoxication, Paris, 1887. Bramwell. Anaemia, etc., p. 22. Bunge. Verb. XIII. Cong. inn. Med., 1895. Chvostek. Wien. klin. Woch., 189,3, p. 487. Clark. Lancet, 1887, II. , p. 1003. Clement. Lyon Med., 1894, T. 75, p. 179. Devoto. Zeit. f. Heilk., Bd. 9. Duclos. Rev. gen. d. clin. et ther., 1887, p. 561. Duncan. Sitzungsber. d. Wien. Acad., 1867. Eichhorst. Cited by v. Noorden. Euren. Schmidt's Jahrbucber, Bd. 240, p. 213. Forcheimer. Amer. .Jour. Med. Sci., Vol. 106. Fraentzel. Zeit. f. klin. Med., 1879. Fuhrer. Archiv f. Heilk., 1856, p. 67. Garrod. Archiv of Path, and Bact., 1892, I., p. 195. Graeber. Therapeut. Monatsbefte, 1887, p. 380. Arb. a. d. med.-klin. Institut Miinchen, Leipsic, 1890. Gram. Fort. d. Med., 1884, p. 33. Grawitz. Klin. Pathol., p. 76. Hammarschlag. Zeit. f. klin. Med., Bd. 21, p. 475. Wien. med. Presse, 1894, p. 1068. Hay em. Du sang., p. 619. Hock, Schlesinger. Beitriige znr Kinderheilk., II., N. F., 1892. Hoffman. Lehrb. d. Constitutionskrank., 1893. Miinch. med. Woch., 1899, p. 949. V. 'Hosslin. Miinch. med. Woch., 1890, p. 248. V. Jaksch. Prager med. Woch., 1890, No. 31. Jolly. Influence d. tuberc. sur Chlorose, Paris, 1890. Jones. Chlorosis, London, 1897, p. 23. Kraus. Zeit. f. Heilk., Bd. 10. Kruger. St. Petersburg med. Woch., 1892, p. 471. Laaehe. Die Anaemic, 1883, p. 81. Cited in Hoffman's Lehrbuch d. Constitutions- krank., Stuttgart, 1893. Lipman, Wulf. Ueber Eiweisszersetzung bei Chlorose, Berlin, 1892. Cited by V. Noorden. Lichtenslern. Untersuch. u. d. Hemoglobingehalt d. Blut., 1878. Limbeck. Grundriss, etc., p. 305. Luton. Bull. Soc. med. Rheims, No. 10. Meincrt. Samml. klin. Vortriige, 1895. Menicanti. Deut. Archiv klin. Med., Bd. 50, p. 407. Morner. Zeit. f. physiol. Chem., 1893, XXIII. Murri. Semaine Med., 1894, 162. Neudorfer. Wien. med. Presse, 1894, p. 1068. V. Noorden. ' Chlorose Wien. , 1897. ^Berl. klin. Woch., 1895, p. 181. ^ibid., 1894, p. 786. Nothnagel. Wien. med. Presse, 1891, No. 51. Ortner. Wien. klin. Woch., 1891, p. 2. PaltauJ. Wien. klin. Woch., 1889, p. 877 ; 1890, p. 172. Peiper. Cent. f. inn. Med., 1891, No. 12. Pick. Wien. klin. Woch., 1891, p. 939. Relhers. Diss. Berlin., 1891. Romberg. Berl. klin. Woch., 1897, p. 533. Rumpf. Cent. f. klin. Med., 1891, p. 441. Rummo, Dori. Rif. Med., 1893, H. III. Scharlau. Theoretisch. Prakt. Abhand., Stettin, 1854, No. 3. Cited by v. Noor- den, Nothnagel, VIII., 2, p. 87. Schaumxinn, Willebrand. Berl. klin. Woch., 1899, p. 9. Schmaltz. Deut. Archiv klin. Med., Bd. 47, p. 145. Siegel. Wien. klin. Woch., 1891, p. 606. Stockman. Brit. med. Jour., 1893, I., p. 880 ; 1895, II., p. 1473. Steindler. Cent. f. inn. Med., 1895, No. 27. Stintzing. XII. Cong. f. inn. Med. Stintzing, Gumprecht. Deut. Archiv klin. Med., Bd. 53, p. 287. Subbotin. Zeit. f. Biol., 1871, Bd. 7. Zumpf. Inaug. Diss. Dorpat, 1891. CHAPTER VII. PROGRESSIVE PERNICIOUS ANEMIA. This term is applied to a peculiar group of diseases of the blood re- sulting from defective hematogenesis and excessive hematolysis, charac- terized by severe and usually fatal anemia, peculiar morphological changes in the red cells, and by characteristic changes in the bone marrow. Although excited by a number of well-recognized conditions, progressive pernicious anemia differs from secondary anemia in the peculiar changes in the red cells, in the lesions of the marrow, in its tendency when once established to progress to a fatal issue, and in the fact that the anemia is apparently out of proportion to the exciting cause, while very often no cause whatever can be assigned. The disease has often been described as primary pernicious anemia, but since typical examples have been positively traced to the effects of intestinal parasites, it becomes necessary to admit that this condition of the blood is not always of cryptogenic origin. Nevertheless its patho- logical features remain so distinctive as to justify the common belief in the peculiar nature of the disease. Hldorlcal. — A review of the early literature on pernicious anemia, as collected by Eichhorst, Musser, and Stockman, shows that while the disease had previously been described by observers in many countries its symptomatology, gross pathology, and separation from chlorosis and secondary anemia, were not clearly stated until Addison in 1855 pub- lished the description of cases observed by him clinically and examined at autopsy during the previous decade. Of the earlier reported cases those of Andral (1821) in France, of Coombe (1823) and Hall (1843) in England, and of Channing (1849) in America, were so clearly described as to leave no doubt that fatal idiopathic anemia was fully recognized by some observers at that early time. Channing, and other Boston physicians, had notes on such cases as early as 1832, and his descriptions of cases in 1849 were sufficiently complete to warrant their acceptance as referring to genuine examples of the disease. Lebert in Zurich had apparently recognized and partially described the disease as early as 1854, but his pathological reports were too scanty to demonstrate the probable specific nature of the malady which he did not claim till 1858, although recognizing it as a fatal form of anemia. Long before this, as shown by Lepine, some prominent text- books had called attention to the occurrence of fatal cases of anemia, considering them usually as instances of fatal chlorosis. (Piorry, Wunderlich, Schonlein, Rokitansky, Canstatt.) PROGRESSIVE PERNICIOUS ANEMIA. 173 Addison, however, offered more abundant clinical and pathological evidence to show that the disease is of idiopathic origin, fully distin- guishable from secondary anemia, but he supposed that the fatty changes were the cause not the result of the anemia. In 1857 Wilks' description of seven cases was very clear and indicates that the condition had then become well identified at Guy's Hospital and was distin- guished from secondary anemia and, by the examination of the blood, from leukemia. In Germany there were fragmentary reports of probable cases by Zenker, 1856; Wagner, 1859; and Grohe, 1861 ; until Biermer in 1868 collected a series of cases observed in Zurich during previous years and gave such a systematic description of the clinical symptoms as to attract general attention to the malady as a special form of anemia. Biermer failed to add to the knowledge of the pathology of the disease and paid little attention to the blood and when Immerman contributed important critical studies in 1874-1877, interest in the microscopical examination of the blood concerned chiefly the proportion of leucocytes. The early observers, especially Wilks, had noted the marked olige- mia, hydremia, and feeble coagulation of the blood at autopsy, and by microscopical examination during life had demonstrated an excessive loss of red cells and a small proportion of leucocytes, but for many years no further advance was attempted in this field. In 1873 Pon- fick found a relative increase of leucocytes in one case, probably from ante-mortem leucocytosis, and noted a greenish tinge of the serum. In 1876 Quincke described the appearance of microcytes and poiki- locytes in several cases, while Eichhorst, at the same time, noted the absence of rouleaux and the increase of Hb in some microcytes, cells which he regarded as pathognomonic of the disease. By this time Bizzozero's and Neumann's demonstration of the function of the bone marrow in blood formation began to turn atten- tion to this tissue. Pepper (with Tyson) making the first contributions in a new field in 1875. Pepper found uncertain evidences of a cellu- lar hyperplasia in some portions of the marrow, that of the radius appearing paler than normal while that of the sternum was quite red. A great advance was made when Cohnheim in 1876 presented a brief communication on the changes in the bone marrow in a case of pernicious anemia. He reported finding an unusual hyperplasia of the red marrow with complete atrophy of fat cells, increase of myelo- cytes, large and small, absence of normal red cells and the presence of megalocytes, microcytes, and large numbers of megaloblasts. In the blood of the vena cava he found a few large nucleated red cells but had failed to find them in the peripheral blood during life. Cohn- heim expressed the opinion that the essential element in the malady was a primary disease of the marrow affecting the development of the red corpuscles, leading to production of megalocytes at the expense of smaller properly functionating cells, the whole constituting a reversion to the embryonal type of blood formation. Cohnheim thus completed the chain of pathological evidence that was required to establish the dis- 174 PROGRESSIVE PERNICIOUS ANEMIA. tinctive nature of the disease, and demonstrated the importance of closer study of the blood as the seat of the primary lesion. The fur- ther study of the marrow was taken up by Osier and Gardner, Neu- mann, Litten and Orth, P. Grawitz,- and many others, whose results have shown that the study of the changes in the marrow gives the deepest insight into the essential process of the disease. The first enumeration of red cells was reported by Sorensen, who, using Malassez's method, found in one case only 470,000 cells ; by Lepine ^ who, with Hayem's method, reported a reduction to 909,000- 378,750, and by Ferrand who found 500,000 red cells and 10 per- cent of Hb. Of morphological changes an increase in size of the red cells during the course of the disease was noted by Lepine ; a slight brownish tinge by Strieker ; the presence of ameboid processes by Scheby-Buch ; the oval form and absence of rouleaux by Bradbury. In 1876 Quincke described the appearance of microcytes and poikilo- cytes in several cases. Eichhorst (1878) found the most striking char- acteristics to be the large size, pallor, and scarcity of the red cells. Yet all these characteristics Eichhorst had seen in a case of gastric cancer, whence he maintained that the blood of pernicious anemia showed no single pathognomonic sign. While Quincke had observed that many microcytes seemed to contain an excess of Hb and that the blood of pernicious anemia was more highly colored than the number of red cells warranted, it was Hayem ^ who first demonstrated that an increased Hb-index is characteristic of the disease, while Laache showed this feature to be referable to the presence of many megalocytes with increased Hb. By 1880 there were only two reported cases in which nucleated red cells had been found during life, when Ehrlich, examining stained dry preparations claimed that while nucleated red cells were found in all cases of severe anemia, megaloblasts were seen exclusively in progres- sive pernicious anemia. Only in recent years has this claim been shown to be but partially valid. In 1891 Luzet first reported the discovery of mitotic nuclei in megaloblasts from the blood of a child, the occurrence of which has since been shown to be a nearly pathogno- monic sign of the disease in adults. With the observations on poly- chromatic degenerative changes in red cells these lesions were found to be specially distinct in pernicious anemia. In one of v. Noorden's dissertations was first mentioned the occurrence of basic staining gran- ules in the bodies of megalocytes, which was later verified by Askanazy and interpreted as evidence of karyorhexis. The contributions on etiology which have steadily accumulated from many sources, and the progress of the knowledge of the chemistry of the blood, chiefly through the work of v. Jaksch, Grawitz, Dieballa, et al., will be considered under those topics. General Etiology. — After the demonstration of striking and pecu- liar changes in the blood and marrow it was generally accepted that progressive pernicious anemia is of idiopathic origin, although Biermer enumerated a list of exciting causes which appeared to have been present SCOPE OF TERM PROGRESSIVE PERNICIOUS ANEMIA. 175 in his cases, viz, insufficient food, bad hygiene, prolonged diarrhea, re- peated hemorrhages. Yet it appeared that the assignable causes were inadequate to account for the result and it was felt that the disease could in no respect be classed with the ordinary cases of secondary anemia such as follow carcinoma, nephritis, hemorrhages, etc. The first retrenchment of this view was rendered necessary by the discovery of typical cases of the disease referable to the presence of in- testinal parasites (Bofhriocephalus) and which recovered promptly after expulsion of the worms. These cases showed that there is not neces- sarily any cryptogenic factor in the etiology of the disease, and that the presence of the typical blood changes does not necessarily imply a fatal course of the anemia. Likewise the changes in the marrow were, for a time, robbed of much of their pathognomonic value by the report that very similar changes are established in some cases of multiple hemorrhages (Neumann), tuberculosis (Litten), and gastric cancer (Eisenlohr). Further, the changes in the blood have now been shown to follow syphilis (F. Muller ^), ^malaria (Bignami,^ the writer) ; atrophy of gastro-intestinal mucosa (Fenwick, Brabazon, Henry, Osier, Noth- nagel, Stengel) ; typhoid fever (Quincke, Rosenstein, P. Grawitz) ; pro- longed diarrhea (Wilks, Frankel, Strieker) ; myeloid sarcoma (Litten) ; in each of which conditions it is occasionally impossible to say that the cause is inadequate to the result, and there remain only the acute cases without demonstrable origin but referred by some to intestinal auto- intoxication, and the cases that follow pregnancy, which can still be held as of cryptogenic nature. Scope of the Term Progressive Pernicious Anemia. — In view of this present state of the subject two plans of classification of the severe anemias of the Addison-Biermer type have been suggested by Birch- Hirschfeld and Grawitz. Birch-Hirschfeld claims that there is no good reason to separate on etiological grounds alone cases of severe anemia of cryptogenic origin from those like the anemia of bothriocephalus, in which an adequate cause is ap- parent. For both show the same anatomical evidence to prove that the pathological process in each is identical. He would, therefore, class together all severe anemic conditions, without regard to origin, in which there is de- generation and loss of red cells, increased destruction of tissue-albumeus, and fatty changes in the viscera, and in which the chief factors are excessive destruction and defective formation of red cells. Against this proposition Grawitz urges that it is clinically inapplicable, because fatty changes in viscera can be recognized only after death, that it classes together anemias arising from such diverse causes as cancer, malaria, sepsis, anchylostomiasis, etc., and that increased destruction of albumens is not a constant feature of the disease, although present in some forms of milder secondary anemia. Grawitz therefore proposes to throw out of the category of progressive pernicious anemia those cases in wliich the etiology is known, e. g., bothriocephalus, and to retain in this class only those cases in which the most thorough examination of the patient fails to show any primary disease. This suggestion accords with the view of those who fail to see in the his- tological changes in blood and marrow, evidence of a peculiar condition in- terpreted by Cohnheim, Ehrlich, and many others, as a reversion to the embryonal type of blood formation. 176 PROGRESSIVE PERNICIOUS ANEMIA. The writer, having failed to find from any source a controversion of the view that the changes in the blood and marrow described in per- nicious anemia result from a peculiar patliological process which is es- sentially different from that seen in the majority of secondary anemias, must accept the opinion that when the blood contains megaloblasts and a considerable proportion of megalocytes with increased Hb, while the lymphoid marrow shows marked hyperplasia of peculiar type, the con- dition should be called progressive pernicious anemia, without regard to its immediate exciting cause. The essential process in the disease is a reversion of the marrow to a type of blood formation which in some respects resembles the embryonal, and it would seem to make very little difference how various the exciting agents may be, whether sypliilis, malaria, anchylostomiasis, or intestinal auto-intoxication, pro- vided they all initiate the same process and stamp the anemia with certain self-perpetuating tendencies not seen in other conditions. Special Etiology. — Progressive pernicious anemia is, as shown by its history, a disease of very general distribution and comparatively frequent occurrence. In the Berlin City Hospital, Lazarus found 274 cases recorded in 10 years, or 2 percent of the admissions. It is con- siderably less common in New York hospitals, but consultants in the city see from 6-12 cases yearly. The female sex is considerably more subject to the disease than are males, owing probably to the predisposing influence of pregnancy. (MuUer, Ehrlich.) The 240 reported cases collected by Ehrlich were distributed be- tween the ages as follows : 1-10 1 case 10-20 22 cases 20-30 ()1 " 30-40 67 " 40-50 47 " 50-60 33 " 60-70 7 " 70-80 2 " Monti and Bergrun collected 16 cases in children, of which 5 occurred between the first and fifth years, and 9 between the fifth and fourteenth years. Ehrlich hesitated to accept these as genuine cases on account of the uncertain significance of changes in the blood in early life. Bothriocephalus Anemia. — The first important inroad made upon the group of idiopathic pernicious anemia was accomplished when Hoffman in 1885, followed by Botkin, Reyher, and Runeberg, showed that the presence in the intestine of many of these worms often leads to a severe or fatal anemia which may be made to disappear by the expulsion of the worms. While Reyher and Runeberg in 1886, and F. Muller in 1889, claimed that the morphology of the blood in these severe cases is identical Avith that of primary pernicious anemia, from the presence of megalocytes with increased Hb, megaloblasts, etc., the complete iden- tity of the condition existing in fatal cases with that of progressive per- SPECIAL ETIOLOGY. Ill nicious anemia was fully established by Schaumann's extensive studies in the same locality. It is of great interest that in some of Schaumann's cases, although the reduction of cells and Hb was not marked, yet Ehrlich's signs of progressive pernicious anemia, megalocytes with in- creased Hb and megaloblasts, were abundantly present. In Askanazy's cases" normoblasts were found in small numbers among many megalo- blasts, the former rapidly replacing the latter as the anemia improved. While the fact that the worms cause the anemia seems to be proved by the rapid recovery observed in some cases following their expulsion (Wiltschur, Askanazy), no satisfactory explanation of the pathogenic action of the parasites has been obtained. It appears that large num- bers of worms are required to produce the severe anemias, as many as 100 having been expelled from a patient of Boetticher's, although in other instances the usual small number was found. The duration of the infection appears to exert little influence, as is evident from the cases reviewed by Askanazy. (The writer secured 4 full-grown speci- mens from a healthy Swede who had been in America for 1 7 years. ) Schapiro and Dehio, finding some dead parasites in their cases, con- cluded that only the decomposing iconn produces the specific toxine which causes the anemia, but the majority of parasites found in severe cases are living. Wiltschur followed this clue further, finding that the majority of parasites were dead, and decomposing, or, as he concluded from alterations in the eggs, " sick." Some cases (Schaumann, Neu- bucher) in which anemia was present while no parasites but only eggs were found in the intestines at autopsy, Ehrlich believes can be ex- plained as resulting from the complete absorption of recent decayed parasites. This reasoning it is difficult to follow, nor can great im- portance be attached to the results of Schaumann and Tallquist who in the course of two weeks considerably reduced the red cells of a dog by injections of a glycerine extract of the crushed bodies of the worms. At present therefore it seems impossible to reach any satisfactory explanation why this parasite should behave at times as a harmless denizen of the intestinal tract, but at others should, apparently, destroy its host through a fatal anemia. The occurrence of fatal anemia referable to the presence of other in- testinal parasites, especially Anchylostoma duodenale, was demonstrated by Griesinger in 1851 and has been fully corroborated by Perroncito, Leichenstern, Sand with, etc. While it appears very probable that the same blood changes are established in these cases as by infection with bothriocephalus, there appear to be no detailed studies of the blood and marrow which can fully support such a view. Indeed, Zappert in a few cases of considerable severity found the ordinary characteristics of secondary anemia, and a low Hb-index, but did not apparently con- sider the morphological changes in the red cells. In Sandwith's re- markable experience, although the red cells were sometimes reduced below a million, the Hb-index was usually low, the clinical symptoms and pathological lesions were not distinctive of progressive anemia, while there is no report of morphological examination of the blood. 12 178 PROGRESSIVE PERNICIOUS ANEMIA. The report of Demme of two cases of fatal anemia in children in whose intestines large numbers of Asearides were found, is of interest but too isolated to be of great importance. Repeated Hemorrhages. — That genuine progressive pernicious anemia may follow repeated hemorrhages has been claimed by many observers who have not insisted upon Ehrlich's signs in their diagnoses. (Ha- bershon, Quincke, Scheperlen, Finney, Greenhow, Stockman.) In recent years reports of genuine cases of this origin have not been so numerous, yet Ehrlich and Lazarus accept as a fact their frequent oc- currence from this cause. The writer has not seen any cases directly referable to hemorrhage. It is probable that the genuine cases observed after pregnancy are partly the result of the losses of blood during and after parturition. Stockman goes so far as to urge that all cases now called idiopathic are the result of repeated minute intestinal hemorrhages, in which opinion he stands unsupported. While granting as does Ehrlich that the disease occasionally follows repeated hemorrhages, it still remains true that the great majority of chronic post-hemorrhagic anemias follow the type of secondary anemia with very low Hb-index, and w'ith leucocytosis. Pregnancy was one of the conditions first recognized as a prominent exciting cause of pernicious anemia, the most numerous cases (29) (35 percent) of those collected by Eichhorst (1878) coming under this heading. The genuineness of some of these cases may now be doubted, and at any rate the patients suffered from a variety of severe symptoms not at present regarded as belonging to the disease, so that Lazarus classes many of them as examples of secondary anemia and exhaustion. In others, however, the pregnancy was uneventful except for the spon- taneous development of severe anemia ending fatally before or after parturition, so that Lebert ascribed to pregnancy a special nervous in- fluence leading to fatal anemia, and Birch-Hirschfeld assumed that placental toxines affect first the plasma and later the red cells. These theoretical considerations have been rendered unnecessary by the grad- ual disappearance from literature of cases of progressive anemia fol- lowing pregnancy. Ehrlich could find only one case of this description in recent literature (Laache, Case 9). At Sloane Maternity Hos- pital from 1892—1899 the writer saw many cases of severe anemia, but none of the progressive pernicious type. Ahlfeld, in a wide experience, could not report a single case following pregnancy, (1898). The rather numerous current reports of severe anemia fol- lowing pregnancy refer to a severe secondary type, and, as Ehrlich concludes, it is impossible to ascribe to pregnancy any peculiar influ- ence in the origin of the disease. Syphilis. — That syphilis is occasionally followed by anemia in which the morphological changes in the blood are identical with those of pro- gressive pernicious anemia, seems to be fairly well attested by the re- ported cases of Laache, Kjerner, and F. Muller. SPECIAL ETIOLOGY. 179 In children suffering from hereditary syphilis the variations in the size of red cells, their increase of Hb, and the large numbers of megaloblasts, seem to leave little doubt that, in this age at least, syphilis figures as an exciting cause of progressive pernicious anemia. (Loos, Luzet.) Yet the reports in recent literature have not strength- ened the view that syphilis may lead directly to this peculiar type of anemia. The changes in the blood seen in the great majority of cases, •iilthough often severe, seem to follow the type of simple secondary anemia. Consequently Ehrlich and Lazarus doubt the existence of the disease as a result of syphilis. The writer has seen excess of deeply staining megalocytes, and megalo- blasts, in several cases of moderate severity shortly after florid syphilis, both in adults and infants, and has come to regard the differential diagnosis of severe secondary anemia in syphilitic subjects from progressive pernicious anemia, as one of the most difficult fields in blood-analysis. From experi- ence at the autopsy-table it has been very apparent, however, that most cases showing advanced or active tertiary lesions have shown types of secondary anemia. The subject requires much more extensive clinical and pathological observation. Malaria. — Although there are in the literature numerous cases of so- called pernicious anemia referred to malarial infection, the fact that the typical changes of the disease may be established as a result of chronic malaria was not demonstrated until Bignami and Dionisi re- ported their study of the marrow of fatal cases. Here were found the signs of megaloblastic hyperplasia in distinct form. Among the writer's Montauk cases there were no less than 19 in which the blood showed a majority of megalocytes with increased Hb, and a considerable number of megaloblasts, and in one of these the lymphoid marrow was found markedly hyperplastic, with excess of megaloblasts, and an enor- mous number of young estivo-autumnal parasites. In all these cases the condition had become established within 8-10 weeks from the be- ginning of the infections, but the malarial element had doubtless been aided by exposure and poor food. The frequency of these cases among malarial subjects and the fact that the bone marrow suffers in a peculiar way in many cases of estivo- autumnal infection, are considerations which bear directly on the myelogenous origin of pernicious anemia. Typhoid Pever. — Quincke and Rosenstein each report the transfor- mation of the cachexia of typhoid fever into pernicious anemia but in neither case do their reports show the presence of typical changes in the blood, both falling in the class of severe secondary anemia. Gastro-intestinal Disorders. — The largest group of cases and those which from the first have given most support to the theory of the cryptogenic origin of the disease have been referred to disturbance of the gastro-intestinal tract. A certain number of these cases have shown at autopsy well-marked lesions of stomach or intestines which have been deemed a sufficient cause of the anemia, while in others the stomach and intestines have been found normal and the theory of in- 180 PROGRESSIVE PERNICIOUS ANEMIA. testinal auto-intoxication has been elaborated, principally by Hunter, to account for the anemia. 1 . Lesions of Gastro-intestinal Tract. — Chronic gastritis or enteritis with fatty degeneration of the secreting cells of the peptic and intestinal glands, followed by atrophy and sclerosis of the mucosa have been demonstrated by Ponfick, Nothnagel, Fenwick, Levy, Brabazon, Nolen, Eisenlohr,^ Osier, Ewald, Marti us, Koch, and Stengel. Ewald found this condition of the gastric or intestinal mucosa in all his subjects examined at autopsy and believes that such changes, which he calls aiiadenia, are responsible for all " idiopathic " cases, but Im- merman and Quincke had previously shown that anadenia is not found in all cases of clearly protopathic origin, and this claim has been verified by recent observers. The writer has secured excellently pre- served peptic and intestinal glands and intact lymphoid structures from an acute case showing no visceral lesions to which the anemia could be referred, and Eisenlohr and Martins have reported extensive atrophy of intestinal mucosa in cases showing no gastro-intestinal dis- turbance and slight anemia. Stenosis of the pylorus as a result of annular carcinoma or of chronic gastritis, with extreme atrophy and contraction of the stomach, has been demonstrated in a considerable number of cases. (Israel, Ehrlich, Renvers.) In two of the writer's autopsies the stomach was markedly contracted, the pylorus was tightly steuosed, admitting a lead pencil with difficulty ; the mucosa showed the changes of well marked chronic gastritis, but the constriction of the pylorus was the result of chronic productive inflammation and not of carcinoma. These cases lasted two and four years, with inter- vals of improyement, but finally died with typical blood changes and with- out marked emaciation. Other cases with annular or circumscribed carcinoma of jyi/Iorus have been reported as showing typical changes in the blood. The writer has not yet seen any case of cancer of the stomach in which the lesions in the blood were indistinguishable from the classical type of progressive pernicious anemia. Lazarus reports genuine cases follow- ing small pyloric cancers only. A "neurotic atrophy'^ of intestinal mucosa has been suggested as an etiological factor by Banti, Jurgens (1882), Blaschko (1883), and Sasaki (1884), who found parenchymatous degeneration of ganglion cells in Meiss- ner's and Auerbach's plexuses, but Scheimpflug has shown that no such im- portant deductions were warranted from the changes demonstrable at that time in sympathetic ganglia, since identical changes occur in many other conditions. Peculiar eosinophile globules have been described by Lubarsch, Koch, and Sasaki, in the atrophic adenoid tissue of chronic gastritis, and these bodies have been regarded as pathognomonic of this condition and of significance in the anemia, but Hammarschlag has seen them in gastric cancer, and the writer, in the normal stomach at autopsy 1 hour after death from ether. The mesenteric lymph nodes have been found hyperplastic, in one in- stance caseous, by Eichhorst, Quincke, and others. This condition was present in some of the writer's autopsies, but not in all, while in a case of SPECIAL ETIOLOGY. 181 profound secondary anemia, probably of syphilitic origin, nearly all the abdominal lymph nodes were enlarged and caseous, but there were no megaloblastic changes in the blood. It is therefore obvious that all the gastro-intestinal lesions thus far de- scribed in cases of jirogressive j^ernicious anemia have been seen in other conditions and that none of them can be regarded as specific lesions in the disease. That many of the above lesions, however, especially the marked atrophic conditions, may act as occasional exciting causes of progressive pernicious anemia, as of secondary anemia, seems equally probable. Yet, not a few observers regard them all as secondary to the anemia, or as unimportant associated lesions, while Faber and Bloch believe that the atrophy of the intestinal mucosa described in many cases is always a cadaveric alteration. 2. Functional Disturbances of Gastro-intestinal Tract. — The prominence of the gastro-intestinal symptoms of the disease has from the first directed special attention to the study of functional dis- turbances in the alimentary tract. The reports of Sandoz were apparently the first to seriously strengthen the impression that some of the most typical cases of per- nicious anemia are of intestinal autotoxic origin, this observer finding that apparently genuine cases were sometimes cured by vigorous gas- tric lavage, enteroclysis, and the administration of intestinal antiseptics and laxatives. This observation has since been fully verified and the opinion has steadily grown that the most frequent if not the essential cause of progressive pernicious anemia is found in a peculiar toxemia of intestinal origin, with or without organic lesions of the mucosa. The evidence supporting this opinion has accumulated from many sides. The results of intestinal antiseptic treatment has steadily pointed in this direction. The observations on bothriocephalus anemia have re- cently given significant evidence to a similar effect (q. v.). Signs of increased intestinal putrefaction have been noted in the excessive indi- canuria of the disease (Senator, Muller, Brieger, Hennige, Grawitz, Schaumann), and in the presence of cadaverin and putrescin in the urine of certain cases (Hunter). In a very acute case examined at autopsy by the writer the odor of H.^S from the intestines was very intense, but in other cases there was very slight evidence of this putrefactive product. Hunter's studies in this field are undoubtedly the most important experimental contribution to the etiology of the disease that has yet been made. His conclusions are, briefly, that pernicious anemia is a specific clinical condition, resulting from excessive hematolysis occur- ring chiefly in the portal system and brought about by intestinal in- toxication in which the products of growth of specific bacteria are probably concerned. Reviewing the knowledge of the pathological changes in the blood and marrow, spleen, and lymph nodes, Hunter concludes that there is no known characteristic lesion of the disease, although the high Hb-index of the blood he regards as very important and an almost pathognomonic sign. Turning 182 PROGRESSIVE PERNICIOUS ANEMIA. to the liver he finds in the studies of Peters proof that in the anemia of wast- ing diseases there is no markedly excessive deposit of iron in the liver, and finds that in no disease bearing any clinical resemblance to pernicious anemia does the deposit of iron in the liver approach in any degree that characteristic of pernicious anemia. Moreover, this deposit is largely limited to the outer two-thirds of the lobules, a position in which the scanty deposits of other conditions (extravasation of blood, chronic congestion) are never limited. From the comparison of chemical analyses he finds that the liver in pernicious anemia contains on the average seven times as much iron as in any other diseases attended with anemia, while the iron content of the spleen is not notably increased. He therefore concludes that per- nicious anemia consists essentially in excessive hematolysis, occurring prin- cipally in the liver. Why Hunter limits the blood destruction to the portal system is not entirely clear, but he reasons as follows : Malaria, par- oxysmal hemoglobinemia, and poisoning by pyrogallic acid cause general hemoglobinemia and hemoglobinuria, with deposits in the liver unlike those of pernicious anemia, while in this disease and in poisoning by toluylendia- min there is never any Hb in the urine, but in each case the iron in the liver is found in the same position. Since toluylendiamin exerts a specific stimu lating action on the liver cells he believes the poison of pernicious anemia must be directed especially against the liver and hence be found principally in the portal sj'stem, a source which is further indicated by the frequency of gastro-intestiual symptoms. Hunter's studies are chiefly important in the demonstration of the pecu- liar behavior of the liver in ridding the blood of iron, but the theory of portal hematolysis by intestinal intoxication, while doubtless strongly sug- gested by his results, is still far from demonstration. He has recently attributed to bacteria from decaying teeth, and to chronic infectious gastro- enteritis from this source, an essential part in the etiology of the disease. Although extensive deposits of iron in the liver and the limitation of the pigment to the outer portions of the lobules have since been shown to occur in other diseases (Russell, Hindenlang), and cases of pernicious anemia have been reported in which no excess of iron was found in the liver (Ran- som), Hunter has demonstrated characteristic if not pathognomonic features in the pathological anatomy of the disease. Myelogenous Origin. — There are recorded a moderate number of ob- servations which suggest that progressive pernicious anemia may re- sult directly from lesions affecting the normal physiology of the bone marrow. Chief among these is the case of general sarcomatosis of bone marrow described by P. Grawitz^ as accompanied by typical changes in the blood. Litten (1877), reported a case of pernicious anemia followed by leukemia in which there were multiple abscesses in the bone marrow, but the diagnosis of pernicious anemia appears doubtful. Waldstein in 1882 saw a similar case in which the marrow was the seat of a peculiar change said to have resembled in structural details a chloroma found in the mediastinum. This case also is ob- scure. Grawitz believes that the severe anemia sometimes following in- fectious processes is to be referred to changes in the marrow instituted during the course of the infection. The frequent occurrence of per- nicious anemia after malaria appears to have similar import. Nervous Origin. — The prominent participation of the nervous system in the anatomical lesions of the disease has often raised a suspicion in many minds that some of these lesions might be primary. Except in HISTOLOGICAL CHANGES IN THE MARROW. 183 the case of the abdominal sympathetic none of the lesions have been long claimed to hold such a position in the pathology of the disease, and at present it is generally accepted that the central nervous sys- tem suffers only secondarily from the anemia or as the result of as- sociated conditions. Likewise the functional disturbances of the nervous system cannot be claimed to act as more than somewhat distant predisposing causes, although McKenzie and Curtin have referred to certain cases as having originated directly from nervous shock. Infectious Origin. — Pernicious anemia has not escaped classification among infectious diseases ; by Klebs, who claimed to have observed flagellate bodies in the fresh blood ; by Frankenhauser and Petrone, who reported the isolation of leptothrices in several cases ; by Bernheim, who isolated a bacillus from the blood of one case after death ; and by Perles, who observed highly refractive, very actively ameboid bodies in several cases. Histological Changes in the Marrow. — Cohnheim's original report on the changes in the marrow of a case of pernicious anemia described the presence throughout the shafts of the long bones of lymphoid mar- row which contained no fat cells but was composed of a small number of colorless myelocytes and of many red cells most of which were nucleated and of unusually large size. This observation was verified by many later investigators, especially by Rindfleisch, Ehrlich, Miiir, Muller,^ but its importance was greatly obscured by the reports by Neumann, Litten, Grohe, and others, that chronic diseases causing cachexia also lead to an hyperplasia of red marrow in the shafts of long bones. Neumann, and Litten and Orth, claimed also to have in- duced identical changes in the marrow by subjecting animals to repeated hemorrhages. In 1886 Geelmuyden claimed that the hyperplastic marrow in such chronic diseases does not differ in structure from that of normal lymphoid marrow except in regard to the Hb-content of the red cells, and that the hyperplasia here represents merely a simple physiological increase in functional capacity. He showed also that while the hyperplastic marrow in pernicious anemia is very extensive, occupying as much as 82 percent of the cavities of long bones, and is firm and cellular, much of the marrow of secondary anemia is apt to be gelatinous with comparatively few cells, while the hyperplasia seldom becomes half as extensive as in pernicious anemia. The writer, from a study of the marrow in cases of grave secondary anemia, showing in the blood many megalocytes, observed that the limits of the hyperplasia are seldom very extensive, that the nucleated red cells are usually relatively scarce and of smaller size instead of being over abundant and of very large size as in pernicious anemia. These observations merely corroborate in minor details the original claim of Cohnheim and Rindfleisch. Yet the essential importance of the changes in the mari'ow is again brought into question by the reports of cases of pernicious anemia in which the usual alterations in this tissue are wanting. 184: PROGRESSIVE PERNICIOUS ANEMIA. Six such cases, reported by Laache and Quincke, and one by Geelmuyden, date from too early a period to figure in the present discussion, but Ehrlich 2 in 1888 reported a case of fatal acute anemia in a "bleeder" who suffered principally from metrorrhagia and in whose blood the red cells were reduced to 215,000, the majority of them being undersized, while nucleated red cells were absent, and there was apparently no hyperplasia of the lymphoid mar- row. Engel has recently reported a very similar case. In 1893 the writer^ observed an apparently similar condition in a boy of 18 years, of Cuban parentage, who had always appeared pale, and whose sister was extremely pale and moderately anemic. The patient suffered from uncontrollable epistaxis, finally from intestinal hemorrhages, for three weeks before death. During the last week the red cells numbered 456,000, the leucocytes were reduced in number, there were no megaloblasts and no megalocytes, the majority of red cells being oval and undersized. No autopsy was secured, but the anemia appeared clearly to be of idiopathic origin. Ehrlich, Lazarus, and Engel drew the conclusion from such cases that pernicious anemia is not necessarily associated with hyperplasia of the lymphoid marrow. The writer has encountered other cases which seem to fall in this class, and for which the term "microcytic type of pernicious anemia" has been suggested, but he has always felt that the grounds for their classification with other acute cases showing megaloblasts and megalocytes are at present entirely inadequate. The reports of Ehrlich do not offer any evidence to prove the relation of these two types of anemia, and it seems entirely pre- mature to draw from such cases any conclusions regarding the pathological anatomy and pathogenesis of true pernicious anemia. The extent of the hyperplasia of lymphoid marrow may vary from the involvement of the entire shafts of all long bones to the appear- ance in smaller foci only of the megaloblastic changes. The latter condition is probably seen daring the transformation of a grave sec- ondary to a pernicious anemia. Hyperplastic lymphoid marrow may be found in one long bone while in others the marrow is normal. (Lazarus.) Histologically, the marrow in milder cases shows a considerable number of normal red cells, always less than in the circulation, a few normoblasts, and a greater number of megaloblasts sometimes lying in " islands," w^hile in a remarkable case of Rindfleisch's nearly all of the cells in the marrow were large megaloblasts. In most of the writer's cases the neutrophile myelocytes were distinctly deficient, the princi- pal colorless cells being large, hyaline, and mononuclear. The eosins are usually deficient. Evidences of active destruction of blood cells are seen in the presence of macrophages inclosing many red cells, and of many pigmented cells. The megaloblasts may reach an ex- treme size, 50 /u (Rindfleisch), their nuclei are usually eccentric, and in various stages of degeneration, while the cell body is polychromatic and very densely staining, or fragmented and pale. Not only the grade, but the type of lesion in the marrow appears to vary in different cases. The writer has encountered three rather dis- tinct conditions in the marrow of ribs, vertebra, and femur, of fatal cases of idiopathic pernicious anemia. 1. The Cohxheim-Rindfleisch type (megaloblastic degex- eration), in which with wide extension of lymphoid marrow an exces- HISTOLOGICAL CHANGES IN THE MARROW. 185 sive proportion of the cells in the marrow, and nearly all the nucleated red cells, are megaloblasts or gigantoblasts. 2. Extreme lymphoid hyperplasia (myeloblastic degexer- ation), in which the majority of cells are small, hyaline, and mononu- clear, while nucleated red cells are scarce, and those remaining are of large size. This type of lesion is identical with that recently described by Nagaeli, under the term " myeloblastic degeneration." According to Nagaeli, in pernicious anemia the hyperplastic marrow con- tains 95 percent of cells, a little larger than lymphocytes, non granular, moderately basophile, with reticulated nuclei. These cells he calls myelo- blasts and claims that they are the ancestral forms of myelocytes. He finds them in small numbers in the blood of pernicious anemia, and secondary anemia, and very abundantly in leukemia. (The writer has been unable to convince himself that these cells are not large and medium sized lympho- cytes.) 3. Recently the writer encountered a case of fatal pernicious anemia of twelve months' duration, without demonstrable origin, in which the marrow showed a remarkable excess of nucleated red cells of normal size. Nature of the Megaloblastic Changes in the Marrow. — If Cohnheim's belief is accepted that changes in the marrow in pernicious anemia constitute a reversion of that tissue to the embryonal type in the for- mation of red cells, pernicious anemia must be regarded as identical with or very closely related to a neoplasm involving the red cells. This view is accepted by Rindfleisch and Ehrlich, but the latter, how- ever, does not apparently press the somewhat strained relation to a tumor formation. The review of otir knowledge of the development of the blood, con- tained in Chapter IV., shows that the cellular processes concerned in the formation of red cells are virtually the same both in adult and in fetal life. In both periods the non-nucleated discs are the result of mitotic division of several series of erythroblasts, each new series con- taining more and more Hb, until by a process not yet fully understood the non-nucleated disc springs from the last series, i. e., the normo- blast. In both adult and embryonal lymphoid marrow the earlier series of nucleated Hb-holding cells appear to be nearly equally repre- sented, while at no period of embryonal life are very large mitotic polychromatophilic cells seen such as abound in the marrow of acute pernicious anemia, and under no other conditions does the marrow con- tain such an excess of the large megaloblasts as in this disease. The essential difference between the normal formation of red cells in the adult and their defective development in pernicious anemia consists in the failure of the later more highly specialized series of erythroblasts, I. e., the normoblast. This of course is the fundamental nature of a tumor formation, in which the highly specialized properly functionat- ing tissues are replaced by excessively cellular, poorly differentiated structures, subject to a great variety of degenerative changes. Although the gigantoblast of pernicious anemia is never seen in em- 186 PROGRESSIVE PERNICIOUS ANEMIA. bryonal blood-forming organs, such aberrant forms seem to result from secondary degenerative processes, while it is the absence of normoblasts which is the more important anomaly of blood formation in the disease. The investigations of the thirty years that have intervened since Cohn- heim concluded (1868) that pernicious anemia is a reversion to the em- bryonal type of blood formation seem to have in a considerable degree verified his belief. To what extent pernicious anemia follows the laws known to govern the occurrence and behavior of tumors is an impor- tant inquiry which cannot here be further considered. Pathogenesis of the Disease. — Progressive pernicious anemia is believed by some to result through defective hematogenesis, and by others primarily from excessive hematolysis, to which is added secondarily defective hemato- genesis. The latter view is based principally upon the abundant evidence of exces- sive hematolysis which is undoubtedly the prominent external feature of the disease, upon the reported occurrence of cases of pernicious anemia which were rapidly fatal without yielding evidence of defective hematogenesis in the megaloblastic changes in the marrow, and upon experimental evidence that the prolonged action of certain blood solvents, toluylendiamin, etc., may lead to a condition very similar to progressive pernicious anemia. To some extent this opinion is favored also by the results of repeated losses of blood which have been shown to lead to a condition closely resembling progressive pernicious anemia both in the blood and in the marrow. (Neumann.) More- over in many cases of grave secoudary anemia the blood often approaches in character that of the cryptogenic anemia, and this fact has been accepted as favoring the hematolytic theory, although in cancer of the stomach, which leads to some of tlie severest grades of secondary anemia, defective blood formation seems quite as likely to figure in the anemia as does excessive hematolysis. The acceptance of this view involves the admission that the changes in the marrow are secondary to the destruction of blood and are therefore not pathognomonic of the disease. The writer has always felt that the correct conception of the disease ac- cording to present knowledge is as a process of defective hematogenesis in which megaloblastic degeneration of the marrow is the pathognomonic tissue lesion and excessive hematolysis a constant result. Hematolysis undoubtedly precedes in many instances defective hemato- genesis, but it is a different kind of hematolysis from that which occurs after the megaloblastic changes have become established. The peculiar de- struction of blood cells seen in pernicious anemia, marked by excessive de- posits of iron, by hemoglobinuria, pathological urobilinuria, by phagocytic englobement of red cells in the marrow, and by peculiar clinical symptoms, appears to be possible only when red cells are defectively formed. Most convincing demonstration of the propriety of this view was encoun- tered on comparing a case of profound secondary anemia from chronic ulcer of the duodenum with a rapidly fatal case of the cryptogenic disease. In the former case, although the Hb registered only 10 percent and the red cells were below one million, there were only slight changes in the size of the cells and the patient improved under iron. In the latter case, the Hb reg- istered 38 percent on first examination, the red cells 1,780,000, but megalo- cytes and megaloblasts were abundant and the patient died after an acute illness of four weeks. The essential difference between these cases appears clearly to have been not in the destruction or loss of blood but in its forma- tion. It has been amply shown that the system can replace enormous losses if conditions are favorable for the normal production of blood. Lazarus quotes the example of a Russian physician whose health remained little affected by CHANGES IN THE BLOOD. 187 the loss of about four times the entire volume of blood by consecutive pul- monary hemorrhages in the course of a few months. This author also refers to cases of cancer of the stomach in which the anemia remained compara- tively harmless until suddenly changing from the secondary type witli low Hb index to the primary with increased Hb-iudex and appearance of megalo- blasts. {Die Anaemie, p. 43.) The attempted experimental production of pernicious anemia by the use of blood solvents, etc., may or may not have been partly successful, as indi- cated bj' Neumann's studies, but their outcome can hardly alter the fact that blood destruction may be continued over long periods without giving the typical picture of pernicious anemia which is at once evident in clinical ob- servation when defective hematogenesis is added to the long-existent hema- tolysis. In view of these and many similar considerations it would seem that the belief is justified that it is the megaloblastic degeneration of the lymphoid marrow which constitutes pernicious anemia a separate clinical and path- ological entity. Hematolysis may, and in some instances does, precede, in point of time, the changes in the marrow, as in the pernicious anemia fol- lowing malarial cachexia, but this destructive influence at once takes on a new character and produces specific results when acting on defectively formed cells, as is shown by the comparison of the iron content of the liver in sec- ondary and in progressive pernicious anemia. The occurrence of rapidly fatal cases of the disease and the transformation of a grave secondary anemia (cancer of pylorus) into true pernicious anemia may be explained by a rapid and general or slow and partial establishment of the marrow changes. In the former case the changes in the blood will be pronounced from the first, in the latter it may be difficult to determine the exact point when a grave secondary becomes a true pernicious anemia. Changes in the Blood. Chemistry. — The specific gravity of the blood is constantly reduced, owing to the loss of both the albumens of the plasma and the Hb of the red cells. The gravity is here a much less reliable indication of the Hb-content than in chlorosis, in which the Hb alone is markedly af- fected. (Hammarschlag, Dieballa.) Nevertheless, it is just in this dis- ease that the information furnished by the specific gravity is of peculiar interest, as it concerns those principles in the plasma of which little can be learned by other, simple, clinical tests. Grawitz finds that the specific gravity of the blood may fall below 1.030, or below that of normal serum. Dieballa had two cases show- ing a gravity of 1.028, with red cells 404,000 and 500,000, while Copeman, using Roy's method, reported a case with a gravity of 1.027. That there is no close relationship between the numbers of red cells and the specific gravity appears in Dieballa's tables in which a gravity of 1.038 was obtained in one case showing 784,000 cells and in an- other with 1.4 million cells. The total albumens of the blood are much reduced, but the loss affects principally the red cells. Grawitz finds the loss of albumens of the serum to be distinctly greater in secondary anemia than in progressive pernicious anemia. The great reduction in total albumen of the blood is indicated by the low percentage of dry residue found by v. Jaksch, Grawitz, and Stintzing and Gumprecht, who in very severe cases obtained between 188 PROGRESSIVE PERNICIOUS ANEMIA. 9 percent and 12 percent (normal 20-24 percent), while the red cells numbered between 534,000 and 1.7 million. That the loss of albumen is due principally to the changes in the red cells appears from the fact that the specific gravity and dry residue of the serum are but slightly altered even in severe cases. Thus Grawitz found a gravity of the serum of 1.024 (normal 1.027- 1.029), dry residue 7.08 percent (normal 10-10.5), while the gravity of the blood fell to 1.036, and its dry residue to 11.45 percent (nor- mal 21-24 percent). While the albumens of the blood were reduced therefore one half, those of the serum fell only one fifth. Similar results were obtained by Dieballa. This loss of albumens of the whole blood must be referred to the loss of fibrinogen and to the diminution of red cells. V. Jaksch's analyses ' of red cells in pernicious anemia led him to conclude that those remaining are abnormally rich in albumens, a view which may seem to accord with the large size of these cells. By estimating the N of 100 gr. of centrifuged red cells he obtained 6.48 grm. N, 40.5 grm. alb. (normal 34.5). Grawitz's analyses however do not support this conclusion. The relatively high proportion of albumen in the serum in perni- cious anemia distinguishes this disease, as Grawitz has shown, from many forms of secondary anemia arising from hemorrhage, malignant tumors, etc., with which the albumens of the serum are much reduced. The resistance of the red cells has not apparently been tested, although Limbeck (p. 163) found their isotonic tension somewhat in- creased (resistance diminished) in two cases of grave anemia. Copeman noted that the hemoglobin crystallized readily from the blood of a case of pernicious anemia, while it is very difficult to obtain such crystals from normal human blood. Morphological Changes. — The whole blood is much reduced in quantity as is shown by the scanty content of vessels and viscera at autopsy, and by the imperfect filling of the superficial vessels during life. This reduction in bulk is often so marked as not to require for its demonstration any attempts at accurate measurement. Coagulation is very feeble so that on standing it was early noted (Ponfick) that the red cells settle to the bottom, the leucocytes form a middle layer of variable depth, while the plasma deposits nearly all its cells without interference from clotting. To the feeble clotting is probably referable the very slow separation of serum from the clot, as observed by Lenoble. The color of the blood may in mild cases be almost normal ; it is usually very pale ; while in extreme cases visible currents of colorless serum mingle with the drop on very slight pressure. On this account a liberal puncture may be required to obtain a drop suitable for exam- ination. A v^ery dark color has been observed by Furbringer and, after transfusion, by Gusserow. Hemoglobin. — The percentage of Hb varies from a moderate reduc- tion (70 percent) to such a low point that accurate estimates are im- possible, while the Fleischl instrument indicates less than 10 percent. PLATE V. Progressive Pernicious Anemia. (Eosin and Methylene Blue.) Fig. I. Figs . 2. Fig. 3- Fig. 4- Figs •5- Fig. 6. Fig. 7- Fig. 8. Fig. 9- Fig. lO. Megalocyte with excess of Hb. Microycte with excess of Hb. Microcyte deficient in Hb. Poikilocyte deficient in Hb. Poikilocyte witli excess of Hb. Polychromasia of Maragliano. Megaloblast, resting vesicular nucleus, basophilic granules in cytoplasm. Microblast. Megaloblast in mitosis. Thread stage. Megaloblast. Pathological mitosis. CHANGES IN THE BLOOD. 189 In the average case the Hb registers between 20-40 percent. A rather high percentage of Hb with severe clinical symptoms and pro- nounced morphological changes in the red cells is not infrequently seen. Lazarus refers to a case showing 65-70 percent but otherwise with pronounced symptoms. The Hb is usually not so low as in secondary anemias of equal severity. A RELATIVELY HIGH Hb-index is One of the characteristic features of the blood. In many cases, especially when the megalocytes are very abundant, the Hb-index is above normal (1-1.75) and in such cases the index is apt to rise as the anemia increases, and to fall dur- ing intervals of improvement. In a larger group of cases the index does not vary greatly from the normal, when the excess of Hb in megalocytes balances the loss in other cells. In other cases the index is for long periods below normal, but still relatively high as compared with secondary anemia and chlorosis. Thus the Hb-index in chlorosis averages about .50 (Cabot, Bramwell), but in pernicious anemia the index seldom approaches that figure. The low Hb-index is usually seen in chronic cases, while the highest indices obtained by the writer were in more rapid cases especially just before death. In some cases of secondary anemia the Hb-index may approach the normal, so that comparatively little diagnostic importance can attach to the lesser grades of increase in Hb. Red Cells. — In the average case of established pernicious anemia the red cells vary slightly above or below one mill ion. Well-marked mor- phological changes may exist, however, when the red cells number over two millions, but earlier stages of the disease have so far escaped notice, indicating that preexisting anemia is an essential predisposing condition. Although Quincke reported 143,000 red cells in a case which recov- ered and Hayem 292,500 in a fatal case, more recent reliable statistics show that life is seldom prolonged when the cells fall below 400,000. Even this extremely low proportion is below the count in many fatal cases, as many patients succumb when the red cells are about one million. Most fatal cases show at death between 300,000 and 600,- 000 red cells. The reduction in cells, in cases under treatment, seldom progresses uniformly, intervals of improvement, followed by rapid re- lapses, continuing till the end. Changes in the form, size, and staining reaction of the cells are even more important than their reduction in number. While the largest cells of normal blood measure about 9 ft across, in pernicious anemia the increased size of the cells is characteristic of the disease, the averas^e diameter being- estimated bv Lazarus at 11-13 fj., while some may reach a diameter of 16—18 /Jt. (Askanazy.) Cases vary greatly in regard to the extent and uniformity of this lesion. In one of the writer's cases, not of unusual gravity, at least 90 percent of the cells measured from 11-16 /jl. Megaloblasts were very numer- ous, and microcytes extremely rare. It may be said that wdess Jo per- cent of the cells are distinctly oversized the diagnosis of pernicious anemia 190 PROGRESSIVE PERNICIOUS ANEMIA. should be made with reserve. Usually, but not always, the proportion of megalocytes accords with the severity of the condition. Lazarus found in 8 severe cases 56—71 percent of megalocytes, in 5 cases dur- ing periods of improvement 33-50 percent, or during complete remis- sions 0-14 percent. The Hb-content of these megalocytes is in typical cases distinctly increased and the cell is flat and uniformly opaque instead of showing the bi-concave discoid shape. The writer has found this character of the megalocytes a very reliable differential sign between true perni- cious anemia and the grave secondary anemia of cancer, etc. In the latter conditions the great majority of megalocytes are usually deficient in Hb. (Cf. Plates V. and VI.) MiCROCYTES are a somewhat less characteristic cell-form in perni- cious anemia. They may be only slightly under normal size or be found in severe cases as mere points of reddish staining protoplasm not over 1 tx in diameter. They may contain an excess or deficiency of Hb. In the early stages of the disease the writer has found them less frequent, but in moribund cases they rarely fail to appear in con- siderable numbers. The division of forms between megalocytes and microcytes is then often complete, no normal cells remaining. Deformed red cells (poikilocytes) of all sizes are commonly present in considerable numbers, probably resulting from the ameboid properties of the cells and frequently from traumatism. The oval shape of many cells both large and small was early noted by Osier and Gard- ner, and in some cases this character is very prominent, for reasons which are not clear. Pear-shaped cells are perhaps the most common form, the narrow point representing a pseudopodium. Cells very defi- cient in Hb are apt to bend into horseshoe or pessary shapes. Cells rich in Hb less often become deformed. Ameboid motion of red cells in pernicious anemia may be observed in fresh preparations. The moving cells usually content themselves with the extrusion of pseudopodia, and with changes in shape which Muir and Gulland believe are not true ameboid phenomena, but the writer has seen active locomotion of lenticular microcytes. Nucleated red cells may be said to occur in all cases of perni- cious anemia, but their demonstration sometimes requires prolonged search. It has been shown by Ehrlich and verified by nearly all later observers, that megaIohlast% constitute the majority of such cells in per- nicious anemia and there is at present no ground on which to deny their pathognomonic significance, when present in large numbers. In typical cases of the disease they are almost or quite exclusively present, while normoblasts are commonly absent. Their numbers seem to de- pend largely on disturbances of the circulation and other accidents, but Ehrlich claims that the extent of the megaloblastic degeneration of the marrow may be determined by the numbers of these cells in the blood. This rule probably holds in general, but in the writer's experience there have been exceptions. In other cases both normoblasts and megaloblasts appear in the blood, in which case it is found that the PLATE VI Secondary Pernicious Anaeniia. (Eosin and Methylene Blue. Figs. I. Megalocytes. Fig. 2. Red cells of normal size, deficient in Hb. Fig. 3. Microcyte deficient in Hb. Fig. 4. Red cell showing granular degeneration. (Punctate basophilia.) Fig. 5. Megalocyte with excess of Hb. (Polychromasia of Gabritschewsky.) Fig. 6. Polynuclear leucocyte. Xodal points of cytoreticiihini. k CHANGES IN THE BLOOD. 191 megaloblasts outnumber the normoblasts, indicating that the defective formation of cells has become uppermost. From the considerations mentioned under the etiology and patho- genesis of the disease it will be seen that the crucial point in the diag- nosis between secondary and pernicious anemia is encountered in prac- tice, in the proportions of normoblasts and megaloblasts to be found in the blood. If, as appears to be fully attested, a grave secondary may be transformed into pernicious anemia, the change must consist pri- marily in a megaloblastic degeneration in some portion of the marrow. The normoblasts of the circulation must then be gradually replaced by megaloblasts. These theoretical deductions appear to be borne out in clinical experience since the appearance of megaloblasts accompanies a more severe type of anemia. Lazarus finds normoblasts rarely associated with megaloblasts, and this has been the writer's experience, except in children, in whom both are frequently found together. Schaumann saw both forms in all his cases of bothriocephalus. Cabot noted that megaloblasts and normoblasts are fre- quently associated in pernicious anemia and that the former increase when the disease grows worse. Askanazy found many megaloblasts in a case of bothriocephalus, while after expulsion of the worms they were steadily re- placed by normoblasts. Dorn observed an idiopathic case promptly improve after the sudden appearance of many normoblasts. From Coles' extensive observations in two cases it is seen that there may be very marked varia- tions in the numbers and to a less extent in the proportions of normoblasts and megaloblasts in the blood, and that in the early stages of the disease normoblasts may predominate. On both clinical and pathological grounds therefore it is justified to class an anemia as pernicious when the megaloblasts out- number the normoblasts, or when any considerable proportion {33 percent) of the red cells are megalocytes with increased Hb. A question frequently arises regarding the value of a prolonged search for megaloblasts in order to establish the diagnosis of pernicious anemia, since if only one nucleated cell be found and that a megalo- blast the chances favor the presence of an excess of such cells over normoblasts. It is quite possible, however, that the first few nucleated cells should be of large size while the next three or four prove to be normoblasts, and it appears much more rational to base the diagnosis upon the general morphology of the blood and on all the clinical data, than to rely upon any single isolated sign. Nevertheless the discovery of a single very large megaloblast with abnormal nucleus is virtually a pathognomonic sign of pernicious anemia. The numbers of nucleated red cells in pernicious anemia vary greatly. Frequently they are scarce and their demonstration requires patient searching. Or, of two slides taken at the same time one may contain several and the other apparently none. Usually they increase with the severity of the blood changes. The occurrence of normoblastic blood-crises has been referred to (Dorn) as a favorable sign. Coles found over 6,000 nucleated red cells per cmm., the majority of large size, in a fatal case. The writer has found them unusually abundant in a case in which nearly all the red cells were very large, and has seen 192 PROGRESSIVE PERNICIOUS ANEMIA. them nearly as numerous as the leucocytes during ante-mortem leuco- cytosis. Distinct microblasts are comparatively rare. The occurrence of mitotic nuclei in megaloblasts, first described by Luzet, is usually limited to the severer stages of the disease, in which their presence is pathognomonic of the condition. Schaumann, how- ever, found them in a case that recovered. The presence of three or more unequal asters in one gigantoblast, as occasionally found in se- vere cases is one of the most significant pathological signs within the range of blood analysis. (Plate V., Fig. 10.) Degenerative changes in the red cells not only cause the appearance of abnormal cell forms but lead to a series of chemical and morphological alterations plainly demonstrable in stained specimens. Poli/chromatophilia, referring to the development of a brownish stain- ing quality of the Hb is constantly present, usually more marked in the severer cases. It appears probable that the change is in some cells preliminary to the solution of Hb in the plasma, but according to Engel it indicates very rapid formation of cells and incomplete meta- morphosis of nucleus. It is most marked and nearly constantly present in megaloblasts, some of which stain diffusely brownish blue. Uemoglobinemia is constantly present in the severer cases and is in- dicated by a distinct brownish red staining quality of the dried plasma. Punctate BASOPHILIA (granular degeneration (Grawitz^)) of red cells is of moderately frequent occurrence both in megalocytes and es- pecially in megaloblasts. As previously shown, this abnormality is sometimes referable in the megaloblasts of pernicious anemia to karyo- rhexis of degenerating nuclei. In other cells the blue granules are probably reappearing particles of the nuclear remnant of ordinary red cells. (Maximow.) ScHiSTOCYTOSis, or the separation of fragments of red cells, is, ac- cording to Ehrlich, the mode of origin of many microcytes in per- nicious anemia. Hydremia of red cells probably causes many smaller cells to appear much larger than normal. It has already been shown that hydremia of the plasma is probably responsible for the increased diameter of many red cells, but in pernicious anemia the majority of megalocytes, especially those with increased Hb, are undoubtedly derived from megaloblasts. Absence of rouleaux is a constant and very characteristic feature of the disease, and results not only from the reduction in the number of cells but principally from chemical changes in their membranes and protoplasm. The Leucocytes. — In well established cases the leucocytes are markedly reduced in number. This condition is clearly referable to the megaloblastic lesion in the marrow, which leads to an excessive pro- liferation of megaloblasts at the expense of other marrow cells. (Rind- fleisch.) Progressive hypoleucocytosis is the rule, the white cells fall- ing very low in extreme cases. (1,500-2,000, Hayem.) Of the colorless cells which remain, a diminished proportion are neutrophile. RESUME OF THE CHIEF FACTS. 193 the majority being large and small basophilic cells. With the reduc- tion in total numbers there is a relative lymphocytosis, very slight in the mild cases, and very marked (79 percent, Cabot) in severe stages. The eoslnophUe cells usually vary within low normal limits. In the severer cases they are apt to be reduced (Grawitz, p. 102), and rarely may be absent (Cabot), but, contrary to Neusser's belief, they are very scarce in mild and present in some fatal cases. In the marrow, the writer has found them usually much reduced in number. Myelocytes are usually present in small numbers, Cabot finding them in 42 of 52 cases. The writer finds these cells in pernicious anemia almost invari- ably of the smaller type, but has seen isolated specimens of the larger (Cornil's) neutrophile cell, which is abundant in leukemia. Mast-cells were increased in some of the Avriter's marrow smears, but he has never seen more than single examples in the blood smear. Hyperleucocytosis in pernicious anemia occurs as the result of com- plications, as an ante-mortem phenomenon, rarely without discoverable cause in course of the disease, and in one case (Dorn) it appeared to be associated with a normoblastic blood crisis, and heralded a favor- able turn in the disease. RESUME OF THE CHIEF FACTS CONCERNING PROGRESSIVE PERNICIOUS ANEMIA. Definition. — The terra should be applied to all forms of grave anemia showing an excess of megaloblasts over normoblasts, or at least 33 percent of megalocytes (or less during remissions). Etiology, — The largest group of cases remains of cryptogenic origin but is possibly to be referred to intestinal intoxication. Various lesions of the gastro-intestinal tract are underlying causes, including chronic gastritis with or without stenosis of pylorus, and small carcinomata of pylorus. Of intestinal parasites, Bothriocephalus latus certainly, and other parasites possibly, lead to the disease. Repeated hemorrhages in rare instances cause typical pernicious anemia. Pregnancy has no specific influence, but the cases which fol- low parturition are probably referable to hemorrhage and various un- hy genie conditions. The disease is referable to pernicious malaria often ; to syphilis, j^ro- longed diarrhea, bad hygienic conditions, probably ; to typhoid fever, yellow fever, and other infectious diseases, doubtfully ; but not to func- tional or organic disturbances of the nervous system. Lymphosarcoma of the marrow was present in one reported case. Pathological Anatomy. — The essential lesion of the disease is a megaloblastic metaplasia of the lymphoid marrow which is invariably present, is pathognomonic of the disease, and in many respects re- sembles a tumor formation aifecting the progenitors of the red cells. Pathogenesis. — The disease does not exist until defectively formed red cells are being destroyed in large numbers by some hemolytic agent. 13 194 PROGRESSIVE PERNICIOUS ANE3IIA. Changes in the Blood. — The red cells may be reduced to about 200,000, but usually number about one million. Megalocytes with increased Hb constitute from 33 percent to 90 percent of the cells, but during remissions they may be very scanty. Megaloblasts are very scarce, or very numerous, and are either ex- clusively present, or outnumber the normoblasts. Microblasts are rare. The Hb-index varies, but in the advancing stages it is usually above normal, or during periods of improvement or remission may be below normal. If below normal, it is usually higher than in any other form of anemia. In the fresh blood coagulation is very feeble, rouleaux do not form, there are doubtful evidences of ameboid motion of poikilocytes, but occasionally active locomotion of microcytes. Hemoglobinemia often exists, but is sometimes, possibly, an artifact. Degenerative changes in red cells are extensive, including poikilo- cytosis, schistocytosis, diminished resistance and solution of red cells, polychromasia (Gabritschewsky), punctate basophilia and karyorhexis in megalocytes and megaloblasts. Leucocytes are usually reduced, in proportion to the severity of the lesion in the marrow and the progress of the disease, and relative lym- phocytosis is the rule. Leucocytosis arises usually from complications, rarely with "blood crises," and often just before death. Eosins are commonly deficient and a few medium-sized myelocytes can usually be found. Chemistry. — The blood is much reduced in bulk. The specific gravity has no constant relation to the number of red cells or propor- tion of Hb, but follows the loss of albumens, of cells, and of plasma. It may fall as low as 1.027. The albumens and specific gravity of the serum are much less reduced than in grave secondary anemias. The resistance of the red cells is reduced and Hb may crystallize read- ily. The hydremia of the plasma affects the red cells and is probably lesponsible for some of their increase in size. The Diagnosis may rest upon the presence of: — numerous megalo- blasts and megalocytes with increased Hb ; 33 percent of megalocytes ; with increased Hb ; an excess of megaloblasts over normoblasts ; a single gigantoblast or megaloblast in pathological mitosis. The diagnosis cannot rest on an extreme reduction of red cells. | The diagnosis may require the complete summation of all clinical 1 and morphological data, as well as observation on the course of the disease, or even the microscopical examination of the marrow. s The cases of fatal idiopathic anemia of microcytic type (Ehrlich, | Engel, et al.) cannot at present be classed with the Addison-Biermer progressive pernicious anemia. The prognosis is very difficult to determine. The most severe case on record recovered. (Quincke's with | 143,000 (?) red cells.) Severe cases of cryptogenic origin may recover. The most unfavorable are the very acute idiopathic cases, and those of slower development from gastro-intestinal lesions. BIBLIOGRAPHY. 195 The sudden development of well-marked changes in the blood com- monly indicates a severe course but such cases have recovered. The slow relapsing cases, with gradual changes in the blood and absence of raegaloblasts usually die from their anemia. A high Hb-index goes with well-marked changes in the blood and marrow and is of unfavorable import. Lower Hb-indices are seen in chronic cases and in remissions of more acute cases. A reduction to 600,000-700,000 red cells in this disease is seldom survived. Marked degenerative changes in megalocytes and megalo- blasts, and marked hemoglobinemia go with the severe cases. Extreme reduction of neutrophile and eosinophile leucocytes is an unfavorable sign ; while leucocytosis commonly precedes death by a few days. Bibliography. Pernicious Anemia. Addison. Constitutional and local effects of disease of Supra-renal glands, London, 1855. Ahlfeld. Lehrb. d. Geburtshulfe, Leipsic, 1898. Askanazy. ^Zeit. f. klin. Med., Bd. 23, p. 80. 2 ibid., Bd. 27, p. 492. Banti. Jaliresb. u. d. gesam. Med., 1881, II., p. 239. Bernheim. Cited by Hayem, Rev. med. de I'Est., 1879, p. 687. Biermer. Tageblatt 42d Versaml. Deut. Naturfors. u. Aerzte, Dresden, 1868. Cor- respb. f. Scliweizer Aerzte, 1872. Bignami. Cent. f. Path., 1899. Blaschko. Virchow's Archiv, 1883, p. 94. Boetiicher. Cited bv Askanazy. ' Birch-Hirschfeld. XI. Cong. "inn. Med., 1892. Botkin. Cited by Schaumann. Brabazon. Brit. Med. Jour., 1878, II., p. 134. Bradbwy. Brit. Med. Jour., 1876, II., p. 244. Brieger. Ueber Ptomaine, Berlin, 1885. Cohnheim. Virchow's Archiv, Bd. 68. Coles. Brit. Med. Jour., 1900, I., p. 759. Copenum. Brit. Med. Jour., 1891,1., p. 161. Curtin. Philada. Med. Times, 1885. Behio. St. Petersburg med. Woch., 1891, No. 1. Demrne. Cited bv Grawitz. , Bieballa. Deut. Archiv klin. Med., Bd. 57, p. 302. Born. Blutuntersuch. bei pernic. Anemic, Diss. Berlin, 1891. Ehrikh. 1 Gesam. Mittheil., 1880. 2 charite Annalen, 1888, Bd. 13. Ehrlich, Lazarus. Die Anemie, 1900, II. Bichhnrst. Die prog, pernic. Anemie, 1878. Eisenlohr. »Deut. Archiv klin. Med., 1877, Bd. 20. "Deut. med. Woch., 1892, No. 49. Engel. Zeit. f. klin. Med., 1900, Bd. 40, p. 17. Ewald. Berl. klin. Woch., 1895, No. 45, 1896, No. 10. Evdng. 1 Jour, of Exper. Med., 1901. ^ Trans. N. Y. Path. Sec, 1896, p. 59. Faber, Block. Cent. f. inn. Med., 1900, p. 404. Fenwick. Lancet, 1877, II., p. 1. Furbringer. XL Cong. f. inn. Med., 1892. Ferrand. Bull. Soc. d. Hop., 1876, p. 349. Finney. Brit. Med. Jour., 1880, I., p. 43. Frankel. Charite-Annalen, 1876, III., p. 283. Frankenhauser. Cent. f. med. Wissen., 1883. Grohe. Berl. klin. Woch., 1881-1884. Virchow's Archiv, Bd. 22, p. 437. Geelmuyden. Virchow's Archiv, Bd. 105. Graudtz. 2 Virchow's Archiv, Bd. 76, p. 353. " Berl. klin. Woch., 1900. Greenhow. Brit. Med. Jour., 1871, II., p. 613. Griesinger. Archiv f. physiol. Heilk., 1854, p. 554. 196 PROGRESSIVE PERNICIOUS ANEMIA. Gulland. Cited by Bramwell. Habershon. Med. Times und Gaz., 1876. Hayem. ^ Cited by Ferrand. Hennige. Cited by v. Noorden. Path. d. Stoffwechsels, p. 347. Hem-y, Osier. Amer. Jour. Med. Sci., 1886, Vol. 91, p. 498. Hmdenlang. Yirchow's Archiv, Bd. 79, p. 492. Hoffman. Vorles. u. allg. Therapie, 1886, p. 14. Hunter. Lancet, 1888, 1889, 1900. Immerman. Ziemsen's Cyclop., Vol. 16, p. 282. Israel. Berl. klin. Woch., 1890, p. 649. V. Jaksch. ' Zeit, f. klin. Med., Bd. 24. Ubid., Bd. 23. Jurgens. Berl. klin. Woch., 1882, No. 23. Kjerner. Schmidt's Jahrb., Bd. 186, p. 26. Klebs. XI. Cong. inn. Med., 1892. Koch. Inaiig. Diss., Berlin, 1898. Ub. Verand. d. Magens ii. Darm. bei pernio. Anemie, Berlin, 1898. Laache. Die Anemie, 1883. Lebert. Wien. med. Woch., 1858, No. 34. Leichienstern. Deut. med. Woch., 1887-1888. Lenoble. Charac. semeiologiques du caillot et du serum, Paris, 1898. Lepine. 'Rev. Mensuelle, 1877. ^L' Union Med., 1876, T. 22, pp. 453, 474. Litten. Berl. klin. Woch. , 1877, 257. Litten, Orth. Berl. klin. Woch., 1877, No. 51. Loos. Wien. klin. Woch., 1892, p. 291. Lowy. Ziegler's Beitriige, 1886. Luzet. Etude sur les anemies de I'Enfance, 1891. Martins, Lubarsch. Achylia gastrica, Leipsic, 1897. Mackenzie. Lancet, 1878, II., p. 834. Monti, Berqgnm. Die Chron. Anemie d. Kindesalter, 1892. Muir. Joiir. of Path, and Bact., 1894. Midler. Die prog, pernic. Anemie, Zurich, 1877. F. Mueller. Charite-Annalen, 1889, p. 253. ^ Deut. Archiv klin. Med., Bd. 51. Musser. Med. News, Vol. 41, p. 404. Nagaeli. Deut. med. Woch., 1900, p. 287. Neubecker. Inaug. Diss. Konigsberg, 1898. Neumann. Berl. klin. Woch., 1877, No. 47. Nolen. Cited, Cent. f. med. Wissen., 1882, p. 767. Nothnagel. Deut. Archiv klin. Med., 1877, Bd. 24. V. Noorden. Charite-Annalen, Bd. 16, p. 217. Osier, Gardner. Cent. f. med. Wissen., 1878. Pepper. Amer. Jour. Med. Sci., 1875, Vol. 140. Perles. Berl. klin. Woch., 1893, No. 40. Perroncito. Compt. Rend. Acad. Sci., 1882, No. 1. Peirone. Lo Sperimentale, 1884, T. 53. Ponfick. Berl. klin. Woch., 1873, Nos. 1-2. Quincke. Volkman's Vortriige, 1876, No. 100, p. 22. Deut. Archiv klin. Med., Bd. 20. Ransom. Cited by Bramwell. Renvers. Cited by Lazarus, Die Anemie, II. Reyher. Deut. Archiv klin. Med., Bd. 39, p. 31. Rindfleisch. Virchow's Archiv, 1880, Bd. 121. Rosenstein. Berl. klin. Woch., 1877, No. 9. Runeberg. Tagebliltt Deut. Naturforsch., 1886, p. 147. Russell Brit. Med. Jour., 1889, I., p. 70. Sandoz. Correspondenzbliitt f. Schweiz. Aerzte, 1887, Sept. 15. Sandwith. Brit. Med. Jour., 1894, I., p. 1362. Sasaki. Virchow's Archiv, 1884, p. 96. Scheperlen. Schmidt's Jahrb., 1880, Bd. 186, p. 22. Schapiro. Zeit. f. klin. Med., Bd. 13. Schaumann, Tallquist. Deut. med. Woch., 1898, No. 20. Schanmann. Zur Kentniss d. Bothrioceph. Anemie, 1894. Scheby-Buch. Deut. Archiv klin. Med., 1876, Bd. 17. Scheimpflug. Zeit. f. klin. Med., 1885, Bd. 9, p. 58. Senator, Muller. Virchow's Archiv, Bd. 131, Suppl. Sorensen. Virchow's Jahresbericht., 1874, II., p. 314. Stengel. Therapeutic Gaz., 1894, p. 378. BIBLIOGRAPHY. 197 Stintzing, Guniprecht. Deut. Archiv klin. Med., Bd. 53, p. 265. Stockman. Brit. Med. Jour., 1895, I., pp. 965, 1025, 1083. Strieker. Charite-Annalen, 1875, Bd. 2, p. 277 ; 1877, p. 287. Wagner. Arcliiv f. phys. Heilk., 1859, p. 415. Waldstein. Virchow's Archiv, Bd. 91. Wilks. Guy's Hosp. Eep., 1857, p. 203. Wiltschur. Deut. med. "Woch., 1893, p. 715. Zappert. Wien. klin. Woch., 1892, p. 347. Zenker. Deut. Archiv klin. Med., 1874, Bd. 13, p. 348. CHAPTER VIII. LEUKEMIA. Historical. — Probable cases of leukemia were observed and com- mented upon from the beginning of the 19th century by Bichat, 1801, Andral, 1823, Hodgkin, 1832, Donne, 1830, and many others, and the disease was imperfectly recognized in current text-books (Piorry, Velpeau, Rokitansky) in which it was regarded as an obscure suppura- tive "hematitis" (Piorry). Donne, however, in 1839, found in his case at autopsy that the blood cells were more than one half " white or mucous globules," and attributed the disease to a failure of transforma- tion of the leucocytes into red cells. The first step in the elucidation of the disease was made when Craigie, in 1841, demonstrated an entire absence of suppurative foci in the body and concluded that the purulent material was absorbed from the en- larged spleen of which the histological structure " does not favor the gathering of pus in abscesses." In the same issue of the Edinburgh Medical Journal, Bennett de- scribed a second case, demonstrating an entire absence of suppurative foci, describing accurately the gross appearance of the blood in bulk, and under the microscope recognizing the leucocytes as identical in ap- pearance with pus cells, and adding to the lesion in the spleen a uni- form enlargement of lymph nodes. Regarding the nature of the disease he concluded that the excess of leucocytes in the blood must be classed as pus ; that in this case the pus was formed, entirely from the action of a zymotic principle in the blood, quite apart from inflamma- tory processes ; and that the enlargement of the liver and spleen and the pyoid condition of the blood " lead us to conjecture that in some way they stand in the relation of cause and effect." Virchow, a few weeks later, described a case complicated by super- ficial abscesses, and termed the condition " weisses blut " or leukemia. In a series of later articles he distinguished more clearly than Bennett had done between pyemia and leukemia, and emphasized the relation between the changes in the blood and the lesions in the blood-forming organs, holding the disease to be a primary affection of these organs. In 1851 Bennett contributed the first monograph on the disease, collecting 37 cases, describing and depicting the microscopical appear- ance of the blood, reporting the first chemical analyses, giving very reliable descriptions of the symptomatology and pathological anatomy, reporting cases of idiopathic splenic tumor without leukemia, and employing the term leucocythemia as preferable to leukemia. In 1847 Virchow observed a case in which the spleen was very HISTORICAL. 199 slightly and the lymph nodes greatly enlarged, and in his review of the subject in 1853 he pointed out the grounds for division of the dis- ease into splenic and lymphatic varieties. The chief of these grounds was the presence in the blood of the above case of cells which resembled, those of the lymph nodes. The leucocytosis was first demonstrated during life by Fuller (1846), who also believed that the alterations in the blood were not of in- flammatory origin. In 1851 Vogel, also, demonstrated the leucocytosis during life and attempted to count the cells. In 1869 Neumann ^ described the lesions of the marrow, demonstrat- ing the origin of the myelogenous variety of the disease, and his obser-, vations were verified and extended by Waldeyer (1871). Ponfick noted that in some cases the marrow is light yellow and puriform, while in others it retains more or less fleshy color, a variation which he referred to the extent of the cellular hyperplasia. The hyperplasia of the gastro-intestinal lymphatic structures and metastatic tumors of the pleura were described by Friedrich in 1857, while the enlargement of the liver was referred to collections of cells probably derived from the blood, by Bottger in 1858. In the morphology of the blood it was shown by Bennett and Vir- chow, that the leucocytes of the blood were identical in appearance with those of pus, although Virchow was careful to point out that they were not therefore of similar significance. These early observers both described the cells as mononuclear, or polynuclear, or devoid of nucleus (Virchow), and granular or hyaline. Virchow and most others mistook the eosinophile granules for fat particles. In his case of lymphatic leukemia Virchow described the cells of the blood as resembling those of the lymph nodes. No important advance in the knowledge of the minute structure of leukemic blood cells was made until Schultz and Erb in 1865 classi- fied leucocytes according to the character of their protoplasm. Neu- mann's discovery of the myelogenous type of the disease also directed more careful attention to the characters of the new leucocytes and strengthened the suspicion that the new cells of leukemic blood were not all of the same varieties as seen in normal blood. The question was finally settled by Hosier" who punctured the sternum during life and secured from the marrow large numbers of the identical cells seen in the peripheral circulation. It was at once concluded that these cells must be characteristic of the myelogenous type of the disease. In 1879-80, Ehrlich's studies aclded greatly to the knowledge of leukemic changes in the blood and were especially important in per- mitting greater accuracy in diagnosis. It then for the first time be- came possible to distinguish all cases of lymphatic from myelogenous leukemia, and to recognize early stages and other obscure phases of the latter type of the disease. The first pure case of myelogenous leuke- mia was described by Litten.* In the diagnosis of leukemia Ehrlich was able to find in the blood three chief signs: (1) Mononuclear cells with neutrophile granules. 200 LEUKEMIA. (2) an increased number of eosinophile cells, and (3) normoblasts or, later, megaloblasts. His conclusion that an increase of eosinophile cells, first noted by Jaderholm and Schwarze, is pathognomonic of the disease was soon relinquished. As shown by MuUer and Rieder, it is not so much the increased numbers but the peculiar forms of eosinophile cells which are characteristic of leukemia, although their numerical excess is often of itself pathognomonic. The ueutropliile myelocytes, as long before claimed by Hosier, were therefore again accepted as the most important cell form. Through the contributions of Hosier,^ Ehrlich, Cornil, Muller,^ Limbeck, Rieder, Troje, and many others, it was shown that these cells are derivatives of the marrow and probably peculiar to the blood of leukemia. The diagnostic importance of Cornil's large mye- locytes with pale eccentric nucleus was established through the obser- vations of Eberth, Eisenlohr, Hosier, Litten, Hayem, and especially of H. F. Huller. Later, it came to be recognized that neutrophile myelocytes are oc- casionally seen in small numbers in many other conditions. Conse- quently there remained as a diagnostic feature of the blood only the presence of a large proportion of myelocytes, and it was seen that the genuineness of some previously reported cases was open to doubt {e. g., Litten's). Absence of ameboid motion in the myelocytes of leukemia was de- monstrated by Biesiadecki, L5wit,^ Hayem, Huller, Gilbert, Rieder, and others. That this loss of ameboid motion is associated with a loss of reproductive capacity was stated by Hayet.^ A ])artial loss of ame- boid activity of other leucocytes was observed by Neumann and Cavafy, while J. Weiss does not admit any difference in ameboid properties of the white cells of leukemia and of other leucocytes. The discovery of nucleated red cells was followed by their demon- stration in the blood of leukemic cadavers by Erb, Bottger^ and Klebs,^ and in the circulating blood by Neumann.^ The comparative absence of these cells in lymphemia was first noted by Hayem, and verified by Wertheim, Rieder, and Dock. Hegaloblasts were first distinguished in leukemia by Ehrlich's pupil Utheman, and by Troje. Likewise the discovery of mitotic division of leucocytes by Flem- ming and Arnold led after a time to the demonstration of mitotic figures in the white cells of both the viscera and the blood of leukemia. (Arnold, 1884; Hayem, 1889; H. F. Huller, 1889.) Acute leukemia was first described by Friedrich in 1857, in a case lasting 6 weeks. Other reports of similar cases were added slowly, so that in 1889 Ebstein collected 16 cases lasting from 5-20 weeks, many of which appear to have been of the lymphatic variety. The knowledge of the chemistry o^ the blood has kept pace with that of the morphology. The first chemical analyses were those of Robert- son, 1850 (January), and of Parkes, while a few mouths later Strecker examined the blood of a case of Yogel's. Later important contribu- tions were made by Becquerel, Robin and Schercr. In these early ETIOLOGY. 201 analyses the chief chemical alterations in the blood were fully indicated. The specific gravity of the blood was found as low as 1.036-1 .049, that of the serum 1 .023-1.029. The iron was shown to be markedly dimin- ished and the proportion of water increased, with a loss of albumens, especially of the red cells. Scherer reported the isolation of several organic acids in considerable traces, and of hypoxanthin, a body pre- viously isolated from the spleen, while the presence of increased uric acid was demonstrated by both Scherer and Parkes, and later by Fowlwarckzny, Koerner, and Mosler. The later chemical analyses of leukemic blood and tissues have been contributed by Stintzing and Gumprecht, Limbeck,^ v. Jaksch, Salkowski, Bockendahl and Land- wehr, Freund and Obermayer.^ Etiology. Age. — The histories of the earliest cases (Bennett) indi- cated that the disease may occur at almost any period of life. Later statistics showed that in the male sex the liability increases from childhood to the thirtieth or fortieth years, thereafter declining. In women most cases occur during active sexual life. In infants, cases have been noted from birth, but the few cases recorded in childhood indicate that this period is comparatively immune. Audeod observed the disease in 56 children; in the 1st year 11 times; 2-4 years, 12; 5-9 years, 12; 10-15 years, 21 ; the majority were boys. Of 135 cases collected by Ehrlich, Grawitz, Litten, 87 were males, 48 females. Heredity has figured in some remarkable cases but its influences can- not usually be traced. Sanger observed a leukemic infant born of a mother who was healthy and remained so. On the other hand, many leukemic women have given birth to healthy children. The influence of direct heredity is rare. Casati cited the case of a girl of 10 years whose father and grandmother were leukemic, while Cameron observed a case in a woman whose grandmother, mother, brother, and two children were leukemic. Collateral heredity, the aifection of brothers and sisters, has been frequently observed. (Nanny n, Duret and Vaquez, Senator, Ortner, Casali, Greene, Eichhorst, Bernier.) Trauma has been demonstrated to be a direct exciting cause of leu- kemia in a large number of authentic cases. The traumatism has been applied directly to the spleen (Velpeau, Wallace, Mosler, Grawitz, Ebstein), or to the bones (Virchow, Grawitz, Mursick) in which case the myelogenous type commonly develops. The traumatism may con- sist in general concussion. Preexisting leukemia was thus rendered rapidly fatal in a case of Greime's. The significance of the alleged cases of leukemia following splenectomy will be considered under that topic. Intestinal intoxication has been urged as the exciting cause of leu- kemia by Vehsemeyer who believes that the failure of conversion of peptone into harmless albumens of the blood is the important causal factor in the disease. This theory was suggested by the frequency of gastro-intestinal disturbances and lesions in leukemia, and by some ob- servations of Kottnitz which indicated that in leukemia the gastro-in- testinal mucosa loses to some extent the power to transform peptones. 202 LEUKEMIA. The effects of bad hygiene have been traced in a large number of cases (Mosler), but no social position nor occupation distinctly predis- poses to the disease. Other diseases predisposing to or directly exciting leukemia have been reported as follows : Syphilis, acquired and constitutional, especially in children, and ex- citing usually the myelogenous type ; by Steinbrugge, Frankel, Mosler. Malaria, either acute or chronic (Muller,^ Gowers, Osier (33 per- cent)). Of 150 cases collected by Gowers, a history of previous malaria was obtained in 39, while in many the spleen had remained enlarged after the malarial attack. The disease has been attributed to diphtheria, by Orth, and to typhoid fever, smallpox, and pneumonia, by Gowers. Influenza has several times been reported as leading to acute or chronic leukemia. (Hinterberger, Frankel,^ Litten.') Rachitis, latent tuberculosis, and hypertrophic intestinal lymph-follicles from chronic catarrh (constitutio lymphatica?) have been noted, singly or combined, in many cases of mixed type in children. (Mosler.^) Stomatitis, ulceration of gums or tonsils, preceding or early ac- companying leukemia, have been reported by many observers (Ebstein, Hinterberger) but are probably secondary lesions. The transformation of pernicious anemia into leukemia has been re- ported in several cases. (Waldstein, Litten,^ Masius and Francotte, Leube and Fleischner, Musser.) It does not appear, however, that any of these cases indicate a relation between progressive pernicious anemia and leukemia. They seem to the writer to be cases of leu- kemia in which there was a temporary reduction in leucocytes at the time of first observation, or instances of marked terminal leucocytosis. The case of Masius and Francotte, in a patient suffering from anchy- lostomiasis indicates that severe secondary anemia may predispose to leukemia. An apparently genuine instance of transformation of pseudo-leukemia into acute leukemia is reported by Mosler,^ and Troje has maintained that pseudo-leukemia is pathologically a forerunner of leukemia. The genuineness of many reported cases of leukemia following in- fectious diseases, splenectomy, childbirth, anemia, etc., has been seri- ously doubted, and it has been agreed on all hands that spurious cases have crept into the literature, especially before the general employ- ment of Ehrlich's methods and criteria. The recent demonstration of numerous myelocytes or lymphocytes in inflammatory leucocytosis (Engel, the writer, etc.), the extreme leucocytosis (1-2), observed by Schede and Stahl in osteomyelitis, and Musser's observation of a pro- portion of one whijte to four red cells in pernicious anemia, render it obvious that the diagnosis of acute leukemia in the absence of micro- scopical examination of the marrow, is often a hazardous undertaking. Infectious Origin. — The chief ground for a belief in the infectious nature of leukemia is found in the rapidly fatal course of some acute ETIOLOGY. 203 cases and their striking clinical resemblance to other undoubted infec- tious diseases. Actual contagiousness is suggested by the report by Obrastzow of an acute case developing in a hospital attendant six weeks after nursing another fatal acute case of the myelogenous type. Biological studies of leukemic blood have led to the reported demon- stration of bacteria and of other micro-organisms of undetermined nature. Bacteria. — Kelsch and Vaillarcl squeezed drops of blood from the fingers and isolated therefrom a diplo-bacillus which was pathogenic for mice. The same bacilli appeared in the vessels of the tumors. Pawlowsky claims to have found one and the same bacillus in six cases, sometimes in the blood, or in the viscera, especially in the liver. In the blood they were found mostly in the plasma. In cultures on glycerin agar the bacillus pro- duced a somewhat characteristic growth. Injected into rabbits this micro- organism could be identified morphologically iii the blood for several weeks, but no symptoms of leukemia were produced. Although Pawlowsky believed he had discovered a specific microbe in the disease, he has not yet verified his results. Bonardi obtained cultures of Stcqihylococcus pyogenes aureus and albus from the blood of" two cases of splenic leukemia. Inoculations. — Mosler and Westphal injected leukemic blood into dogs and rabbits with negative results. Bollinger with the blood, and Gilbert, Cadiot and Roger, with the leukemic tissues of diseased dogs, failed to produce any significant eflects by injections into healthy animals. The injection of leukemic blood and of the fresh juice of the spleen into various animals was without results in the hands of Mo-sler and Westphal, Bollinger, Nette, Eichenbush, and Muller.'* Nette injected defibrinated blood under the skin, into the peritoneum, and into the veins of the ear, epigastrium, and marrow, with negative results. The juice of leukemic nodes was injected into animals without eiFect by Troje, Litten, and Gilbert. Negative cultures of blood were obtained in acute cases by Ebstein, Roux, Westphal and Mosler, Guttmann, Mayet, Eichhorst,i Pfeiffer, Litten, and Sittmann. Micro-organisms of undetermined nature have been described in the blood, or in the leucocytes, or in metastatic deposits, or in cultures, by Klebs,2 McGillavry, Osterwald, Mayet, 2 Roux, Bramwell, Fermi, and Klein. None of these writers has been able to follow up his observations. The observations of Mannaberg and of Lowit ^ point to the possible existence of protozoa in leukemic blood. In 8 percent of the lymphocytes of a case of lymphatic leukemia Manna- berg found hyaline bodies 1-4 /^i in diameter, exhibiting ameboid activity such as is seen in the quartan malarial parasites. They failed to exhibit a nucleus by Romanowsky's stain, but after fixation in Hermann's fluid they stained well with saffranin. In the peripheral blood and viscera in mj'^elemia Lowit has recently re- ported the discovery of an ameba which he classes Avith the sporozoa and terms hemameba leukemise magna. In lymphemia he finds another somewhat different ameba, hemameba leukemia; I'ivax. which is scarce in the blood, but abundant in the viscera. In some mixed cases both amebse were present, and they were also seen in pseudo-leukemia and in v. Jaksch's anemia. Lowit has found the same amebpe in leukemic animals and claims to have transferred the disease from one animal to another. The method employed by Lowit in the demonstration of the parasite of leukemia is as follows : 204 LEUKEMIA. The blood smeared in a thin layer is heated 1-2 hours at 110°-120° C. and stained one-half hour in a concentrated watery solution of thioniu. After thorough washing in tap-water, and drying in air, the specimen is laid for 10-20 seconds in solution of iodine (I, 1; KI, 2 ; Aq. dest. 300), again washed, dried, and mounted in balsam. The parasite in its various stages then appears green or olive green, or greenish black, the basophile granules and degenerated products of nucleus and cell body bluish red or brownish red, the protoplasm of leucocytes and their granules yellowish or yellowish brown. The blood smear must be thin. The solution of thionin should be several weeks or months old, long exposed to sunlight, and preferably con- taining a considerable growth of micro-organisms which seem to assist in "ripening" the stain. In the absence of an old solution good results may be obtained with a mixture of the thionin solution, 30 ; Loeffler's methylene- blue, 15 ; but the green color of the parasites may be obscured by a diffuse dark stain from the methylene- blue. The fading which ordinary prepara- tions suffer within a few days may be avoided by mounting the specimen in balsam which has been mixed with a solution of iodine in xylol until the balsam is of light yellowish ccSlor. Lowit regards the greenish stain produced by his method as the specific reaction of the parasite, and finds reason to believe that this color is assumed by a substance identical with or closely related to cellulose. Turk believes that the hemameba of Lowit is an artifact resulting from the partial solution and deformation of mast-cell granules by the action of watery dj^es. He claims to have found similar bodies in the blood of healthy men and rabbits. Pathological Changes in the Viscera. — Only the portions of this extensive subject will here be considered which bear upon the patho- genesis of the disease. Marrow. — In the early stages of the leukemic process the lymphoid marrow exhibits a slightly lighter color and greater consistence than is normal. This change results from a cellular hyperplasia, which obliterates blood sinuses, diminishes the blood content, and causes atrophy of fat cells. With this cellular hyperplasia of the marrow of ribs, vertebrse, etc., there is an extension of lymphoid marrow through- out the shafts of long bones where also the marrow appears firm and light colored. While in early stages the marrow may retain a fleshy tint, in advanced cases it is almost invariably very light colored and firm. In some advanced cases sclerotic and degenerative processes may follow the cellular hyperplasia, the writer having seen consider- able areas of connective tissue and foci of mucoid degeneration in the marrow of old cases. In acute leukemia the marrow may be strikingly puriform in ap- pearance, though invariably less diffluent than pus. In one case of the writer's this resemblance to pus was so marked as to require micro- scopical section before the hyperplastic marrow of the femur could be positively identified. In myelogenous leukemia the affection of the marrow has in all un- doubted cases been distinct and if Litten's case can be accepted, as is doubtful, the leukemic process may be limited to the marrow. Leube and Fleischer (1881) attempted to prove the contrary by a case show- ing a marked leucocytosis (1-1'2) and dying of progressive gangrene, in which no hyperplasia of the marrow was found. In lymphatic PATHOLOGICAL CHANGES IN VISCERA. 205 leukemia it was early shown that the marrow may be very slightly or not all involved (Heuck/ Fleischer, Penzoldt), although in many such cases the usual extensive hyperplasia is observed (Muller). Hirschlaff has recently reported two obscure cases, one of which appears to have been an example of pure splenic leukemia Avith peculiar blood changes. In the other no distinct leukemic lesions were found in the vis- cera. Similar negative results at autopsy have been recorded by Eichhorst - and by Ambros. The microscopical examination of the marrow in leukemia shows that the hyperplasia affects the variety of cells seen most abundantly in the blood. In myelogenous leukemia these cells are largely the neu- trophile myelocytes which are found in very excessive numbers, of very large size, and in mitotic division, or in various stages of degener- ation. In the same cases also there is almost invariably an excessive number of large mononuclear hyaline cells. Eosins and giant cells may long persist, but do not appear to take a prominent part in the hyperplasia. In pure lymphatic leukemia the lymphocytes alone are found in the hyperplastic areas (Muller^) and eosinophile and neutrophile cells are very scarce, but there appear to be all grades between the pure lym- phatic and myelogenous types of leukemia, in which the changes in the marrow correspond in general to those in the blood. It is held by some (Ponfick, Hosier, 2 Muller 2) that the leukemic process may at one period of its cour.se affect principally the lymphoid structures, at another the marrow, so that the blood may at one time exhibit the changes of lymphatic, at another those of myelogenous leukemia. Fleischer and Penzoldt's case illustrates such a variation. The red blood cells of the marrow are not greatly affected in the early stages of leukemia, as is indicated by the frequent absence of marked anemia, although in very cellular areas they are diminished in number from the first. Later, in chronic myelogenous cases the same changes are seen as in pernicious anemia, but the excessive numbers of megaloblasts are seldom so prominent. In some acute cases, especially those of lymphatic type, the writer has seen almost entire absence of nucleated red cells in considerable sections. The distribution of the lesion in the marrow of lymphatic leukemia is often very irregular. In an acute case the writer has found entirely normal structures in the ribs and femur while the usual hyperplasia was present in the bodies of vertebrae. Moreover the lesion, at least in its early stages, appears to be focal rather than diffuse. These pe- culiarities render it essential that very complete examinations should be made before it can be positively claimed that leukemia can exist without lesions in the marrow. Hirschlaff, Kormoczi, and Pappenheim report cases of lymphatic leu- kemia in which the lesion was exclusively in the marrow, and Pappenheim holds that lymphemia is always of myelogenous origin. This view is apjtar- ently disproved by the case of Rosenfeld's in which no change was dis- covered in the marrow. 206 LEUKEMIA. Liver. — The enlargement of this organ is usually referable to the presence of numerous metastatic growths, which follow the course of development described for all secondary leukemic deposits, and to in- filtration of capillaries with proliferating leucocytes. In myelogenous leukemia the liver is seldom free from such deposits, but in the lym- phatic type the infiltration is usually limited to the portal canals aud intralobular capillaries, although it may produce circumscribed tumors. It is of great interest to note that the liver in myelogenous leukemia gives evidence of having resumed its embryonal function of developing red cells. Neumann and Heuck first called attention to the presence of considerable numbers of nucleated red cells in the hepatic capillaries. In Heuck's case they were distinctly more abundant than in the gen- eral circulation. The resemblance of the leukemic to the embryonal liver is completed in the presence of mitotic leucocytes aud giant cells, as described by Neumann,^ Muller, and others. Spleen. — The pathological process in the spleen consists in hyper- plasia of large and small mononuclear cells. These cells appear in excessive numbers throughout the pulp cords, while the limits of the Malpighian bodies become less defined. As a result of this cellular hyperplasia the sinuses are choked with cells, and hemorrhages, in- farcts, degeneration, and necroses occur, which lead to a variety of gross appearances in the enlarged organ. Regarding the exact nature of the cellular processes it has been shown that in myelogenous leukemia, while the majority of the cells in the spleen are hyaline, yet neutrophile myelocytes occur in moderate numbers. These are generally regarded as derivatives of the marrow brought by the blood stream, although Muller has once seen mitosis in a splenic myelocyte. Numerous giant cells are sometimes found, which Muller regards as originating in situ. For the increased num- ber of large and small lymphocytes the chief mode of origin must be mitotic division in situ, especially in the Malpighian bodies, as has been demonstrated by Muller. Along with this excess of colorless cells, the organ commonly contains an excess of blood, and the usual process of destruction of red cells may everywhere be observed. Nucleated red cells are visible in smears of splenic pulp in most cases of myelogenous leukemia. The writer has been unable to con- vince himself that they are more numerous than in the general circu- lation, or if slightly more numerous, that they are not mechanically sifted from the blood. Heuck, however, found them in the splenic smears when they were absent in the general circulation and concluded that the spleen in leukemia may resume its embryonal function of developing red cells. Following the stages of acute hyperplasia, comes a later period marked by continuous inci'ease in the size of the organ, and great in- crease in consistency, owing to fibroid changes beginning in the trabecula and often extending diifusely throughout the pulp. In such stages the cellular content is comparatively diminished without any apparent effect upon the character of the blood. PATHOLOGICAL CHANGES IN VISCERA. 207 In lymphatic leukemia the same course of events is observed, but the new cells are limited to the large and small lymphocytes. In this form of the disease the lymph follicles are apt to be enlarged. The general appearance of sections of the spleen in myelogenous leukemia, the congestion, the indiscriminate filliug of sinuses with red and white cells, the destruction of Malpighian bodies, the prominence of fibroid changes, and the usual absence of areas of distinct leukemic deposits, strongly indicate that the enlargement of this organ results largely from mechanical sifting of red and ichite cells from the circulation and subsequent inflammatory changes. In lymphatic leukemia, how- ever, the proliferation of lymphocytes is usually very marked and the Malpighian bodies must be considered one of the most important pri- mary seats of the disease. In some cases it may perhaps be the ex- clusive seat of the primary lesion. (Cf. Hirschlaif, Rosenfeld.) Lymph Nodes and Other Lymphatic Structures. — Swelling of the lymph nodes is one of the earliest of observed symptoms, having in some lympathic cases preceded all other signs for some months. They may reach a considerable size, that of a hen's egg, but seldom become excessively hirge. Their capsules are almost invariably intact unless ulceration occurs. The stage of cellular hyperplasia is succeeded, here as in the spleen, by one of fibrosis with diminution in the number of cells. In myelogenous leukemia the smears of lymph nodes yield almost the same appearances as those of the spleen, except for a relative ex- cess of lymphocytes. Neutrophile myelocytes are scanty, but here again Muller has found mitosis in myelocytes. On section, the out- lines of the lymph follicles are lost, and the tissue consists of a diffuse mass of mononuclear cells, large and small, resembling lymphocytes and often seen in mitotic division. In some cases distinct prolifera- tion zones appear in small foci where the multiplication of leucocytes is very active. Although the sinuses are usually choked and indistinguishable, yet, according to Birch-Hirshfeld, most of the lymph nodes of leukemia differ from those of j^seudo-leukemia, in the fact that substances injected into the capsules pass through the leukemic node into the efferent vessel, but in pseudo-leukemia the enlarged nodes are impervious. The writer has been unable to find in literature, or in sections and smears of nodes from his own cases, any constant microscopical differences distinguishing the lymphatic from the myelogenous type of the disease. In the former the small lymphocytes are rather more numerous, and of course the hyperplasia is usually much more marked. In addition to the lymph nodes, spleen, and marrow, all preexisting lymphoid structures may become hyperplastic, especially in lymphatic leukemia, and new growths of lymphoid tissue have been found of nearly universal distribution. From all of these localities new leuco- cytes are undoubtedly contributed to the circulation. Character of Metastatic Deposits. — Although Rindfleisch, Cornil and Ranvier, and Zeigler, held that leukemic deposts arise by mechanical 208 LEUKEMIA. lodgment of increasing numbers of leucocytes from the blood, Virchow early maintained that these deposits arise in situ. It has since been shown that \Yhile the early minute collections of cells are ivhiie cell thrombi, the further growth of these masses takes place principally by mitotic division of these cells. (Bizzozero, Muller, and Neumann.) Bizzozero has pointed out that the structure of these metastases closely resembles that of Ivmph nodes with reticulum and sinuses, and is not at all that of an indiscriminate deposit of blood cells. He found in such deposits quite as many mitotic nuclei as in similar nodules of metastatic carcinoma. Leukemic deposits must therefore be regarded as conforming in many important respects to the laws governing meta- static neoplasms. The lymph nodes and spleen when affected exhibit an hyperplasia of the lymphocytes, which choke the sinuses, obliterate the follicles, and cause an overflow of these cells into the blood and probably also into the lymph stream. In lymphatic leukemia the new cells are of small size ; in myelemia the cells are either of small size or larger, and with a distant mass of protoplasm about the nuclei. The writer has made several attempts to demonstrate neutrophile granules in the cells of the lymph nodes in myelogenous leukemia but without success. In chronic cases, the stage of hyperplasia is regularly followed by one of fibrosis, which in the spleen greatly increases the size and con- sistence of the organ. Pathogenesis. — The conclusions regarding the pathogenesis of leu- kemia which seem to be warranted from the foregoing review of the etiology and pathological anatomy of the disease are to a large extent those formulated by Virchow and Neumann, whose original ideas have been partially readjusted from time to time to suit newly acquired facts. Leukemia must now be considered a primary disease of the blood- producing organs, principally or exclusively affecting the marrow or the lymphoid structures. The essential process consists in an excessive hyperplasia of the myelocytes or lymphocytes, or both varieties of blood cells, with secondary increase of leucocytes in the blood, secon- dary lesions in the blood-forming and other organs, and leading even- tually to disturbances in the production of red cells and to pernicious anemia. In the myelogenous type, the marrow and the neutrophile myelocytes are chiefly affected and the development of red cells is soon disturbed. In lymphatic leukemia the lymphoid structures (lymph nodes, spleen, lymph cells of marrow) are almost exclusively affected and red cell formation is less early disturbed. In many cases the type of the disease is mixed, and it appears probable that the disease may principally affect one organ (the marrow) at one time, and another organ (the lymph nodes) at a later stage. In either case the spleen is affected secondarily by mechanical lodgment of leucocytes and red cells, probably also by toxic chemical influences. In lymphatic leu- kemia the spleen may take active part in the initial process and furnish many new lymphocytes to the blood, but in myelogenous leukemia its PATHOGENESIS. 209 behavior is largely passive, and it seldom becomes the seat of true leu- kemic new growths but may as usual furnish lymphocytes to the blood. Pure myelogenous leukemia appears not to have been demonstrated, unless in some very acute cases, or in some reported cases of lymphemia. The importance of jiroJiferation of leucocytes in the circulation has been maintained especially by Biesiadecki and Lowit,^ who believe that leukemia is a primary disease of the blood. While nearly all ex- perienced observers agree that leucocytes multiply to some extent in the circulation, there is almost equal agreement that the extent of mitotic division in the blood is inconsiderable compared with that in the viscera. The leukemic deposits of the viscera develop by proliferation of original thrombi of leucocytes, which under favorable conditions un- dergo active mitotic division, develop a supporting stroma and behave in essential particulars like true neoplasms, while probably continuing to furnish new leucocytes to the blood. The red cells during the early stages of myelemia suffer from the disturbances of development such as exist in chlorosis. In later stages red cell production reverts to the embryonal megaloblastic type, mitotic megaloblasts are seen in the marrow, spleen, and liver, and the changes of pernicious anemia are established in the blood. The theory that polynuclear ueutrophile leucocytes are developed from lymphocytes met an insurmountable difficulty in the blood of lymphatic leukemia, and gave birth to a prominent theory of the pathogenesis of the disease now maintained chiefly by Lowit. Lowit, representing the views of most of the early observers, believes that the essential process in leukemia is a failure of full development of mononu- clear cells (lymphocytes, myelocytes). This theory, the discussion of which has cumbered the literature of the subject to an enormous extent, becomes untenable since the lymphocytes appear to have no developmental relation to ueutrophile cells. The partial resemblance of leukemia to a neoplasm affecting the lymph nodes, spleen, marrow, or blood, was early recognized and this theory of origin has been discussed by most writers and openly main- tained by Kottman, Bard, Herzfeld, and lately by Gilbert, who com- pares lymphemia to a sarcoma affecting lymphocytes. To what extent the origin and course of leukemia follows the laws known to govern neoplasms is an interesting inquiry which cannot here be pursued at length, but which may well be kept in mind. Accord- ing to such a theory it is necessary to suppose that lymphatic leukemia is a tumor originating in the lymphocytes, myelemia a tumor affecting myelocytes. There are, however, other tumors affecting lymphocytes, the lympho- sarcomata, which behave in a very different manner from lymphemia, in growth and especially in their metastatic characters. It seems not impossible, however, that all these differences may be referred to the involvement of the blood in the latter case and that all gradations ex- ist between lymphatic leukemia through pseudo-leukemia, to genuine 14 210 LEUKEMIA. lympho-sarcoma. The cases of lymphemia arising in the course of lympho-sarcoma are here of great interest. Changes in the Blood. The bulk of blood is apparently little altered by the uncomplicated leukemic process but in autopsies on chronic cases, the same diminu- tion in the total quantity of blood may be noted as in pernicious anemia, and from the same cause. The color of the blood may in well-marked cases be entirely normal, as anemia is of slow progress. Usually it is slightly lighter in color and more fluid, owing to pronounced anemia and great excess of leuco- cytes. The exuded drop has been described as lymphoid or puriform, but when properly expressed, blood of such character must be ex- tremely rare. When the red cells are below one million and the leu- cocytes almost equally numerous the blood drop still resembles blood, though of very light color. When obtained in bulk at autopsy the blood coagulates slowly, the red cells settle to the bottom followed by a characteristic thick layer of leucocytes and fibrin, while the very clear serum floats above. In the cadaver pure deposits of leucocytes re- sembling abscesses may be found in various situations often in the heart and pulmonary vessels. The coagulability is greatly diminished and may be entirely lost, but less often than in fatal pernicious anemia. To the touch, the blood may have a cohesive mucous quality, as noted by Grawitz. When leukemic blood is smeared and dried on a glass slide it ex- hibits a peculiar granular opaque appearance which is readily identi- fied as belonging to a great excess of leucocytes. The Red Cells. — In average chronic cases the red cells are usually reduced to 2-3 millions, falling to one million or lower in fatal cases, but not usually reaching the extreme reductions seen in pernicious anemia. On the other hand Cabot reports a case with over 5 million red cells, 134,000 white, Hb 78 percent. At death, the numbers of red cells depend upon the presence or absence of hemorrhage, and upon the length and character of the disease. There are no marked differences in the grade of anemia in lymphatic and myelogenous leu- kemia. Morphological changes in the red cells are invariably present. In the average chronic case the uniform loss of Hb in red cells showing nor- mal rouleaux and very uniform size and shape, with a rather free ad- mixture of normoblasts, gives to the stained blood smear a very peculiar and almost pathognomonic character. Possibly the abundance of leucocytes controls the movements of the red cells in the smearing process, but from some cause the appearance of the red cells in stained specimens of leukemic blood is very characteristic. (See Plate VII.) At this early stage the normoblasts are frequently present in greater numbers than in any other similar condition. PLA 1 h. VII, ^ • • • • • • '.*•>-• . • ♦. r^'.V^t-' . * .t'»" 'rX- •;>♦/.•/ 10 w Myelogenous Leukemia. (Triacid Stain.) Figs. I. Normal-sized red cells, deficient in Hb. Figs. 2. Pear-shaped poikilocyte. Fig. 3. Normoblast. Fig. 4. Myelocytes (Ehrlich's). Fig. 5. Myelocyte (Cornil's). Fig. 6. Myelocyte (eosinopbile). Fig. 7. Eosinopbile leucocyte (normal). Fig. 8. Blood plates. Fig. 9. Lympbocyte. Fig. 10. Polynuclear leucocyte, nucleus subdivided. THE LEUCOCYTES. 211 Later, the red cells begin to show the changes of more severe chlo- rotic anemia, differences in size and shape appearing, degenerative changes occurring with greater frequency and extent, till in many old severe cases, the features of pernicious anemia are fully established. Even in such stages, however, the appearance of the red cells usually differs distinctly from that of pernicious anemia, in the more uniform loss of Hb, and in the larger number of nucleated red cells among which a few are almost invariably normoblasts. Mitotic nucleated red cells are rare. In most other respects the same changes are seen as in pernicious anemia. Hemoglobin. — Estimates of Hb in leukemia are at present some- what unsatisfactory, owing to the opacity produced in the diluted blood by an excess of leucocytes. In the early stages of the disease the loss of Hb follows the reduction in red cells as in chlorosis, a diminution being demonstrable at times before any change in the num- ber of red cells, and the Hb-index remaining about .5-.6. Later, the estimates are untrustworthy, but the Hb-index probably rises. The Leucocytes. — The white cells are usually so much increased as to leave no doubt, from their numbers alone, as to the nature of the disease. In cases of very moderate severity there are usually from 100,000-200,000 leucocytes; Cabot's average of first examinations in 39 cases was 438,000. Rarely a million or more white cells are pres- ent in the cubic millimeter of expressed blood, but it appears doubtful if any such excessive number actually exists throughout the circula- tion . As a rule, the, severity of the general condition is proportionate to the increase of leucocytes, except in acute leukemia, when the num- bers of leucocytes may not exceed that of inflammatory leucocytosis. Variations in the number of leucocytes may be observed at differ- ent periods of the day, as shown by Hayem, who found 122,500 at 10 A. M., 235,000 at 4 P. M. of the same day. These rapid changes must be referred largely to vaso-motor disturbances leading to unequal distribution of cells, or possibly to a sudden, but temporary, increase of the leucocytes discharged from the marrow. Other variations in number of leucocytes will be considered under the course of the disease. Morphology of Leucocytes in Myelemia. 1. Neutrophile Myelo- cytes. — These are large mononuclear cells with neutrophile granules. It is of some moment to distinguish two varieties of neutrophile myelocytes in leukemia. (a) Cells of about the same size as normal polynuclear leucocytes, with well-staining, central nucleus, as described by Ehrlich and Uthe- man. These cells are abundantly present in leukemic blood and are generally the only form of myelocyte seen in secondary anemia. (Plate VII., Fig. 4.) (6) Cornil and H. F. Muller have called special attention to the large myelocyte loith pale eccentric nucleus. This cell is not usually seen except in leukemia but in some other conditions, especially in the 212 LEUKEMIA. secondary anemia of children, it may appear in moderate numbers. These cells have frequently been seen in mitotic division. (Plate VII., Fig. 5.) As first shown by Mosler, and abundantly verified by later observers, the presence of large numbers of neutrophile myelocytes is pathogno- monic of leukemia. In 28 cases (chronic?), Cabot found that between 20-60 percent (average 35 percent) of all leucocytes present were myelocytes. With the extreme leucocytosis of chronic leukemia there are nearly always enough myelocytes present to establish the diagno- sis as against any other condition that has yet been observed. In acute, myelogenous leukemia, however, this rule may fail and both the numbers and j^roportions of myelocytes may not exceed those seen by En gel in fatal diphtheria. Of recent cases of acute leukemia in which the type of the disease was satisfactorily determined, Frankel (1895) could find only three of myeloge- nous type, the great majority being of the lymphatic variety. In 1897 the writer saw three cases of rapidly fatal leukemia, verified by autopsies, and all of the myelogenous type. They occurred in the services of Drs. Thom- son and Delafield at Roosevelt Hospital. In two of these the changes in the blood were from the first observation typical of the condition and the mar- row was puriform, while the spleen and lymph nodes were but slightly af- fected. In the third case, on the first examination, with a leucocytosis of ordinary inflammatory grade, 5 percent of the cells were with difiiculty recognized as myelocytes. Eosinophile myelocytes and normoblasts were absent, and there was then no sufficient ground on which the diagnosis of leukemia could rest. Later the leucocytosis and the proportion of myelo- cytes increased and the diagnosis of acute leukemia, made with reserve just before death, was fully verified at autopsy. In this case, on account of the small number and indistinctness of the granules in the myelocytes, as well as the moderate grade of leucocytosis, the writer did not feel certain of the diagnosis until after the microscopical examination of the bone marrow. Degenerative Changes in Myelocytes. — The myelocytes of leukemia are usually deficient, and may be entirely lacking, in neutro- phile granules. In the latter case they are indistinguishable from large lymphocytes except by the great pallor of their nuclei (triacid stain). Such forms occur especially in acute leukemia. The nuclei of degenerating myelocytes may undergo hydropic degeneration. It is then practically impossible to determine the origin of these altered cells. (Plate IX., Figs. 1, 2.) (Cf. Grawitz, p. 122.) Some very small mononuclear cells with neutrophile granules seen in leukemic blood Ehrlich regards as the result of subdivision of polynuclear leucocytes (neutrophile pseudolymphocytes). In the fresh condition myelocytes fail to exhibit ameboid motion, a character which is regarded by many as indicating a loss of further power of development. 2. Polynuclear neutrophile leucocytes are excessively numerous in myelocythemia but are commonly found in diminishing proportions, though in increasing numbers, as the percentage of myelocytes in- creases. Cabot, using the triacid stain, found between 17—72 percent PLATE IX. Degenerating Leucocytes in Myelogenous Leukemia. (Triacid Stain.) Fig. I. Fig. 2. Fig. 3- Fig. 4- Figs ■ 5 Fig. 7- Fig. 8. Myelocyte deficient in neutrophile granules. Necrotic myelocyte, complete karyolysis ; hydropic degeneration ; loss of neutrophile granules. Degenerating myelocyte. Hydrops of nucleus. Polynuclear leucocyte, deficient in neutrophile granules ; hydrops of nucleus. 6. Polynuclear leucocytes. Loss of neutrophile granules. Advanced subdivision of nucleus. Myelocyte in mitotic division. ( Eosin and methylene blue.) Eosinophile leucocyte. Some granules basophilic. I THE LEUCOCYTES. 213 (average 46 percent) of polynuclear leucocytes. Even when the mye- locytes are relatively few their presence seems always to be at the ex- pense of polynuclear cells, a fact which indicates a slower progress to full development as well as increased production of these cells in the marrow. The most successful demonstration of nuclear figures in these cells {e. g., by Nocht's stain) shows very numerous transitional forms between the spheroidal and the multilobate nucleus, while after the triacid stain it is usually impossible to distinguish between many single and polymorphous nuclei in neutrophile cells. In lymphatic leukemia polynuclear leucocytes are usually scarce and may not be found at all. Degenerative changes in the polynuclear leucocytes are very common and very marked. Their cohesiveness is increased and they appear in large inseparable groups with myelocytes. Their nuclei are usually pale (karyolysis) and may become excessively faint. Gump- recht suggests that this pallor of the nuclei of neutrophile cells indicates the transformation of the nucleins into xanthin bodies. Or, the lobes may become more numerous, entirely separate, very compact and densely staining (pyknomorphous), while the granular cytoplasm is replaced by homogeneous highly refractive material. (Plate IX., Figs. 5, 6.) Or, the nuclei may undergo hydropic degeneration. (Plate IX., Fig. 4.) All grades of deficiency of granules may be observed, but the writer has been unable to demonstrate fat or glycogen in the neutrophile cells of leukemia. Considerable variations in the size of these cells may be noted in some cases of leukemia. 3. Eosinophile cells are usually much increased in number in myelo- cythemia but their proportions to other forms of leucocytes commonly vary within normal limits. In some cases, however, the proportions are increased but never so much as in cases of pemphigus, etc. Their proportions are not pathognomonic of the disease, but in most cases their total numbers greatly exceed those found in any other conditions. Ehrlich places their numbers between 3,000-100,000 per cmm., thereby prac- tically demonstrating, as he originally claimed, that the excess of eosins is among the pathognomonic signs of the disease. In some spec- imens of blood the eosinophile cells vary greatly in size, some of them being very small while their neighbors are hypertrophic. A few large basophilic granules may be found in the polynuclear as in the mono- nuclear eosins of leukemia. These cells are actively ameboid. In acute leukemia of both types and in chronic lymphocythemia eosins are scarce or absent. Eosinophile myelocytes are mononuclear cells with eosinophile granules. In some of these cells the granules are of uniform size and staining quality, or there may be some basophile granules among the eosino- phile, or the granules may vary greatly in size and in density of stain. Eosinophile myelocytes 7vith granules of very unequal size and density of stain are, in the writer's experience, pathognomonic of myelocythemia. (Plate VII., Fig. 6.) Bignami, however, speaks of the occurrence 214 LEUKEMIA. iu pernicious malaria of eosinophile myelocytes " such as are seen in leukemia." • In the majority of cases of myelocythemia eosinophile myelocytes constitute a large proportion of all eosinophile cells present. In lym- phatic leukemia they are almost always absent, and in acute leukemia they are scarce or absent. 4. Lymphocytes. — The numbers and proportions of lymphocytes in myelocythemia vary in different cases and at different periods in the same case. It may still be stated as a rule that the more purely mye- logenous the disease the smaller is the proportion of lymphocytes. But Muller has shown that the exact character of the leukemic process in the viscera is not always indicated by the changes in the blood, and that the greatest intensity of the process may be transferred from one organ to another. One must therefore be prepared to find marked variations in the proportions of lymphocytes without being able always to attach much significance to such changes. In most cases the lym- phocytes are notably diminished in proportion while probably always increased in actual numbers. Cabot found an average of 10.6 percent, but when at tiieir lowest (2 percent) they were still more numerous than in normal blood. 5. Large mononuclear leucocytes with very faint cytoreticulum and vesicular nucleus are seen in considerable numbers in most cases of leukemia, but they appear to lack special significance. They are very apt to suffer damage in the smearing process and to appear in the dry specimen as large coarsely reticular nuclei without demonstrable cell body. (Plate VIII., Fig. 3.) These cells may contain granules giving the reaction of glycogen. The larger forms may show horseshoe-shaped nuclei. In the cases of acute leukemia, described by Frankel and others, the majority of cells were large mononuclear leucocytes. Degenerative changes are noted in the small and often in the large lymphocytes of leukemia. The nuclei of the small cells, instead of remaining compact, may become incurved and finally bilobed or tri- lobed, while the cell body remains basophilic. (Rieder, Ehrlich.) Litten describes," in two acute cases, very large cells in which the nucleus was obscured by many globules of fat blackened by osmic acid. There are some other reports of fatty degeneration of leucocytes in leukemia, but apparently none from very recent literature. Mitotic figures in lymphocytes of lymphemia are reported by Wertheim. 6. Mast-cells. — These cells are very constantly increased in chronic myelogenous leukemia, but in some cases a prolonged search is required for their discovery. Although occasionally seen in other conditions, in leukemia they are usually so markedly increased as to constitute a very reliable diagnostic feature of the blood. Thev are at times more abundant than the eosins. They are usually absent in lymphemia and the writer has failed to find them in acute myelocythemia. Strauss found 5 percent of mast-cells among the leucocytes in blister- fluid from a case of myelogenous leukemia, and Milchner reports 23.9 percent of mast-cells in the sediment from ascitic fluid in a similar case. PLATE VIII. "^S^" . */ Lymphatic Leukemia. (Eosin and Metliylene Blue. Figs. I. Small lymphocytes. Figs. 2. Medium-sized and large lymphocytes. Fig. 3. Degenerating basket-shaped nucleus of large lymphocyte, without cytoplasm. Fig. 4. Polynuclear leucocyte. Nodal points of cytoreticulum. Fig. 5. Red cell. Polychroraasia of Maragliano. SPECIAL CHARACTERS IN LYMPHEMIA. 215 Special Characters of the Blood of Lymphatic Leukemia. — In this type of the disease the lymphocytes are the only form of white cell appearing in the blood in increased numbers. In cases in which a positive clinical diagnosis is possible the nature of the condition is at once evident from the great abundance of these cells, but owing to the numerous other causes of chronic lymphocytosis, it is at present impos- sible to state what is the lowest proportion of such cells seen in the blood of genuine cases. In well-marked cases they are quite as numer- ous as the leucocytes of myelocythemia, while some of the highest counts on record have been reported in cases of this type. Usually the lymphocytes are of small size and normal structure and the writer has specimens of one very marked case in which almost all the leucocytes are small mononuclears. Their percentage commonly runs between 80-90. In rather rare cases reported by Frankel, Grawitz, Cabot, and others, and in one child observed by the writer, the great majority of the cells were unusually large and their cytoreticulum faintly staining. Between the two extremes are cases showing various proportions of small, medium-sized, and large lymphocytes. The small lymphocytes are usually more abundant in chronic cases and in adults, the larger cells tending to become prominent in acute cases and in children. Gerhardt refers to a case in which large lymphocytes, abundant in the early acute stage, were replaced during a chronic course of six months by small lymphocytes. Frankel believed that these cells could be regarded as pathognomonic of acute lymphatic leukemia, but his claims have been fully disproved. (Grawitz.) Large mononuclear cells with hyaline bodies were common iu the writer's cases of acute myelocythemia, and reasons have been given to show that some of these cells may result from loss of neutrophile granules of myelocytes. Hirschlaff has recently reported 2 acute cases in Avhich most of the leucocytes were large mononuclear cells without granules. In a moderate proportion of them neutrophile granules and in a larger number, eosinophile granules were noted. This variation in size of lymphocytes may have an anatomical basis in the structure of the hyperplastic lymph nodes, as in some cases these nodes con- tain mostly small lymphocytes ; at other times large mononuclear cells. (Birch-Hirschfeld, Benda.) These and other considerations oppose the interesting view recently ex- pressed by Rosenfeld, that when the lesion of lymphemia is in the marrow many large lymphocytes appear in the blood, anemia rapidly develops, and the disease runs a severe course, while if the lesion affects only spleen and lymph nodes large lymphocytes are absent from the blood and anemia is of slow development. Myelocytes, both neutrophile and eosinophile, and mast-cells, are usually absent or extremely scarce in lymphemia, but are occasionally seen in scant numbers. (MuUer.) Polynuclear leucocytes, both neu- trophile and eosinophile, are also comparatively scarce. The red cells suffer the same changes as are seen in myelocythemia, but nucleated red cells are usually very scanty and may not be found at all. 216 LEUKEMIA. Variations in the Blood Changes in Leukemia. — Both the num- bers and the proportions of leucocytes in leukemia are subject to con- siderable variations from many causes. 1. Intercurrent Diseases. — A considerable number of cases have been reported showing that intercurrent infections may greatly alter the appearance of the blood in leukemia. These cases have been col- lected by Frankel and Marischler, Froelich, Cabot, Kormoczi, and McCrae. Some of the reports refer to terminal septicemia which has resulted usually in a marked and rapid decrease of leucocytes, without great change in the proportions of the different forms. Frankel, who saw the leucocytes fall from 220,000 to 1,200, refers this result to pure leucocytolysis. Such cases have been observed after typhoid fever, pneumonia, empyema, erysipelas, septicemia, tuberculosis, carcinoma, etc., but distinct changes in leukemic blood do not invariably result from such intercurrent diseases. In cases where the infection is local- ized (empyema, erysipelas) the normal activities of the marrow seem to be stimulated, and while the leucocytes diminish the proportion of polynuclear cells increases. (Freudenstein, Kovacs, Muller.') Kraus has recently reported a reduction of from 393,000 to 4,000 leucocytes in a few days as the result of double pneumonia and empyema. The vis- cera (marrow, liver, spleen) showed no evidence of leukemic infiltra- tion, and Kraus suggests that the suppurative process had not only transformed the blood, but had resolved the essential visceral lesions as well. Usually, intercurrent infections diminish the volume of spleen and lymph nodes even when, as in MuUer's case, the leucocytes are increased. After the subsidence of the infection the blood soon resumes its original condition (Grawitz), or even before the infection subsides, the first effect may pass off, as occurred in Eisenlohr's case within fourteen days. Although no permanent improvement resulted iu any cases siiffering from intercurrent infections, several attempts have been made to favorably influ- ence the course of leukemia hy artificial leucocytosis. Jacob reduced the leucocytes from 850,000 to 282,000 in a few days by repeated injections of glycerine extract of spleen, but could not report any improvement in the patient. The same result was obtained by Richter using spermin. Richter and Spiro claim to have increased the leucocytes in leukemia from 170,000 to 560,000 within 3 hours after injection of cinnamic acid, followed by prompt return to the previous condition. Heuck2 has also reduced the leucocytosis of leukemia by injection of tuberculin. In mild inflammatory processes there may be no effect upon the leucocytes of leukemia, as indi- cated by the cases of Richter and Heuck.i Ante-mortem leucocytosis of considerable degree (172,000) was ob- served by Thorsch in a case complicated by pneumonia. In a case of lymphatic leukemia dying of septicemia Muller found, four days before death, 400,000 leucocytes when there had previously been but 180,000. A differential count was not made, but Muller regarded the increase as referable to a polynuclear leucocytosis. Chronic infections have much less effect in altering the character of leukemic blood. Quincke, and Stintzing, reported a general improve- I CHEMISTRY. 217 ment of the leukemic process during acute miliary tuberculosis and during an exacerbation of chronic phthisis. In a case of lymphatic leukemia dying with carcinoma of the kidney, Marischler found a decrease of leucocytes from 96,000 to 48,000, with marked increase in the proportion of polynuclear cells. 2. Spontaneous Changes in the Blood of Leukemia. — A few instances are recorded in which the disease, as indicated by the blood changes, was slowly transformed from one type into another. The first of these cases was that of Fleischer and Penzoldt who observed splenic leukemia pass into lymphatic. It seems possible that this case may be placed with others described by FrankeP and Gerhardt, in which lym- phatic leukemia began acutely with large lymphocytes but progressed more slowly with small lymphocytes. Wey observed a case of mye- locythemia in which, within ten weeks, the polynuclear cells fell from 33.5 percent to 3.7 percent, the mononuclears rose from 66.5 to 96.3 percent among which were neutrophile myelocytes, large hyaline cells, and a few lymphocytes. Gerhardt observed the blood of a case of leukemia of marked grade pass into that of pernicious anemia, the excess of leucocytes disappear- ing in three days. The possible transformation of pernicious anemia into leukemia has already been discussed. Pseudo-leukemia has been shown to change into leukemia in several authentic cases. (See Pseudo-leukemia.) Chemistry. Specific Gravity. — The specific gravity of the blood is usually reduced owing principally to the loss of Hb. But leukemia is one condition in which Hb is often replaced by other albumens, so that the gravity of the blood is relatively high in comparison to the Hb-content, and in exact proportion to the increase of leucocytes. Thus Dieballa reports a case with 2.6 million red cells, 750,000 leu- cocytes, and gravity of 1.060. The lowest observations are placed by Grawitz between 1.029-1.023. The low figures are usually seen in cases complicated by hemorrhages or other causes of secondary anemia. Alkalescence. — The alkalescence of leukemic blood has been found by v. Jaksch, Peiper, and others, to be much diminished. Early observers found the blood distinctly acid soon after death (Scherer), probably owing to the post-mortem formation of acids, v. Noorden re- ferred the diminished alkalinity to the development of acids during life. Lactic and formic acids have been isolated from the fresh blood by Scherer, Hosier, et ciL, and acetic acid, after death only, by Hosier. Albumens. — Fibrin has been found in excessive quantity, 5.7 per- cent (normal .25 percent), by Parkes, but in chronic cases with severe anemia it is diminished. (Robin.) Peptone. — Considerable interest attaches to the demonstration, first by Bockendahl and Landwehr, of considerable traces of peptone in the splenic pulp and in the blood of leukemia. This observation has lately been verified by Hathes, who, however, has shown that the principle in question is not Kuhne's peptone, but deutero-albumose. Freund believes that the retarded coagulation of leukemic blood is largely ref- 218 LEUKEMIA. erable to the presence of albumose. The products of the excessive destruction of leucocytes have been traced in the presence of their vari- ous derivatives in blood and urine. NucLEO-ALBUMEX has been isolated from the serum by Mathes ; mucin or a closely-allied substance by Scherer and others ; and a prin- ciple resembling gluten by Salkowski. Xanthin bodies, which represent further decomposition-products of leucocytes, have been found in the spleen and blood by Scherer and many others. The xanthin bodies of Kossel are more abundant and more easily recognized in leukemic than in normal blood. Uric acid has been found in traces by Hosier, Fowlwarczny, Klem- perer, and Weintraud. An excessive excretion of uric acid in the urine of leukemia has often been observed, and Magnus-Levy reported as high as 8 gr. daily, but since the most abundant appearance of leu- cocytes need not indicate the time of their greatest destruction, there is no parallel between the excretion of this principle and the excess of white cells in leukemia. (Minkowski.) An increased quantity of glycogen has been extracted by Salomon and by Gabritschewsky. The fat-extract has been found distinctly in- creased by Robertson, Isambert (.72 percent), and Freund and Ober- mayer. Tyrosin was isolated by Fowlwarczny, and lecithin and choles- TERIN from the blood of a hematoma by Freund and Obermayer. Of Inorganic principles the iron has uniformly been found di- minished (2.24-2.97 percent of ash) usually in proportion to the loss of Hb. (Strecker, Scherer, Freund and Obermayer.) Phosphorus and sulphur, and sodium were much increased and potas- sium and chlorine much diminished in Freund and Obermayer's ^ case, while the total salts were moderately increased. Charcot-Leyden Crystals. — Robin first observed these peculiar crys- tals in the tissues of a leukemic cadaver, and they were described by Charcot and Robin in 1853. Zenker later (1855) claimed priority in their discovery. They were found in all parts of the circulation in a case examined after death by Wallace in 1855. Later they were care- fully studied by Charcot and Vulpian, who described them as color- less, refractive, elongated octahedra, .016 x .005 mm., or occasionally of much larger dimensions, insoluble in cold water, alcohol, ether, or glycerine, soluble in hot water and in most acids and alkalis. After death they appear in the blood, exudates, and especially in the spleen, and their numbers increase with post-mortem changes in the tissues. Like most other crystals of post-mortem formation they are often seen in or on the leucocytes. Their occurrence in sputum and relation to eosinophile cells was pointed out by Charcot and Leyden. According to Neumann ^ they are found in leukemia only when the blood con- tains many large cells with abundant cytoplasm and large nuclei, and hence are not seen in lymphatic leukemia. Chemically they were re- garded by Charcot as albuminates, by Salkowski as mucinous. Lately, Pohl has shown that they are identical with Bottcher's spermin crystals. THE DIAGNOSIS OF LEUKEMIA. 219 which are a product of the destruction of the nucleins of disintegrating cells, and probably consist of spermin phosphate. Scherer also con- cluded that they are composed of phosphoric acid and an organic base. They were obtained in the blood drawn during life by Neumann and later by many others. They are not peculiar to leukemia, but appear in the sputum of asthma, in feces and mucus surrounding intestinal parasites, and have been found in the fetal blood in simple anemia. (Gowers.) The Diagnosis of Leukemia. — In no other department is the ex- amination of the blood to be interpreted with greater certainty in one case or greater reserve in another, than in its application to the va- rious types of leukemia. 1. TJie changes in the blood may yield positive diagnostic signs of leukemia. This result is obtained in the great majority of cases of both types, whether acute or chronic. These signs are briefly : In myelocythemia. — An excessive leucocytosis (150,000-1,000,000). A large number and considerable proportion of neutrophile myelo- cytes (20-60 percent). A large number of eosinophile cells (3,000-100,000) of which many are mononuclear and exhibit very large densely staining granules. An excessive number of polynuclear neutrophile cells. Character- istic changes in the red cells. To these may be added : The presence of many mast-cells, many nucleated red cells, mitotic nuclei in leucocytes, extreme and peculiar degenerative changes in leucocytes. In chronic lyinphocythemia. — An excessive leucocytosis (150,000- 1,000,000). The presence of 85-99 percent of lymphocytes. Scanty numbers of myelocytes, eosins, nucleated red cells, and mast-cells. For the positive diagnosis of either type of the disease the essential point is the excessive leucocytosis, at least 150,000 cells, of which a considerable proportion are myelocytes or a still higher proportion are lymphocytes. No other condition thus far observed yields such characters in the blood, in the presence of which the diagnosis of leukemia is established beyond doubt. 2. The changes in the blood may justify only a probable diagnosis of leukemia. This situation is encountered under several conditions. (rt) In myelocythemia, the eifect of intercurrent infections may so reduce the number of leucocytes and especially the proportion of my- elocytes that the blood does not differ from that of some cases of acute inflammatory leucocytosis with 5-16 percent of myelocytes. (Diph- theria (Engel), pneumonia (Turk).) (6) In the less marked stages of lymphocythemia, especially of the acute form, the number and proportion of lymphocytes may be far from 220 LEUKEMIA. characteristic of leukemia. When there are less than 150,000 white cells and less than 90 percent of lymphocytes, the condition of the blood does not differ from that seen in some forms of inflammatory leucocytosis, or of lymphocytosis in the secondary anemia of children, or of the obscure condition called " v. Jaksch's anemia." In inflam- matory lymphocytosis, however, there are always a fair proportion of polynuclear leucocytes, which are very scanty in lymphemia, and the lymphocytosis is usually transient. In secondary anemia the lymphocytosis is seldom excessive, the pro- portion of lymphocytes is usually lower, and a few myelocytes, eosins, and nucleated red cells are commonly present. In " V. Jaksch's anemia " the lymphocytes are usually less numer- ous, there are more large hyaline mononuclear leucocytes than are usually seen in lymphemia of equal chronicity, and myelocytes, eosins, and many nucleated red cells, are usually present. Frankel has classed as acute lymphemia certain obscure cases attended with moderate hyperplasia of lymph nodes and the presence in the blood of a considerable number of cells regarded by some as large lymphocytes. Frankel's cases have not been fully accepted by his countrymen, and Grawitz especially has shown that the above condition of the blood is not pathognomonic of lymphemia, while Benda has paved the way for the divi- sion of these cases into other categories, including acute myelocythemia. Frankel's cases may belong in the class of lymphocythemia, but the condi- tion of the blood is not characteristic and the result of his autopsies is in- conclusive. 3. In some stages of leukemia the blood may fail to furnish indica- tions of the nature of the disease, which may then be overlooked. (a) Intercurrent infections have been shown to temporarily trans- orm the blood of leukemia into that of inflammatory leucocytosis. (6) In a case of acute myelocythemia, with the disease fully estab- lished, the writer found on first examination 5 percent of myelocytes with a leucocytosis of ordinary inflammatory grade. Later the myelo- cytes increased to the lower limits of leukemia, but a positive diagnosis was not established until the marrow was examined microscopically. (e) Most writers agree that the early stages of leukemia often escape detection, and that many spurious cases have appeared in literature. Bibliography. Leukemia. Ambros. Tnaug. Diss. Miinchen, 1893. Arnold. Virchow's Archiv, Bd. 97, p. 107. Audeod. Traite de Mai. de I'enfance. (Grancher), T. II., p. 112. Bard. Lyon Med., 1888, p. 239. Becquerel, Rodier. Recherches sur le Sang;., 1845. Benda. XV. Cong. inn. Med., 1897, p. 371. Bennett. Edin. Med. Jour., 1845, Vol. 64, p. 413. Monthly Jour. Med. Sci., 1851-52, XII., XIIL, XIV. Biesiadecki. Wien. nied. Jahrb., 1876, p. 233. Bignami. 20th Cent. Pract., Article, Malaria, 1900. Birch-Hirsehfeld. XV. Cong. inn. Med. Bizzozero. Virchow's Archiv, Bd. 79, p. 378. Bockendahl, Landwehr. Virchow's Archiv, Bd. 84, p. 561. Bollinger. Virchow's Archiv, Bd. 59, p. 341. BIBLIOGRAPHY. 221 Bonardi. Rev. gen. Ital. di clin. med., 1889, No. 5. Bottger. i Vircliow's Archiv, 1858, Bd. 14, p. 483. ^ibid., 1866, Bd. 36, p. 342. Bramivell. Cited by Pawlowsky. Cadiot, Roger. Path, du Sang. Traite de Med. (Charcot), Paris, 1892. Cameron. Cited by Stengel, 20th Cent. Practice, Vol. II. Casali. Eev. clin. Bologna, 1872, II., p. 118. Charcot, Robin. Compt. Rend. Soc. Biol., 1853, p. 44. Charcot, Vidpia7i. Gaz. Hebdom., 1860, p. 756. Cornii. Archiv d. Phys. norm, et path., 1887, T. 10, p. 46. Cornil, Ranvier. Histol. path. Craigie. Edin. Med. Jour., 1845, Vol. 64, p. 400. Dock. Anier. Jour. Med. Sci., Vol. 106, p. 152. Donne. Cited by Gowers, Lancet, 1878, I., p. 550. Eberth. Virchow's Archiv, 1868, Bd. 43, p. 8. Ebstein. Deut. Archiv klin. Med., Bd. 44, p. 343. Ehrlich. Gesam. Mittlieil., 1891. Elchhorst. 'Virchow's Archiv, Bd. 130, p. 365. 2 ibid., Bd. 113. Eisenlohr. Virchow's Archiv, 1878, Bd. 73, p. 56. Engel. Deut. med. Woch., 1897, pp. 118, 137. Er'b. Virchow's Archiv, 1865, Bd. 34, p. 138. Eiving. N. Y. Med. Jour., 1895, II., p. 161. Fermi. Cited by Litten. Fleischer, Pemoldt. Deut. Archiv klin. Med., Bd. 26, p. 368. Fowlwarczny. Deut. Zeit. f. prak. Med., 1875. Fran/eel. 'Deut. med. Woch., 1895, p. 666. 2 XV. Cong. f. inn. Med. Freudenstein. Inaug. Diss. Berlin, 1895. Freund, Obermayer. ' Zeit. f. physiol. Chera., Bd. 17, p. 318. ^ibid., Bd. 15, p. 310. Friedrich. Virchow's Archiv, Bd. 12. p. 37. FroeUch. Wien. med. Woch., 1893, p. 285. Fuller. Lancet, 1846, II., p. 43. Gabritschewsky. Archiv f. exper. Path., Bd. 28, p. 272. Gerhardt. XV. Cong. inn. Med., p. 382. Gilbert. Path, du Sang. Traite de Med. (Charcot), Paris, 1892, IT., p. 457. Gowers. Reynold's Syst. Med., 1879, V., p. 216. Grawitz. Klin. Path, des Blutes. Green. N. Y. Med. Jour., 1888, Feb. Greiwe. Berl. klin. Woch., 1892, No. 33. Giimprecht. Deut. Archiv klin. Med., Bd. 57, p. 523. Guttmann. Berl. klin. Woch., 1891, p. 1109. Hajek. Wien. klin. Woch., 1897, No. 20. Hayem. Du Sang., pp. 382, 856, 857. Herzjeld. N. Y. Polyclinic, 1894. Heuck. 'Virchow's Archir, Bd. 78, p. 475. ^T>e\xt. med. Woch., 1891, p. 747. Hinierberger. Deut. Archiv klin. Med., Bd. 48, p. 324. Hirschlaff. Deut. Archiv klin. Med., Bd. 62, p. 314. Isambert. Cited by Mosler. Jacob. Cited by Grawitz, p. 127. Jaderhohn. Cited by Litten (Nothnagel, Spec. Path.), p. 148. V. Jaksch. Zeit. f. klin. Med., Bd. 23, p. 187, also Klin. Diagnostik. Kelsch, Vaillard. Annal. de I'Institut Pasteur, 1890. Klebs. 'Virchow's Archiv, 1867, Bd. 38, p. 190. ^Eulenberg's Realencyc, Bd. I., p. 357. Klein. Cited by Pawlowsky. Klemperer, Weintraud. Deut. med. Woch., 1895, No. 40, V. B. Kormoczi. Deut. med. Woch., 1899, p. 775. Kottman. Inaug. Diss. Bern., 1871. Kottnitz. Berl. klin. Woch., 1890, No. 35. Kovacs. Wien. klin. Woch., 1893, p. 701. Kraus. Prag. med. Woch., 1899, No. 41. Lenbe, Fleischner. Virchow's Archiv, Bd. 83, p. 124. Limbeck. Prag. med. Woch., 1893, Nos. 12-14. Litten. 'Berl. klin. Woch., 1877, p. 256, also Krankh. d. Milz. ( Nothnagel' s spec. Path.), p. 133, etc. ^ xi. Cong. f. inn. Med., p. 159. Lowit. ' Sitzungsber. kais Acad. Wien., 1885, also ibid., 1887, Bd. 95, III. Abt., p. 22. 2 Die Leukemic als Protozoeninfec. Also, XVIII. Cong. inn. Med., pp. 251, 322. Magnus, Levy. Cited by Litten. 222 LEUKEMIA. Mannaberg. XIV. Cong. f. inn. Med., p. 252. Marischkr. Wien. klin. Woch., 1896, p. 686. Masius, Fmncotte. Cited by Ebstein. Mathes. Berl. klin. Woch., 1894, pp. 531, 556. 3Iayet. ' Compt. Rend. Acad. Sci., 1888, p. 762. ^Soc. med. de Lyon, 1888, T. 28, p. 47. McCrae. Brit. Med. Journ., 1900, I., p. 760. McGiUavry. Schmidt's Jahrb., Bd. 192, p. 19. Milchner. Zeit. f. klin. Med., Bd. 37, p. 194. Minkowsky. XVII. Cong. f. inn. Med., 1899. Mosler. ' Berl. klin. Woch., 1864. 2 Ibid., 1876, pp. 701, 720. ^ Virchow's Archiv, Bd. 114, p. 461. Mosler, Westphal. Semaine Med., 1889, p. 372. Midler. 1 Deut. Archiv klin. Med., Bd. 48, p. 51. 2 ibid., Bd. 50, p. 47. " Cent. f. Path., 1894, p. 628. 3ruller, Rieder. Deut. Archiv klin. Med., Bd. 48, p. 96. Mursick. Amer. Jour. Med. Sci., Vol. 69, p. 449. Musser. Trans. Phila. Co. Med. Soc, 1885. Neumann. 1 Archiv f. micr. Anat., Bd. 2, p. 507. 2 Cent. f. Med. Wissen., 1868, p. 689. 3 Ibid., 1869, p. 292. *Archiv f. Heilk., 1874, p. 441. ^ Berl. klin. Woch., 1878, p. 607. V. Noorden. Path, des Stoffwechsels, 1893, p. 342. Obastzmv. Deut. med. Woch., 1890, p. 1150. Osterwald. Archiv f. Ophthal., Bd. 27, p. 224. Parkes. Med. Times, 1851. Pawloivsky. Deut. med. Woch., 1892, p. 641. Ffeiffer. Cited by Litten (Nothnagel's Path. ). Poht. Deut. med. Woch., 1895, p. 475. Ponfiek. Virchow's Archiv, Bd. ()7, p. 368. Quincke. Miinch. med. Woch., 1890, p. 19. Benaut. Archiv de Physiol., 1881, T. 13, p. 649. Richter. Cited by Grawitz. Richter, Spiro. Archiv f. exper. Path., Bd. 34, p. 289. Rieder. ' ' Leucocytose. ' ' Rindfleisch. Lehrbuch d. path, gewebel. Robertson. Cited by Mosler. Roux. La Prov. Med., 1890, p. 280. Salomon. Charite-Annalen, 1878, V., p. 139. Salkowski. Virchow's Archiv, Bd. 81, p. 166, also Bd. 50. Sanger. Cited by Stengel, 20th Cent. Practice, Vol. II. Sehede, Stahl. Mittheil. a. d. Chir. Abt., Friedrichshain Hosp. Berlin, Leipsic, 1878, cited by Ebstein. ^ Scherer. Annal. de Chimie et de pharm., T. 194. Schwarze. Ehrlich's Gesamt. Mittheil. Senator. Berl. klin. Woch., 1882, p. 5.33. Sittmann. Deut. Archiv klin. Med., Bd. 53, p. 323. Steinbrugge. Zeit. f. Ohrenheilk., 1886, p. 238. Stintzing. Miinch. med. Woch., 1890, p. 19. Stintzing, Gumprecht. Deut. Archiv klin. Med., Bd. 53, p. 264. Strauss. Charite-Annalen, Bd. 23. Thomson, Eiving. N. Y. Med. Record, 1898, L, p. 333. Thorsch. Wien. klin. Woch., 1896, p. 395. Troje. Berl. klin. Woch., 1892, p. 285. Turk. Wien. klin. Woch., 1900, p. 293. XVIII. Cong. inn. Med., pp. 251, 325. Utheman. Ehrlich's Gesamt. Mittheil. Vehsemeyer. Miinch. med. Woch., 1893, p. 564. Virchow. Froriep's Notizen, 1845, No. 780. Med. Zeitung, Berlin, 1846, p. 157. Virchow's Archiv, 1847, Bd. I., p. 563. Vogel. Virchow's Archiv, 1851, Bd. 3, p. 570. Waldeyer. Virchow's Archiv, 1871, Bd. 52, p. 305. Waldstein. Virchow's Archiv, Bd. 91, p. 12. Wallace. Glasgow Med. Jour., 1855, III., p. 1. Weiss, J. Haematolog. Untersuch. Wertheim. Zeit. f. Heilk., 1891, Bd. 12. Wey. Deut. Archiv klin. Med., Bd. 57, p. 300. Ziegler. Lehrbuch d. Allg. u. Spec. Path. Anat. Zenker. Deut. Archiv klin. Med., 1876, p. 125. CHAPTER IX. PSEUDO-LEUKEMIA. Definition. — Pseudo-leukemia is a primary disease of the lymphatic structures, characterized by progressive enlargement of various chains of lymph nodes and of the spleen, by metastatic growths of lymphoid tissue in many localities, and by progressive anemia, but diifering from leukemia chiefly in the absence of marked leucocytosis. The disease does not belong with aifections of the blood, but on ac- count of its pathological resemblance to leukemia, is properly described in this connection. Historical. — The first attempt to demonstrate the specific character of the lymphadenopathy of the disease appeared in 1828 when Craigie called attention to the dilEference between certain firm tumors of lymph nodes and the caseating scrofulous and the cancerous enlargements. In 1832 Hodgkin described several cases observed chiefly at Guy's Hospital, some of which were cancerous, tuberculous, or syphilitic, but two of which were undoubtedly genuine examples of pseudo-leu- kemia. He distinguished the "■ organized " nodes from the caseous and cancerous ones. Before the discovery of leukemia therefore it was known that there are enlargements of the lymph nodes apart from cancer and tuberculo- sis. In 1839 Velpeau drew attention to the occurrence of hypertro- phied lymph nodes apart from scrofula. In 1856 Wilkes described several cases which he regarded as a special form of disease of lymph nodes but did not fully distinguish the lesions from those of tuberculosis. The first complete description of the malady appears to have been that of Wunderlich, 1858, who noted an idiopathic origin of the dis- ease, and the cellular character of the new growths in both nodes and spleen, mentioned the severe anemia, and called attention to the ab- sence of the extreme leucocytosis which characterizes leukemia. In the same year, 1858, Billroth described the histological structure of the enlarged nodes, noting the limitation within the capsule, the oblit- eration of follicles and of lymph sinuses and vessels, the proliferation of cells by nuclear division, and concluding that the hyperplasia is closely related to and may pass into sarcomatosis. Recognizing in the condition a tumor of lymph nodes of fatal tendencies but differing from sarcoma, he employed the term " malignant lymphoma " as spe- cially applicable to this condition. In 1864 the general pathological features were described by Virchow under the term " lymphosarcoma." Cohnheim in 1867 described a case, and noting especially the absence of leucocytosis suggested the term "pseudo-leukemia." The nearly constant afi'ection of the spleen was noted by the earlier 224 PSEUDO-LEUKEMIA. observers, AVilks, Woillez, Griesinger, Muller, and Strumpel, and the prominence of this symptom in some cases led Griesinger to employ the term " splenic anemia " for such forms of the malady. In France Bonfils described a case in 1858, using the designation '- cachexia sans leukemia'^ on account of the absence of leucocytosis. A few other examples of the disease were reported before 1865 when Trousseau described in detail the clinical characters and proposed the term " adenia." French writers from Trousseau to Gilbert have regarded the condition as very closely related to or identical with leu- kemia, employing the term " lyinphadenie aleukemique.'' In 1870 a full description of the disease, clinical and pathological, with historical notes to date was published by Murchison and Sander- son. Cornil and Ranvier in 1867 very fully described the histological structure of the nodes, regarding the process as a true lymphadenoma. In 1887 Ebstein^ described an acute form of the malady. Later writers have described similar cases possibly including among them some conditions not related to pseudo-leukemia. (Banti, Potain, Bruhl.) The peculiar localization of the lesions to different structures has been illustrated in many reported cases. Bonfils and Trousseau showed that the spleen need not participate at all in the lesion. A special aflFection of the tonsillar ring was described by Demange, and of the intestinal mucosa by Gilly, and later by many others. A testicular form has been described by Monod and Terillon. A case of Reine- berg's indicates that the lesion may be limited to the marrow. Many have regarded the anemia infantum pseudo-leukemica of v. Jaksch as the splenic form of Hodgkin's disease occurring in an infant. (Luzet, Gilbert.) The dermal type of the disease has been described by many writers, including Biesiadecki, Kaposi, Hochsinger and Schiff, and Joseph, but appears to have been first described in France, by Gillot. Anatomical Characters. — The lesions in reported cases of the disease have included inflammatory hyperplasia of lymph nodes, lym- phadenoma, and various forms of sarcoma of lymph nodes. 1. Simple inflammatory hyperplasias have been described in genuine cases in slightly swollen glands, not yet reached by the true lymph- adenomatous process. (Gowers.) 2. Lymphadenoma or Lymphoma. — In the majority of cases the lesions encountered in the nodes are those of lymphadenoma, i. e., the auton- omous growth of lymphoid tissue with small lymphocytes supported by reticular connective tissue. In the typical chronic cases one lymph node in a chain is affected, fol- lowed by the other members of the chain, or the entire chain is affected simultaneously. The disease seldom spreads by gradual extensions from one chain to another, but suddenly involves a new chain on the same or opposite sides of the body. Sarcoma of lymph nodes affects the intervening tissues between chains of nodes. For a long period the swollen nodes are retained in their capsules, but late in progressive cases the members of the chain are variously fused and their capsules largely obliterated. This fusion may result either from rupture of the capsules or from periadenitis. ANATOMICAL CHARACTERS. 225 Microscopical examination shows the new tissue to be composed of lymphocytes supported by reticular tissue. There are considerable variations in the character both of the cells and the reticular tissue. The cells are usually lymphocytes of small and medium size, among which are a few polynuclear leucocytes and occasionally (Goldman, Kanter) eosinophile cells. They are either diffusely distributed, or the normal follicles may be preserved for a long period. In the early stages, or in less active growths, the lymph paths are preserved, but later are obliterated. The reticular tissue is frequently less abundant than in the normal node, later a diffuse fibrosis occurs and the propor- tion of cells diminishes. The varying proportions of cells and tissue yield soft and cellular or hard and fibrous growths. The latter is probably a later stage of the former condition, but all stages are com- monly seen in different nodes of the same subject, and dense nodes are said to have been replaced by softer ones, (Gowers.) There is little tendency in these structures toward caseation, or suppuration, or hya- line changes of chronic tuberculosis. The lymph nodes of pseudo-leukemia may not differ in microscopical structure from those of lymphatic leukemia in any known particular. In myelogenous leukemia the lymph nodes commonly show a diffuse growth of larger mononuclear cells. According to Birch-Hirschfeld ' the nodes of leukemia can be injected through afferent vessels and the fluid will pass through into the efferent vessels, while in pseudo-leukemia the injection is imperfect and fluid fails to pass through. Attempts to demonstrate lymph vessels about pseudo-leukemic nodes have often failed (Billroth) but sometimes they are found distended with lymph. The occlusion of lymph paths may prevent the afflux into the general circulation of cells from lymph nodes, but it still remains an obscure fact that the very abundant and unusually diffuse lymphoid deposits in the viscera fail to cause lymphocytosis. 3. Lymphosarcoma. — In many cases, especially of more acute type, the lymph nodes in the group are fused together, capsules and sur- rounding structures are infiltrated and destroyed, different chains are united by a continuous growth, and the process has all the gross char- acters of a malignant tumor. Such nodes show no traces of sinuses or follicles but are composed of a diffuse growth of cells usually larger than lymphocytes, often containing giant-cells, and sometimes masses of fusiform cells. Two views are possible regarding the classification of such growths. One may enlarge the scope of the term pseudo-leukemia to include lymphosarcoma or may throw out such cases of lymphosarcoma and limit the application of the term to the lymphadenomata. In the present state of our knowledge, there apjiears to be no escape from the former alternative. There appear to be all transitional forms between the chronic lymphomata and the rapidly growing infiltrating giant- celled lymphosarcomata. In at least one case it has been possible to observe the transformation of pseudo-leukemia into sarcoma of lymph nodes. (Eisenmenger.) Moreover, of the cases of pseudo-leukemia 15 226 PSEUDO-LEUKEMIA. which have passed into leukemia some have shown the histological structure of giant-celled lymphosarcoma. There remains as a separate class of sarcoma of lymph nodes, the spindle-celled sarcomata or endo- theliomata, which arise from the reticular stroma. Acute Pseudo-leukemia. — Various acute and fatal processes of somewhat uncertain nature have been classed with pseudo-leukemia. 1. That true pseudo-leukemia may run an acute course appears from the reports of Cohnheim, Eberth, and Falkenthal, in which the disease lasted from 11 days to 4| months, and in which microscopical exami- nation showed the presence of lymphadenomatous hyperplasia of the nodes and metastatic growths in the viscera. These cases occurred in children under 15 years. 2. Chronic pseudo-leukemia may suddenly assume a grave character and prove rapidly fatal from the ulceration of enlarged lymph follicles in the intestines, as first shown by Berthenson's cases, and now not infrequently verified at the autopsy table. 3. Acute lymphosarcoma may run a rapidly fatal course and lead to changes in the viscera which are very similar to those of pseudo- leukemia. Fagge collected several cases to illustrate this fact, but their real ifature is not always apparent, and the microscopical reports were meager. They were all attended with purpuric symptoms. The most rapid case of lymphosarcoma that the writer has seen lasted four months. The nodes showed a diffuse growth of mononuclear cells much larger than Ij^mphocytes and few giant cells. There was prompt recurrence of the tumor in loco, rapid metastases, marked anemia, and no leucocytosis. Much more rapidly fatal cases have been observed, while Birch-Hirschfeld's case of true Hodgkiu's disease following typhoid fever was fatal in 6 weeks. Relation of Pseudo-leukemia to Leukemia. — The identity of the process in the lymphoid tissues in pseudo-leukemia and leukemia is strongly indicated by the histological features already described. The process in the former condition differs from that in the latter in the occlusion of lymph vessels, in the more diffuse and abundant metastatic growths, usually in the more rapid development, often in the tendency to infiltrate and rupture the capsules, in the lesser ten- dency to retrogressive changes, and in the more frequent transformation into giant-celled lymphosarcoma. Pseudo-leukemia may, it is claimed, lead to the same changes in the marrow as are seen in lymphatic leukemia. (Perrin, Schulz, Dyren- futh, Kelch, Poufick,' Schmuziger.) The transformation of pseudo-leukemia into leukemia has been reported in several cases. Most of these on examination prove to be of uncertain nature or were undoubtedly instances of terminal or moderate temporary leucocytosis in the course of some disease of the lymph nodes. Only three of these cases appear to be genuine, and all refer to the appearance of extreme leucocytosis during the course of pseudo-leukemia. M osier reports an acute case in a child of 14 years, whose blood a few weeks after the onset of the disease showed no leucocytosis, while about RELATION OF PSEUDO-LEUKEMIA TO TUBERCULOSIS. 227 three weeks before death the white cells were as numerous as the red. The varieties of cells were not stated. Senator has reported a case of trans- formation of V. Jaksch's anemia into leukemia. The case of Fleischer and Peuzoldt is the most significant. Their patient was a male of 40 years who suffered from enlargement of many lymph nodes for 16 months before death. Four months after the beginning of the illness there was no leucocytosis, but after 12 months there was one lymphocyte to eight red cells. The spleen was moderately enlarged, the lymph nodes extremely large ; there was dif- fuse Ij'mphoid infiltration of the liver, but the marrow of the femur Avas normal. The microscopical structure of the nodes was not reported. It thus appears that the statement that pseudo-leukemia may pass into leukemia rests upon the observation of one undoubted case only, in which the report meets present requirements, and that one was pe- culiar in other respects. The more recent case of Posselt is an ap- parently genuine example of the same transformation but the clinical history is meager. The remarkable case of Lucke's, of lymphosarcoma with rupture into a vein and development of lymphatic leukemia is also of interest in this connection. One must therefore accept with considerable reserve the opinion that some forms of pseudo-leukemia represent an aleukemic preliminary stage of leukemia. The vast ma- jority of cases of pseudo-leukemia pursue a chronic course showing no tendency to develop leukemia, and recent experience does not support Rothe's belief that early pleurisy and pneumonia carry off many cases of pseudo-leukemia before the leukemic stage has had time to appear. Eisenmenger reports a case of pseudo-leukemia which, after pursuing the ordinary course for four years, developed malignant tendencies in- filtrating surrounding tissues and perforating the larynx. The pres- ence of 90,000 leucocytes in this case would have led many observers to suggest the presence of lymphatic leukemia as well. Relation of Pseudo -leukemia to Pernicious Anemia. — Runeberg reported a case showing pernicious anemia, which he regarded as an ex- ample of purely myelogenous variety of pseudo-leukemia, on account of the lymphoid changes in the marrow. The transformation of perni- cious anemia into pseudo-leukemia was claimed to have occurred in a case described by Laache, and there can be no doubt that the marrow in some cases of pernicious anemia greatly resembles that of pseudo- leukemia, as first noted by Scheby-Buch, and Pepper. On evidence of this sort Sevestre suggested that all cases of pernicious anemia fall in the class of pseudo-leukemia but this view cannot be supported on either clinical or anatomical grounds. Relation of Pseudo-leukemia to Tuberculosis. — That cases of tuberculous lymphadenitis cannot always be distinguished clinically from pseudo-leukemia was early recognized, and so commonly observed that the scope of the disease was by some widened to include such cases. Many of the cases of so-called chronic intermittent fever with swelling of lymph nodes described by Pel, Ebstein,^ Renvers, and many others, probably belong in this class, as shown by Combemale. The exact significance of these cases appears not to have been sus- pected until Askanazy in 1888 demonstrated tubercle bacilli in the 228 PSEUDO-LEUKEMIA. swollen nodes of snch a case. In most of the cases in which micro- scopical examination of the nodes was reported there were distinct evidences of tuberculous inflammation, in the presence of miliary tubercles, or caseous foci, or at least abundant hyaline degeneration, and often there was a general tuberculosis (Delafield), but Waetzoldt's case furnished some nodes showing pure lymphoid hyperplasia, with onlv very scanty and minute foci of hyaline material, although coyitain- inq many tubercle hacilll in seetions. Finally Brentano and Tangl have described a chronic case in which, with tuberculous lesions in other regions, the lymph nodes showed no recognizable evidences of tuber- culosis, not even minute hyaline areas, nor bacilli in section. Inocu- lation experiments, however, demonstrated the tuberculous nature of the process. Sabrazes has reported a similar case. On this evidence, of which further confirmation appears desirable, it is necessary to admit either that (1) tuberculosis of lymph nodes may follow the most typical course of acute or chronic pseudo-leukemia, or (2) that one form of pseudo- leukemia is tuberculous, or (3) that tuber- culosis and pseudo-leukemia may co-exist. It is evident also that mi- croscopical examination of lymph nodes is not a sufficient test of the tuberculous nature of lymphomata, and that the entire subject requires readjustment on the lines indicated above. That all cases of pseudo-leukemia are not tuberculous, however, is shown on this same basis by the negative results of inoculation ob- tained by Westphal, Sciola and Carta, and others. Infectious Origin of Pseudo -leukemia. — Acute or chronic, non- tuberculous, in-flammatory hyperplasia of lymph nodes may reach such extreme degree as to resemble the condition found in pseudo-leukemia. The reported cases of this character are numerous, and have arisen from a great variety of infections, as from carious teeth (Ebstein,- Stengel), quinsy (Ponfick '), ulcerative pharyngitis (Chvostek), typhoid fever (Birch-Hirschfeld"). Other cases have been referred to otitis media, chancroid, eczema, etc. The writer has seen more extensive lymphoid hyperplasia in the intestine in typhoid fever than existed in two cases of pseudo-leukemia dying with intestinal ulcerations. Microscopical examination, where reported, has shown that the hy- perplasia of nodes is inflammatory in character and not lymphomatous, as in the cases of Lannois and Courmont, while metastatic growths are wanting. Many of these cases have resembled purpura hemorrhagica, as pointed out by Kossler and regarded by him as evidence of infec- tious character. Lannois, Courmont and Guillerraet isolated pyogenic cocci in their cases. The occurrence of these febrile forms of the disease and its fre- quent development after many cases of suppurative inflammation, espe- cially of the mucous membranes, early suggested the idea that pseudo- leukemia is, in all instances, of infectious origin. This view was reached by Westphal after a full review of the subject in 1893, and has since been maintained by Barbier, Verdelli, and others. Verdelli collected fifteen cases of pseudo-leukemia in which the presence of pyo- THE CHANGES IN THE BLOOD. 229 genie organisms was demonstrated in the blood or lymph nodes, viz., Staphylococcus jyi/ogenes mireus, 7 times ; Streptococcus pyogenes, 3 times ; pneumococcus of Frankel, once ; unidentified cocci or bacilli, four times. Barbier refers the failure of suppuration in these cases and the nega- tive results of some inoculations to a diminished virulence of the germ or to relative insusceptibility of the individual. This accumulating evidence points to the possibility of separating from true pseudo-leukemia a considerable group of inflammatory hyperplasias of bacterial origin. At the same time it must be admitted that the micro-organisms in some of the above cases were probably the cause of secondary infectious only, and there are opposed to these positive results of culture a number of negative reports. It may be added that Lowit claims to have found the hemameba leukemice magna in a case of jjseudo-leukemia. Summary of Etiology. — It thus appears that the group of cases now regarded as falling in the class of pseudo-leukemia passes by in- sensible gradations at one point into the inflammatory hyperplasias of lymph nodes, at another into the true lymphosarcomata, and at another into lymphatic leukemia. The field may undoubtedly be greatly simplified by invariable resort to inoculation to demonstrate the presence or absence of tuberculosis. These and all other cases of inflammatory origin should be separated from the genuine disease. There remains a considerable number of cases in which the clinical character of the disease and the microscop- ical structure of the nodes resemble those of sarcoma. It seems best, as suggested by Kundrat, Dreschfeld, and others, to separate such cases also from the others, although the distinctions between frank lympho- sarcoma and pseudo-leukemia are not sharply drawn, and transitional stages occur. (Sharp.) Likewise the increasing number of apparently authentic transformations of Hodgkin's disease into lymphatic leukemia strongly suggests a close relation between these two conditions. The evidence at hand seems to throw lymphatic leukemia, pseudo-leukemia and lymphosarcoma into a single group possibly connected by some common etiological factor. The Changes in the Blood. Red Cells. — It is a uniform observation that in the early stages of pseudo-leukemia the anemia may be very slight. The red cells fre- quently number five millions or more, when the nodes are distinctly swollen. With the progress of the lesion there is usually a progressive loss of red cells, which, however, is less marked than in corresponding stages of leukemia. In fatal cases there may be surprisingly little anemia, usually the cells fall below three millions, and occasionally the condition of the blood resembles secondary pernicious anemia. In acute cases the anemia may rapidly increase, and the disease resembles a septic infection or a malignant new growth. In morphology the red cells usually show the changes of simple secon- dary anemia of chronic course. Laache has called attention to the 230 PSEUDO-LEUKEMIA. very uniform size of the red cells in the average case. In some of the Avriter's cases the cells have been very uniformly undersized. Later, megalocytes deficient in Hb may appear, but they are seldom numerous, and do not lead to confusion with pernicious anemia. Laache also re- ports an obscure case regarded as showing the transformation of perni- cious anemia into pseudo-leukemia. Nucleated red cells are usually very scarce, even in late stages, and when present are usually of moderate dimensions. Jawein reports an obscure case of splenic type, and afebrile course, in which there were 3^840-5,914 normoblasts. The condition of the blood resembled that of von Jaksch's anemia, but the patient was an adult. The Hb is in all cases diminished. A low Hb-index is commonly seen in early stages or with slight diminution of red cells, while in ad- vanced cases with marked reduction of red cells the Hb-index is usually higher. These characteristics are common to most secondary anemias. The anemia of the typical case of Hodgkin's disease stands in an in- termediate position between simple secondary and pernicious anemia. It is usually much less marked than in corresponding stages of leu- kemia and almost never approaches the type of pernicious anemia. On the other hand it is usually more marked than the general condition of the patient would lead one to suspect, thus differing from most forms of secondary anemia. As Grawitz has pointed out, the character of the blood changes offers little encouragement of the tendency to find in this disease a condition related to either pernicious anemia or leukemia. Leucocytes. — The number of leucocytes in the blood is considerably influenced by the character of the process in the lymph nodes. In the majority of cases the leucocytes are normal or diminished in number, and there is a tendency toward relative lymphocytosis. In many cases however the leucocytes are continuously increased, with periods of well- marked leucocytosis. The increase of white cells may be considerable but does not pass beyond the limits of inflammatory leucocytosis, 50,- 000-60,000. Exceptions to this rule must be allowed in the cases of pseudo-leukemia which appear to have passed into leukemia, while in Eisenmenger's case, which passed into sarcoma, there were 90,000 leucocytes. A few eosins are commonly seen in afebrile cases, while an occasional myelocyte has sometimes been encountered. The cases showing few white cells the writer has found to be usually the chronic ones of slow progress and without fever, while a high pro- portion of poly nuclear cells with or without absolute increase usually belongs to the more rapid or the febrile cases. Limbeck finds that when the polynuclear cells are increased the nodes usually show in- flammatory changes, but when lymphocytosis is found the nodes ap- proach more to the sarcomatous type. In lymphosarcoma also,' not simulating pseudo-leukemia, the small lymphocytes may be very nu- merous. The writer has been unable to find any features in the blood distinguishing between pseudo-leukemia and frank chronic tuberculosis of lymph nodes. Diagnosis of Pseudo-leukemia. — 1. From myelogenous or pro- BIBLIOGRAPHY. 231 noimced lymphatic leukemia the diagnosis is readily accomplished by the examination of the blood, demonstrating the absence of character- istic leucocytosis. 2. From less pronounced cases of lymphatic leukemia the diagnosis may not in every instance be possible. Usually the excess of leuco- cytes in pseudo-leukemia affects the polynuclear type, but as the histo- logical structure of the nodes approaches sarcoma, tlie lymphocytes may be greatly increased. 3. From tuberculosis of lymph nodes the diagnosis may require microscopical examination of the nodes. If tuberculous, these may show (a) distinct tuberculous lesions, miliary tubercles and cheesy areas ; (6) small areas of hyaline material associated with demonstrable tubercle bacilli ; (c) lymphoid hyperplasia, without any trace of distinct inflammatory changes and without bacilli in demonstrable numbers. Such nodes should be tested by inoculation. 4. Non-tuberculous inflammatory hyperplasia of lymph nodes usu- ally causes irregular fever, polynuclear leucocytosis, and suppuration of the nodes. It is quite possible that some pyogenic infections of lymph nodes may run the course of subacute pseudo-leukemia. 5. The splenic form of Hodgkin's disease, or " splenic anemia/' usu- ally follows the type of chronic pseudo-leukemia with moderate changes in the blood. The blood in splenic anemia may be indistinguishable from that of old malarial cachexia, and the latter condition has been said to pass into the former. (Gowers.) The writer has found the clinical condition of splenic anemia, with blood changes indistinguish- able from pseudo-leukemia, proven at autopsy to be associated with : (a) Syphilis of the spleen (gummata). (b) Large round-celled sarcoma of spleen. (c) Chronic splenitis of the type of epiilieliome primitive of Gaucher. id) Cellular and fibrous stages of the ordinary lesion of the spleen in pseudo-leukemia. Bibliography. Pseudo-Leukemia. Ashanahy. Zieglei-'s Beitrage, 1888, 111., p. 411. Banti. De 1' Anemia splenica, Florence, 1882. Barbier. Gaz. hebdom., 1894, No. 5. Berthenson. St. Petersburg med. Woch., 1879, p. 101. Biesiadecki. Wien. med. Jahrb., 1876. Billroth. Beitrage z. path. Hist., 1858, p. 158. Birch-Hirschfeld." ^ Lehrbuch d. Path., 1895, p. 191. 2 Ziemssen's Handb. , Bd. 13, p. 100. Bonfils. Soc. Med. d'Observ., Paris, 1856. Brentano, Tangl. Dent. med. Woch., 1891, p. 588. BruM. Archiv gen. de Med., 1891, I., II. Chvostek. Cited by Stengel. Cohnheim. Virchow's Archiv, Bd. 33, p. 451. Combemale. Eevue de Med., 1892, p. 540. Comil, Ranvier. Man. d. Hist. Path., p. 251. Courmont. Archiv de Med. Exper., 1892, No. 1, p. 115. Craiaie. Path. Anat., 1828, p. 250. Deldfield. N. Y. Med. Eecord, 1887, I., 425. Bemanye. Th^se de Paris, 1874. 232 PSEUDO-LEUKEMIA. Dreschfekl. Deut. med. Woch., 1891, p. 1175. Dyrenfuth. Diss. Breslau, 1882, cited by Grawitz. Eberl'h. Virchow's Archiv, Bd. 49, p. 63. Ebstein. ^ Berl. klin. Woch., 1887, No. 31. ^ Deut. Archiv klin. Med., Bd. 44, p. 389. Eisenmenger. Wien. klin. Woch., 1895, p. 505. Falkenthal. Diss. Halle, 1884, cited by Ebstein. Fleischer, Penzoldt. Deut. Archiv klin. Med., Bd. 26, p. 368. Gaucher. These de Paris, 1882, Gilbert. Traite de Med. (Charcot), II., p. 528. Gillot. Th^se de Paris, 1869. Gilly. Th^se de Paris, 1886. Goldmann. Cent. f. Path., 1892, p. 665. Gowers. Pseudo-leukemia (Reynolds' System of Medicine). Griesinger. Berl. klin. Woch., 1866, p. 212. Guillermet. These, Lyon, 1890. Hochsinger, Schif. Viertelj. f. Dermat. u. Svph., 1877. Hodgkin. Med". Chir. Trans., 1832, p. 69. Jawein. Berl. klin. Woch., 1897, No. 33. Joseph. Deut. med. Woch., 1889, No. 46. Kanter. Cent. f. Path., 1894, p. 299. Kaposi. Oester. Med. Jahrb., 1885, p. 129. Kelsch. Bull, de Soc. Anat. , 1873. Kussler. Jahrb. d. Wien. Krankenanstalten, 1893, p. 565. Kundrat. Wien. klin. Woch., 1893, pp. 211, 234. Laache. Die Anemie, 1883. Lannois. Lyon Med., 1890, No. 34. Lucke. Virchow's Archiv, Bd. 35, p. 524. Luzet. Th^se de Paris, 1891. Monod, Terillon. Archiv gen. de Med., 1879, II., 34, 325. Hosier. Virchow's Archiv, Bd. 114, p. 461. Muller. Berl. klin. Woch., 1861. Murchison. Trans. Path. Soc. London, 1870, p. 372. Pel. Berl. klin. Woch., 1885, No. 1 ; 1887, No. 35. Pepper. Amer. Jour. Med. Sci., 1875, Vol. 70, p. 313. Penin. Bull, de Soc. Anat., 1861, p. 247. Ponfick. ^ Virchow's Archiv, Bd. 50, p. 550. ^Qtgj by Stengel. Pos'seli. Wien. klin. Woch., 1895, p. 407. Potain. Semaine Med., 1887. Eamond, Picou. Archiv de Med. Exper., 1896, p. 168. Eenvers. Deut. med. Woch., 1888, No. 37. Rothe. Inaug. Diss. Berlin, 1880. Runeberg. Deut. Archiv klin. Med., Bd. 33. Sabrazes. Soc. Anat. et Physiol, de Bordeaux, Feb. 8, 1892. Scheby-Buch. Deut. Archiv klin. Med., Bd. 17. Schulz. Archiv f. Heilkunde, 1874, p. 200. Sehmtiziger. Archiv f. Heilk., 1876, p. 279. Sdokb, Carta. Gaz. d. Osped., No. 5, 1894. Senator. Berl. klin. Woch., 1882, p. 533. Sevestre. Prog. Med., 1877, No. 34. Sharp. Jour, of Anat. and Phvsiol., Vol. 30, p. 59. Stengel. 20th Cent. Pract., Vol. II., pp. 445. Strumpel. Archiv f. Heilk., 1876, Bd. 17, p. 547. Trousseau. Clinique Med. Velpeau. Lecons Orales de Clinique, T. III. VerdeUi. Archiv Ital. de klin. Med., 1894, T. 32, 4, p. 595. Virchow. Krank. Geschwulste, II. Waetzoldt. Cent. f. inn. Med., 1890, No. 45. Wagner. Deut. Archiv klin. Med., Bd. 38. Westphal. Deut. Archiv klin. Med., Bd. 51, p. 83. Wilkes. Guy's Hosp. Rep., 1856, p. 114. Woillez. Soc. med. d. Hop., 1856. Wunderlich. Archiv f. Physiol. Heilk., 1858, p. 123. 1 CHAPTER X. ANEMIA INFANTUM PSEUDO-LEUKEMICA. SPLENECTOMY. ANEMIA INFANTUM (V. JAKSCH). Historical. — In 1889-90 von Jaksch described a form of infantile anemia clinically resembling leukemia but failing at autopsy to show the visceral lesions of leukemia. The disease was said to be characterized by grave anemia, high and persistent leucocytosis, marked enlargement of the spleen, slight swell- ing of the liver and occasionally of the lymph nodes, and was to be distinguished from leukemia by the disproportionate size of the spleen as compared with that of the liver, by the more moderate leucocytosis, by the more favorable prognosis, and by the absence of leukemic infil- tration of the viscera. Peculiar forms of grave anemia in infants had previously been de- scribed by Italian observers, Cardarelli, Somma and Fede, under the term " infective splenic anemia." Mosler and Senator also had long before recognized and described numerous cases of infantile anemia which they placed in an intermediate position between leukemia and pseudo-leukemia. Shortly after v. Jaksch's first reference to this condition Hayem de- scribed a case in which the blood contained numerous nucleated red cells, many showing mitotic figures. The leucocytes were chiefly mono- nuclear, but the eosins were increased. In Luzet's case nucleated red cells were not v^ery numerous, but the eosins were increased and some mitotic leucocytes were seen. Thereafter, reports of cases multiplied rapidly so that in 1892 Monti and Berggrun were able to collect 16 cases, not including those of Hausse and Loos, and added four of their own. In recent years the knowledge of the condition has not been greatly extended, and opinions regarding its significance are still at variance. Etiology. — The typical condition is usually limited to infants and children between one-half and four years of age. The majority of the reported cases have occurred in girls. Rachitis is of undoubted im- portance in the disease since it was prominent in 16 of the 20 cases of Monti and Berggrun. In the cases of severe rachitis reported by Hock and Schlesinger ^ the changes in the blood resembled those of v. Jaksch's anemia, even when the spleen was not enlarged. Syphilis, chronic in- testinal catarrh, and chronic tuberculosis were found in other cases, col- lected by Fischl. An infectious origin was held for very similar cases as early as 234 ANE3nA INFANTUM PSEUDO-LEUKEMICA. 1880 by Cardarelli, and in 1887 by Somma and Fede. Later Mya and Trambusti, Toeplitz, and Gianturco and Pianese, again maintained the infectious nature of the malady. Kecently Lowit reports finding the " hcmameba leukemice magna " in one case. Pathological Anatomy. — The changes in the viscera have been de- scribed principally by v. Jaksch (Eppinger), Luzet, Baginsky, Aude- oud, and Rotch. The spleen is much enlarged, and usually rather firm. Histologic- ally the changes are those of simple hyperplasia of all elements, while the sinuses contain an excessive number of leucocytes. Luzet found some mitotic normoblasts, but in his cases nucleated red cells were not numerous in the blood. Baginsky found many eosinophile cells in the spleen. Audeoud described extensive proliferative changes in the splenic follicles and pulp, gorging of sinuses with leucocytes, and occa- sional extravasations of blood. The liver, in the majority of cases, has been found moderately en- larged, usually less so than the spleen, and of uniform color and normal consistence. Luzet could not verify the claim that the liver is less af- fected than in leukemia, while in a case of the writer's the liver was quite as large as in fatal cases of leukemia in young children. Histo- logically there is an absence of leukemic infiltration, but Luzet found between the liver cords a considerable number of large cells (15-25 /j. in diameter) which he regarded as progenitors of the red blood corpuscles. In a case of the writer's, examined in 1896, within the liver capil- laries there were small collections of nucleated red cells and leucocytes, of which some of both were found in mitotic division, but the charac- ters of leukemic infiltration were entirely wanting. The infiltration with these groups of small cells gave much the same appearance as the late fetal liver. In the fetal liver, however, these foci are composed almost exclusively of nucleated red cells. In one of Rotch's cases also a similar condition was described. The lymph nodes were moderately enlarged in 12 of 20 cases, but in no degree comparable to the changes of leukemia. The marrow was described by Luzet as diflFusely reddened and moist, and as showing evidence of excessive multiplication of red cells. The Changes in the Blood. — The red cells are always markedly diminished, v. Jaksch's case, in which they numbered 820,000, showed exceptionally severe anemia, the majority of cases having 1.5- 3.5 million red cells. The ordinary changes of grave secondary anemia are to be noted in the red cells. Alt and Weiss found poikilocytosis very prominent in their cases, but this condition, together with mega- locytes and microcytes deficient in Hb, are common to other anemias. An excessive abundance of nucleated red celes has been shown to be one of the most characteristic features of the blood of this condition, in which they may be even more numerous than in leukemia. Yet they are not always present, in which case the disease becomes difficult to distinguish from pseudo-leukemia. Luzet, Alt, and Weiss, and others, have noticed that an unusual number of these cells are SIGNIFICANCE OF " V. JAKSCH'S ANEMIA." 235 found in mitotic division. In well-marked cases the nucleated red cells may all be of normal size for this age, but in the graver stages of anemia megaloblasts reach a considerable proportion or a majority. The usual degenerative changes in the red cells of severe anemia have been described in several of these cases. The leucocytosis is an important characteristic. Usually the white cells number between 20,000 and 50,000, but in an apparently genuine case of Baginsky's they varied between 122,000 and 40,000. In more typical cases the uniform persistence of the leucocytosis without marked variations has been somewhat peculiar. Some cases have recovered with diminishing leucocytosis. In most instances in wliich differential estimates are reported, the mononuclear cells have been slightly in the majority. In apparently genuine cases (Rotch, Hock and Schlesinger ") the mononuclear cells, large and small, formed 80 percent, 84 percent, and 75 percent, of a greatly increased number of leucocytes (in one case 116,000). In other cases the poli/nuclear cells have been the more numerous. The proportion of eosinophile cells varies. Although considerably increased, up to 6 percent (Vickery), they do not reach either the proportions or numbers seen in leukemia. Myelocytes have been noted by Klein and in the more recent reports, but only in small numbers. In one of Vickery's cases 10 percent of myelocytes are reported among 22,000 white cells. Usually these cells are not so abundant as to suggest leu- kemia. Great variations in the size of the neutrophile leucocytes have been described, but cannot be regarded as a special character. The great variety of degenerative changes seen in all the blood cells, especially in the leucocytes, has been very fully described and depicted by Engel. Significance of " v. Jaksch's Anemia." — The attempt to determine the true nature of the cases described under this term must be guided by the known characteristics of the blood of infants. The more im- portant of these characters are: (a) The relative lymphocytosis ; (b) the more active leucocytosis excited by chemotactic influences ; (c) the tendency to enlargement of the spleen in all chronic anemias of infants ; (d) the hyperemia of the mar row associated with rachitis ; {e) the facility with which blood formation in infants partially irgresses toward the embryonal type. With these disturbing factors in view, it would appear that some of the reported cases of anemia pseudo-leukcniica could better be classed as pernicious anemia, by referring the peculiar leucocytosis to the spe- cial tendencies of infants' blood. Such a case is that of v. Jaksch in which there were 820,000 red cells and 54,660 leucocytes of undeter- mined varieties. Other cases, like that of Rotch, with 1,311,500 red cells, 116,500 leucocytes, 80 percent lymphocytes, enlarged lympli nodes, no autopsy, etc., might perhaps better stand as a case of lymphatic leukemia with unusual proportion of nucleated red cells. Yet the present tendency is to accept too uncertain evidence in the diagnosis of lymphatic leu- kemia, and the writer agrees with Rotch in his classification of this 230 ANEMIA INFANTUM PSEUDO-LEUKEMICA. case. One of Senator's cases also showed enlargement of lymph nodes and a ratio of 1-10 between leucocytes and red cells. Likewise Vick- ery's case, with 22,000 leucocytes, 35.8 percent lymphocytes, 10 per- cent myelocytes, and 6.2 percent eosins, and no autopsy, may have been an early stage of myelogenous leukemia, but myelemia in infants, even if acute, ought to give more than 22,000 leucocytes and 10 per- cent myelocytes, while if chronic its characters are usually unequivocal. The real nature of these cases must remain obscure in the absence of repeated and very complete examinations of the blood, and microscopical study of the viscera. Finally, it is*possible that many of the cases are to be regarded as grave anemia with leueocytosis of peculiar character. This view is sup- ported by the occasional record of recoveries and of gradual transforma- tions into grave anemia with ordinary leueocytosis. (Monti, Berggrun.) Nevertheless, in spite of the resemblance of many cases to pernicious anemia, leukemia, or grave anemia with leueocytosis, there appears to be, in the clinical history, in the morphology of the blood, and especially in the condition of the viscera, sufficient ground on which to separate, at least for the present, certain peculiar forms of chronic anemia in children from any of the above conditions. In the clinical history these signs are, chiefly, the chronic course, the frequent association with rachitis, syphilis, or chronic intestinal catarrh, the pronounced enlargement of the spleen, and the usual absence of distinct features of pernicious anemia, leukemia, or of a cause of leu- eocytosis. In the blood the changes are rarely such as to cause confu- sion with pernicious anemia, though the condition appears at times to precede pernicious anemia. While the acute cases with many lym- phocytes or myelocytes may be indistinguishable from leukemia (one case of Luzet's (XXV.) having passed slowly into leukemia), in other more numerous instances the peculiar condition of the blood persists unchanged for months and shows no tendency to declare itself as leu- kemia. In the average case of grave secondary anemia the leueocytosis fails to show the peculiar characters of this condition. The exces- sive numbers of nucleated red cells and the very active multiplication of these cells and of leucocytes, as indicated by the large proportion of mitotic nuclei, and the abundance of myelocytes and eosinophile cells, or of large lymphocytes, are signs not necessarily connected with grave anemia in children, but here constitute a striking and distinctive blood- picture. Considering the changes in the viscera, it is found that while the other features of the disease simulate pernicious anemia or leukemia, there are neither the leukemic infiltrations and hyperplasias, nor the exces- sive iron-content of the liver, or the megaloblastic changes in the mar- row, of pernicious anemia. The Avriter is inclined to believe that the peculiar groups of mitotic red cells and leucocytes found in the hepatic capillaries of the single case which he has had opportunity to examine may indicate the essen- tial nature of the condition. The presence of these cells shows that BIBLIOGRAPHY. 237 the liver had resumed or retained its fetal function of blood cell for- mation. Similar groups of cells resembling leucocytes have been de- scribed by Rotch and similarly interpreted, but without mention of Hb-content or mitotic figures in the cells. Luzet also mentioned the presence of peculiar groups of cells in the liver. If these observa- tions can be verified, a specific anatomical condition will have been established for v. Jaksch's anemia, viz, the resumption by the liver under the influence of grave anemia or toxemia of its fetal function of de- veloping red and possibly also white blood cells. The conditions would then differ from pernicious anemia in the absence of extreme megaloblastic changes in the marrow, and in the extension, rather, of normoblastic red cell formation beyond the marrow and into the next most available tissue, which is the liver. On this anatomical basis it would still be impossible to explain the frequent presence of many lymphocytes or myelocytes which, however, may be referred to the co- incident affection of lymphoid tissues, especially of the marrow, a tis- sue which in this disease has as yet received inadequate attention. In a case observed by the writer in 1896 the patient was a markedly rachitic infant of 20 months. The spleen extended 3 inches below the costal border. The liver was considerably enlarged. The lymph nodes were not enlarged. The red cells numbered 1,820,000, about 10 percent of which were pale megalocytes. Nucleated red cells were extremely abundant and usually of normal size. Hb not taken. The leueocj'tes numbered 48,000, of which 22 percent were small lymphocytes, 34 percent large lymphocytes, 3 percent myelocytes, 38 percent polynuclear, 3 percent eosins. The spleen was the seat of uniform hyperplasia of pulp cells, while the follicles were much reduced in size and number. Throughout the pulp there were numerous small collections of small cells with compact or mitotic nuclei. Many of these were nucleated red cells. Mitotic leucocytes were not identified. There was beginning increase of connective tissue and the sinuses were often obliterated. Eosins were very abundant. The livei- ex- hibited numerous small intracapillary foci of cells, many of which contained mitotic nuclei. Of these the majority were nucleated red cells, others were much larger and granular. Occasionally they appeai'ed fused together in one cell mass, resembling the multinuclear masses of Luzet. These groups of cells resembled the foci of nucleated red cells of the embryonal liver, but were much less numerous. The liver cells were not fatty. The marrow of the femoral shaft was cellular throughout, no fat cells being found in this situation which in normal subjects of this age is at least partly fatty. There was general hyperplasia of cellular elements, but the cells were not densely packed as in leukemia. The islands of nucleated red cells were very numerous, th?se cells being distinctly in excess. Many mitotic leucocytes with granular protoplasm were identified. The sinuses were obliterated. Eosins were not over abundant. The condition of the spleen, liver, and marrow indicated that excessive demands were being made upon the blood-forming organs, and that these demands were being met by the marrow, the liver, and probably also by the spleen. Bibliography. Anemia Infantum. Alt, WeUs. Cent. f. med. Wissen., 1892, Nos. 24-25. Amleond. Kevuc nie^d. de la Suisse Rom., 1894, p. 507. Bagimky. Arcliiv f. Kinderheilk, 1892, Bd. 13, ]>. 304. Cdrtlarelli. Atti d. priino Cong. Pediatr., Naples, 1891. 238 SPLENECTOMY. Engel. Yircliovr's Arcbiv, Bd. 135, p. 369. Fede. Atti d. priiiio Cong. Pediatr., Naples, 1891. Fischl. Zeit. f. Ileilk., 1892, Bd. 13, 277. Gianturco, Pianese. Uaz. d. Clin., Naples, 1892, III., p. 305. Hausse. Diss. Miinchen, 1890. Hayem. Gaz. des Hop., 1889, No. 30. Hock, Schlesincjer. ' Beitnige zur Kinderheilk. Wien., 1892. K'ent. f. klin. Med., 1891. r. Jaksch. Wien. klin. Wocli., 1889, Nos. 22-23. Prager med. Woch., 1890, No. 99 'Klein. Berl. klin. Woch., 1890, No. 31. Loos. Wien. klin. Woch., 1891, p. 26. Lowit. Die Leukemic als Protozoeninfec. Wiesbaden, 1900. Luzd. Archiv gen. de Med., 1891, I., p. 579. These de Paris, 1891. Monti, Berggrun. Die chron. Anemie, etc., Leipsic, 1892. Master. Cited by Senator. Mya, Trambusti. Lo Sperimentale, 1892. Botch. Pediatrics, Philadelphia, 1897, p. 359. Senator. Berl. klin. Woch., 1882, p. 533. Somma. Allg. Wien. med. Zeit., 1891, Nos. 4-11.^ Toeplitz. Jahrbuch f. Kinderheilk., Bd. 33, p. 367. Vickery. Med. News, 1897, Vol. 71, p. 731. SPLENECTOMY. The effects of spleaectoray upon man are usually the combined re- sults of severe hemorrhage, or preexisting anemia, of the loss of the function of the organ, and very often of intravenous infusion practiced immediately after the operation. How much of these effects are referable solely to the loss of the organ can better be determined by comparing the blood changes fol- lowing splenectomy in animals with those observed in the human sub- ject. Effects of Splenectomy in Animals. Hosier was one of the first to study the effects of splenectomy upon healthy dogs. He found a diminution of red cells, persisting for several months, but no changes in the leucocytes. The blood-forming function, then regarded as inherent in the spleen, he believed to have been trans- ferred to the marrow, in which tissue he found changes comparable to those of leukemia. The chief permanent chemical alteration he found to be a loss of iron. Malassez found the oligocythemia to persist only one month, while the Hb remained deficient much longer. Zesas found in rabbits a marked leucocytosis, reaching a maximum in the tenth week, when the red cells begin to diminish. The blood was restored to the normal six months after the operation. Winogradoff found that the red cells in dogs diminish for 200 days, and many megalocytes appear. Irf the second year, smaller cells reappear and the total number gradually becomes normal. In two of three dogs there was considerable leucocytosis. Diminution of specific gravity and of Hb was constantly noted. The lymph nodes of the animals were found to be smaller, and the marrow hyperemic. Gibson found the maximum diminu- tion of red cells 60 days after splenectomy (three dogs). Pronounced leucocy- tosis was observed the day after the operation. The restoration of the blood required six months. More recently Kourloflf followed the course of the leucocytosis of splenec- tomy, finding, in the fir.st year following the operation varying grades of lymphocytosis, the proportion of these cells rising from 30 up to 60 percent. The proportion of granular leucocytes fell from 40-50 percent to 20 percent, or less. The numbers of large mononuclear cells did not change, indicating SPLENECTOMY IN MAN. 239 that the spleen cannot be considered as their place of origin During the second year, a prominent alteration of the blood was a marked eosinophilia, while in this period the lymphocytes fell to the normal proportion. Emlianoff demonstrated in splenectomized dogs a slight diminution of red cells and an extreme leucocytosis marked by an initial loss of small cells, and a sudden increase of larger (polynuclear) cells. Vulpius also observed in several rabbits a moderate leucocytosis lasting not longer than nine weeks, and a loss of 20 percent in the number of red cells, which were restored within five to six weeks. It has therefore been show^n, by the above observers among others, that splenectomy in animals is followed by moderate reduction in red cells lasting from one to twelve months, by relatively greater loss of Hb more slowly restored, and in some cases by the appearance during the first year of raegalocytes. Leucocytosis follows the operation, but its extent and duration are very variable. A polynuclear leucocytosis is observed during the first days or weeks, followed by relative or absolute lymphocytosis during the first year, while during the second year distinct eosinophilia may be observed. With these changes in the blood are associated a marked cellular hyperplasia of the marrow, approaching at times that of leukemia, and often also afiFecting the lymph nodes. In the SAVollen nodes an exces- sive number of nucleated red cells have been foimd by Winogradoff, Tizzoni, Gibson, Kourloif, and Grunberg. Splenectomy in Man. The most marked changes in the blood following splenectomy in man are seen in those cases in which the organ has been removed for rupture or idiopathic enlargement. Successful cases are recorded by Czerny,^ Regnier, Hartley, McBurney, and others. In Czerny's case of idiopathic enlargement (1,500 grm.) the Hb fell to 56 percent, while the red cells were very slightly reduced and reached over 5 millions within a month, but four years later the patient, for reasons not stated, was foimd to have only 3.3 million red cells with 85 percent of Hb. The leucocytosis was very marked, reaching 70,000 within a week and persisting for at least eight weeks. In Regnier's case of rupture of spleen with severe hemorrhage, the Hb fell to 20 percent, rising gradually to 80 percent in the eighth week. The red cells, falling to 2.5 millions, rose to 4.7 millions by the eighth week. An acute leucocytosis, 25,000, appeared soon after the operation. One month later the polynuclear cells had been largely replaced by an equal number of lymphocytes. Nucleated red cells and eosins were scarce. Hartley's case (examined by the writer) was complicated by infu- sion, and possibly by preexisting malaria. The anemia on the fourth day was profound ; the leucocytosis was extreme, estimated roughly at 75,000, of which 77 percent were polynuclear. Many " splenocytes " and a moderate number of myelocytes with deficient neutrophile gran- 240 SPLENECTOMY. ules, and a few normoblasts, were present, so that the blood strikingly resembled that of leukemia. At the end of three weeks the young patient's blood had greatly improved, and the leucocytes were normal. After three months the only abnormality was the presence of a few mod- erate-sized megalocytes, and an apparent deficiency of neutrophile gran- ules. Later examinations, extending over three years, failed to show, at any time, an absolute increase of either lymphocytes or eosins, or any persistent anemia. In McBurney's case the writer found the anemia and the resemblance to leukemic blood even more pronounced. The same prompt improve- ment followed, and six months later the blood was practically normal. After splenectomy for various other general indications not includ- ing malaria and leukemia, rather variable results have been noted. In the cases collected by Vulpius, Hartley, and Litten, and reviewed by various authors, the writer finds a general resemblance of the blood changes when reported to those described after experimental splenec- tomy. The grade of anemia and the period required for the restora- tion of the blood seem to vary with the general condition of the patient and the circumstances of the operation. Dominici describes a case in a tuberculous subject in whose blood, 1 2 days after the operation, a very large number of nucleated red cells began to appear, 9,800 per ccm. having been counted at one time. After three weeks they dis- appeared. In the majority of the cases of splenectomy for wamlenng spleen and other conditions not seriously affecting the patient's health, the blood was described as normal before and after the operation, but in some the usual reduction of red cells and leucocytosis have been observed. The slight changes commonly noted are probably referable to the effects of the laparotomy, as Hartraann and Vaquez found that after every such operation there is slight anemia and leucocytosis. The malarial spleen has been excised by many surgeons with favorable results. Jonnesco, who reports a series of cases, found a prompt increase of red cells after a temporary diminution (.5-2 mil- lions) and a somewhat persistent leucocytosis (15,000-30,000). The operation affected a prompt cure of long-persistent cachexia. Similar favorable results were obtained by Hartley, Vulpius, and others. In letikemia the spleen has been extirpated in 28 cases with im- mediately fatal results in 25. One case survived 13 days, another 8 months, and a third, in which the diagnosis must be doubted (Franzo- lini, 1882), was reported cured. In the two genuine cases the opera- tion was followed by steady diminution of red cells and increase of leucocytes. (Bardenhauer, Burckhardt.) Visceral Changes Following Splenectomy. — Enlargement of lymph nodes has been observed after splenectomy in man by Czerny,^ Kocher, Lennander, and Regnier. In Regnier's case the enlargement of the nodes was associated with lymphocytosis. The marrow was examined by Regnier four weeks after the splenectomy and found very hyper- emic. There had been marked anemia in this case. BIBLIOGRAPHY. 241 Resume. — In comparatively healthy subjects, splenectomy has often been performed without affecting the blood more than does any other laparotomy. In many graver cases the loss of blood and the shock of the opera- tion give rise to a considerable grade of secondary anemia. The red cells are, in favorable cases, restored in 1-3 months, but in less favor- able cases there may be more persistent anemia. The restoration of Hb seems in some cases to fall behind the improvement in cells rather more than in most secondary anemias. The operation is usually fol- lowed by considerable polyuuclear leucocytosis (15,000—50,000), which commonly persists for 2—6 weeks, but may continue for months, in which case the polyuuclear cells may be largely replaced by lympho- cytes. Eosinophilia has been observed in a few cases during the second and third years. In traumatic cases suffering from large hemorrhages, splenectomy, especially when complicated by infusion, may lead to very profound anemia, marked by extreme loss of red cells, the presence of many very large, pale, sometimes polychromatic, and dissolving, red cells, nucleated red cells, and to a high grade of leucocytosis. Among the leucocytes there may be a considerable proportion of large, pale mono- nuclear cells and myelocytes, so that the blood resembles that of acute leu- kemia. This condition however is transitory and the blood may im- prove rapidly. Leukemia and the amyloid spleen are contraindications to splenec- toDiy. In other conditions the choice of operation may depend entirely upon the general condition of the patient. Beyond- a moderate persistent leucocytosis or lymphocytosis, and possibly a slight delay in the restoration of Hb, there are no specific effects known to follow splenectomy in man. Bibliography. Splenectomy. Bardenhauer. Deut. Zeit. f. Chir., Bd. 45, p. 181. Burckhardt. Archiv f. klin. Chir., Bd. 43, p. 446. Czerny. i Cited by Vulpius. ^ wien. raed. VVoch., 1879, p. 333. Dominki. Compt. Rend. Soc. Biol., 1898, p. 1193. Emilianoff. Archiv des Sci. Biol., St. Petersburg, Bd. II., No. 2. Fmnzolini. Gaz. med. Ital., 1882, Guil. Gibson. Jour, of Anat. and Physiol., 1886, Vol. 20, p. 674. Grunherg. Diss. Dorpat, 1891. Hartley. Med. News, 1898, Vol. 72, p. 417. Hartman, Vaquez. Compt. Rend. Soc. Biol., 1897, p. 126. Jonnesco. Archiv f. Sci. med. (Bucharest), 1897, p. 301. Koeher. Cent. f. Chir., 1889, p. 14. Kourloff. Wratsch, 1889, 1892, cited by Ehrlich, Die Anemie. Lennander. Wien. klin. Woch., 1893, No. 30. Malassez. Gaz. med. de Paris, 1878, p. 317. McBurney. N. Y. Med. Record, Vol. 53, p. 601. Mosler. Leukemia, 1872. Regnier. Berl. klin. Woch., 1893, p. 181. Tizzoni. Archiv Ital. de Biol., 1882, T. 1. Vulpius. Beitriige zur klin. Cliir., 1894, Bd. 11. Winogradoff. Cited by Laudenbach, Archiv de Physiol., 1896, p. 724. Zesas. Langenbeck's Archiv, Bd. 28, p. 157. 16 PART III. THE ACUTE INFECTIOUS DISEASES. INTRODUCTORY SECTION. THE BLOOD IN FEVER. While all observers have agreed that the blood in fever suffers a reduction in red cells, it still remains a matter of doubt whether a fe- brile process alone actually destroys red cells or only causes them to be unequally distributed in the body. The theory of unequal distribution of red cells in the acute fevers is supported by a variety of observations. Maragliano demonstrated a contraction of arterioles during the height of a febrile process, fol- lowed by dilatation during defervescence, and he was able to verify these results by watching the effects of antipyretics. Breitenstein found an excess of red cells in the livers of heated animals coincident with a deficiency in peripheral vessels, while Naunyn under the same conditions could find no evidence of destruction of red cells. Increase in specific gravity of the blood was demonstrated by Stein, with ris- ing temperature, followed by lowering of gravity in defervescence. Reinert believes that excessive loss of fluids diminishes the volume of blood during the height of the fever, while retention of fluids results from lower blood pressure and heart weakness after defervescence. Tumas believes that the blood in fever may be reduced in bulk as well as in proportion of red cells. In favor of the view that the red cells are destroyed in fever is the demonstration of an excess of potassium (Salkowski), and of hydro- bilirubin (Gerhardt, Hoppe-Seyler) in the urine. Yet Bunge has shown that the red cells which are very rich in potassium may take up and discharge large amounts of this element without themselves being destroyed. In pure types of experimental pyrexia Naunyn and others have found no solution of Hb in the blood-serum and no evidences of destruction of red cells. Werhowsky, however, exposed rabbits to a temperature of 38.5-40° C. for 2-29 days and found a steady diminu- tion, reaching 30 percent, of Hb, followed by progressive loss of red cells, moderate leucocytosis, and deposits of hemosiderin in the marrow and spleen. Similar indications were obtained long ago by Mobitz in septic animals whose blood at first showed from day to day consider- able variations in the red cells but eventually a permanent loss. These results leave no doubt of the capacity of prolonged fever to de- 244 THE ACUTE INFECTIOUS DISEASES. stroy red cells, but the length of time re(ju'ired furnishes opportunity for factors other than pyrexia to intervene. Consequently Lowit ^ concludes that the real nature of the oligocythemia in early febrile process is still doubtful. Fever is usually accompanied by marked disturbance of the coagula- tion of the blood, which has sometimes been found increased, at others diminished or entirely inhibited. From the studies of Schmidt and his pupils it has been shown that in septic fever coagulation is diminished at an early stage and increased at another later stage. The variations depend, according to Bojanus, upon changes in the quantity either of fibrinoplastic substances or of fibrin ferment, and are connected with the presence of dissolved Hb in the plasma. In non-febrile pyrexia, insolation, the coagulability of the blood is lost. Changes in Chemical Composition. — A progressive loss of albumens of the blood in febrile diseases has been noted by many observers, but it is not yet known whether the loss is referable to the pyrexia or to other associated factors. While there is good reason to believe that the less diffusible globulin should suffer less than the albumen, as Gottschalk claims to have found in some instances, yet Limbeck and Pick found that no general rule could apply to variations in the globulin of the blood in infectious diseases. A contrary conclusion also was reached by Emmerich and Tsuboi who found the globulin of the blood to diminish when rabbits are being immunized against hog cholera. Resistance of Red Cells in Fever. — A diminished resistance of the red cells and increased isotonic tension of the plasma were demonstrated by Maraffliano in severe anemia, and bv Celli and Guarnieri in various fevers, while the morphological characters which indicate this change in the red cells have been described by Maragliano, Gabritschewsky, and Grawitz. Yet diminished resistance of the red cells is not found in all fevers and may, as Limbeck ' has shown, be at times increased. While Hamburger has shown that the resistance of the red cells de- pends principally upon the osmotic tension of the plasma, Demoor refers the low resistance in fever to the presence in the blood of bac- terial toxines and of excess of CO^, while Lowit ^ connects it more closely with changes in the vessel walls. Febrile Hydremia. — A marked diminution of the albumens of the plasma has been demonstrated in many infectious diseases by Hammar- schlag, Limbeck,^ Biernacki,^ Wendelstadt, and others, and has been referred not to increase in the total bulk of water in the blood but to destruction of albumens in the febrile process. Herz believes further that the febrile hydremia is characterized also by swelling of the red cells with corresponding oligoplasmia, and Limbeck and Steindler found the volume of the serum in three healthy subjects to average 72.6 percent, while in eight acute febrile cases it fell to an average of 54.8 percent. On the other hand Pfeiffer denies that the red cells swell in febrile diseases, finding rather that they shrink, and Biernacki denies the existence of any uniform febrile hydremia. The conclusions THE BLOOD IN FEVER. 245 of each of the above observers appear to be justified by their results, and their conflicting opinions must be referred to technical errors or to particular conditions existing in the subjects of their study. Alkalescence. — Abundant sources of acid metabolic products are believed to exist in fever as a result of imperfect oxidation of albumens and the formation thereby of fatty and lactic acids, as well as in the ■development of acids as a result of bacterial growth. Numerous obser- vations, also, by Senator, v. Jaksch, Kraus, Klemperer, et ciL, have apparently shown by different methods that febrile processes are regu- larly marked by diminished alkalescence of the blood. Yet Minkowski has from the first pointed out that at the height of the fever the loss of alkalescence is not proportional to the pyrexia. Limbeck and Steindler in a large series of febrile subjects found in both blood and serum variations in alkalescence quite within normal limits. On the other hand Lowy and Richter, using a special method which they claim to be more reliable than others, find that in a certain period •of febrile processes corresponding to the stage of hypoleucocytosis the blood shows an increase and not a decrease of alkalescence. The cause of this condition is not yet explained, but Lowy's results have been verified by Lowit, Biernacki,- and Strauss. Important contributions to the subject were made by Fodor and Rigler. These observers found that the serum of rabbits infected with anthrax at first shows an increase of alkalescence, but after 24 hours a rapid and marked diminution. With rabies the alkalescence dimin- ished from the first. Between the quantity of diphtheria toxin injected and the final loss of alkalinity therefrom they found a parallel. In- jections of vaccine increased the alkalescence for 7— S days, of diph- theria antitoxin, for only 48 hours. Since the alkalinity of the serum- ash did not vary with that of the serum they concluded that the property depends on the presence of organic substances. Other im- portant studies of the alkalescence of the blood by Cantani, Calabrese, etc., have shown that the important variations in the quality of the blood are dependent less upon fever than upon other obscure chemical processes. Lowit^ concludes that the alkalescence of the blood may be increased at one thne and diminished at another period of an infectious disease, that this property is not dependent in any large measure upon the leucocytes and that its sic/nijicance is still unexplained. Action of Bacteria and their Products on the Blood. — Many of the phenomena classed among the effects of fever on the blood can be di- rectly traced in part to the action of bacteria and bacterial products which are the exciting causes of the febrile process. Bouchard has demonstrated for the bodies of Bacillus pyocyaneus, and Gley and Charrin for their filtered products, a vaso-constricting action, and therefore a tendency to concentrate the blood. An in- •creased flow of lymph and probably a similar concentrating effect upon the blood has been shown to follow the injection into animals of pep- tone and animal extracts, by Heidenhain, of tuberculin, pyocyanin, 246 THE ACUTE INFECTIOUS DISEASES. pneumotoxin, by Gartner and Romer. Rapid variations in the gravity of the blood have, in Grawitz's ^ hands, followed the injection of cul- tures of cholera, diphtheria, and pyogenic organisms. Small doses of toxines of various bacteria were found by Bianchi and Mariotti to in- crease the isotonic tension of the blood quite beyond normal limits, while large doses of many germs or even small injections of Bacillus typhosus had the opposite effect. Specific bacterial products, toxalbumius, bacterio-proteins, and pto- maines, are believed by most observers to be the active agent in the destruction of the albumens of blood and tissues in infectious diseases, acting at times without the accompaniment of fever. (Muller.) As a specific elFect of the chemotactic action of bacteria and their products upon the white blood cells there is the whole series of phe- nomena of leucocytosis, the significance of which is considered in an- other section. Conclusions Regarding the Blood in Fever. — From the foregoing brief review it will be seen that the changes in the blood in fever form an extremely complex subject, about which our knowledge is still rather fragmentary. The same phenomena have been encountered in differ- ent lines of investigation and attributed to different single factors, though probably referable to many. While some main facts are rather distinctly apparent, other questions must remain undecided until fully adequate technical methods are devised. Decrease or relative increase in the proportion of red cells, but end- ing always in a loss in their total numbers, must be accepted as accom- panying all cases of pyrexia, although requiring some time to become clearly apparent. Coagidab'ility varies in different stages of febrile diseases, but is not clearly connected with the pyrexia as such. The progressive loss of albumen of the blood is probably essentially connected with the febrile process, but occurs in increased degree when the fever is of infectious origin. Febrile hydremia is an accidental condition which may or may not occur as a result of the loss of albumens of the blood. Diminished re- sistance of red cells occurs in the majority of fevers, and depends on a variety of factors. Variations in alkalinity are frequent and consider- able in fever, but are not proportional either to the height of the tem- perature or to the toxic condition of the blood. Bibliography. Blood in Fever. Blancld, Mariotti. Wien. med. Presse, 1894, No. 36. Biernacki. iZeit. f. klin. Med., Bd. 24. 2 Ibid., Bd. .31, p. 312. Bojanus. Inaug. Diss. Dorpat, 1881. Bouchard. Lecons sur les Autointoxications, Paris, 1887. Breitenstein. Archiv f. exper. Path., Bd. 18, p. 42. Bunge. Zeit. f. Biol., Bd. 9, p. 129. Calabrese. Policlinico, 1896, fasc. 1, 2. Cantani. Cent. f. Bact., Bd. 20, p. 556. BIBLIOGRAPHY. 247 Celli, auarnierl Fort. d. Med., 1889, No. 14. Demoor. Jour. med. de Bruxelles, 1896, No. 36. Emmerich, Tsuboi. XI. Cong. f. inn. Med., 1892, p. 202. Fodor, Rigler. Cent. f. Bact., Bd. 21, p. 134. Gabritschewsky. Arcliiv f. exper. Path., Bd. 28, p. 83. Gartner, Romer. Wien. klin. Woch._, 1892, No. 2. Gerhardt. Ueber Hydrobilirubin, Diss. Berlin, 1889. Gley, Charrin. Cent. f. Bact., 1894, p. 688. Gottschalk. Zeit. f. physiol. Chem., Bd. 12, cited by Limbeck. Orawitz. 'Zeit. f. klin. Med., Bd. 22, p. 411. Hamburger. Virchow's Archiv, Bd. 140, p. 503. Hammarschlag. Zeit. f. klin. Med., Bd. 21, p. 475. Herz. Virchow's Archiv, Bd. 133. Hoppe-Seyler. Virchow's Archiv, Bd. 124, p. 30. V. Jaksch. Zeit. f. klin. Med., Bd. 13. Kraus. Zeit. f. Heilk., Bd. 10. Limbeck iKlin. Path, des Blutes, 1896, p. 158. 2 Prager med. Woch., 1893, Nos. 12-14. Limbeck, Steindler. Cent. f. inn. Med., 1895, p. 649. Lowit. ' Die Lehre V. Fieber. 2 Ibid., p. 151. 3 ibid., p. 164. Louy, Bichter. Deut. med. Woch., 1895, No. 33. Maragliano. Berl. klin. Woch., 1887, No. 43. Minkowski. Archiv f. exper. Path., Bd. 19, p. 215. Mobitz. Inaug. Diss. Dorpat, 1883, cited by Lowit. Midler. Zeit. f. klin. Med., Bd. 16, p. 496. Naunyn. Archiv f. exper. Path., Bd. 18, p. 81. Pfeiffer. Cent. f. inn. Med., 1895, No. 4. Reinerl. Zahlung d. Blutkorp, p. 174. Schmidt. Pfliiger's Archiv, Bd. 11, pp. 291, 515. Senator. Untersuch. u. d. lieberhaften. Process, Berlin, 1873. Stein. Cent. f. klin. Med., 1892, No. 23. Strauss. Zeit. f. klin. Med., 1896, Bd. 30, p. 315. Tumas. Deut. Archiv klin. Med., Bd. 41, p. 323. Wendelstadt. Zeit. f. klin. Med., Bd. 25. Werhowsky. Cong. f. inn. Med., 1895, p. 345. CHAPTER XI. PNEUMONIA. DIPHTHERIA. PNEUMONIA. The gross characters of the blood in pneumonia were accurately described by Piorry, who noted the prompt and firm clotting due to excess of fibrin, while the distinct crusta phlogMica, most marked about the seventh or eighth days, was the earliest recorded observation con- cerning the leucocytosis of this disease. Fibrin has since been shown to be abundant in nearly all cases of pneumonia, and very much in- creased in the great majority. The excess often continues beyond defervescence. In very severe cases without leucocytosis, fibrin is usu- ally deficient. (Turk.^) The total volume of blood is probably somewhat reduced in severe cases of pneumonia, owing to the concentration which occurs in fever, from cyanosis, and from the loss of cells and plasma in the exudate. Evidences of this concentration have accumulated from several sources, among which are : the increase in specific gravity during the height of the process (Monti, Berggrun), and the persistence of a high propor- tion of red cells and Hb which usually continues until the crisis, after which there is a rapid decline in the quality of the blood. Large exudations into the lungs are believed by Bollinger to reduce the quantity of blood to such an extent that oligemia may usually be noted at autopsy. In 65 percent of his cases marked diminution of blood was thus noted at the post-mortem examination, a loss which was fully explained by the extent of the exudate, which averaged over 1,000 gr. in weight. The Red Cells. — Although destruction of red cells undoubtedly occurs in the disease, as shown by the increased excretion of hydrobili- rubin, and the occasional appearance of jaundice, yet the red cells re- main high during the active febrile period, when this destruction is taking place, to diminish only when the temperature begins to fall. This result can be referred only to the concentrating effects of fever, exudation, and local vasomotor phenomena, and is to be seen in other infectious fevers. In. spite of these complicating factors the red cells usually show a slight but steady decline during the course of the fever, as shown by the reports of Sorensen, Boeckman, Halla, Tumas, Sadler, and others. Slight polycythemia (maximum 7 millions) frequently appears in the reports of Sadler, while Sorensen observed an inorease of red cells dur- ing the febrile period in one case. The decrease of red cells often LEUCOCYTES. 249 occurs suddenly with the crisis, Tumas reporting one case with a fall of 600,000 on the day of defervescence. The diminution continues for a very variable period depending upon somewhat complex factors, but commonly ceases within 10-14 days, or in mild cases much sooner. The grade of anemia established is usually not very marked. In many recorded estimates the loss of cells did not exceed the limit of error. A loss of .5—1 million is not uncommon, while a reduction of 2 million cells is recorded by several observers. Such changes in the red cells, however, must be interpreted with great caution. There was, for instance, a loss of 2 million cells in one of Sadler's cases, yet no anemia was established as the blood had been concentrated and the lowest count was 5.1 millions. Morphological changes in the red cells are usually slight. Polychro- masia is seen in the severe cases only. Nucleated red cells were present in 7 of Turk's 18 cases. These were usually normoblasts, occasionally megaloblasts. The cases Avere all very severe, but only one of the seven was fatal. A similar phenomenon may occur in other severe infections. The Hb suifers considerably more than the red cells, being almost invariably reduced after defervescence, but in the absence of complica- tions seldom falling below 65 percent. Chemistry. — The specific gravity was found to remain normal or to distinctly increase during the febrile period in 9 children studied by Monti and Berggrun.* The albumens of the blood were slightly diminished, while those of the serum were normal or increased in 7 cases reported by v. Jaksch.' The to.viciti/ of the blood serum was found much increased in pneu- monia and other infectious diseases by Albu, who locates the poisonous principle in the albumens of the serum. Leucocytes. — Leucocytosis appears in the great majority of cases of pneumonia, failing only in very mild attacks and in very severe in- fections with feeble reaction and bad prognosis, and when some pre- existing conditions have already excited leucocytosis or debilitated the system. In leukemia a complicating pneumonia reduces the leuco- cytes, while the onset of pneumonia in the course of some infectious diseases is not always traceable in the blood. Although the leucocytosis of pneumonia was noted by Virchow, Nasse, and other early writers, its closer study was begun by Sorensen in 1876, and continued by Boeckman, Halla, Tumas, Hayem, and Limbeck.^ Even Boeckman (1881) gathered from the literature that most febrile dis- eases are accompanied by levicocytosis and that typhoid fever and malaria are exceptions to the rule. Halla noted that the leucocytosis is not pro- ^ portional to the fever, having encountered three cases without leucocytosis, all fatal. Hayem and Limbeck followed the course of the leucocytosis over considerable periods, pointed out relations between the grade of leucocytosis and the severity of the disease, discussed the time and manner of its disap- pearance, and reported other fatal cases without leucocytosis. Limbeck's study was especially minute and he was able to detect a fall in 250 PNEUMONIA. DIPHTHERIA. the leucocy tosis some hours before the crisis, and to note the absence of any change during pseudo crises. He drew the important general deduction that leucocytosis precedes and is a part of the inflammatory exudate, and that non-exudative diseases are unaccompanied by leucocytosis. Rieder, in 1892, was probably the first to demonstrate that the leucocytosis of pneu- monia is not determined by the height of the fever, or the extent of the exudate, but depends upon the intensity of the infection and the degree of resistance of the subject, and this view was fully supported shortly afterward in the writer's series of cases. Since Limbeck's study, the leucocytosis of pneumonia has been a favorite field of investigation, so that an extensive literature in many languages has accumulated, and many interesting and important details have been added to the knowledge of the subject. Course of the Leucocytosis. — Leucocytosis appears very early in the course of pneumonia, simultaneously with the chill, according to Klein, preceding the exudation, according to Limbeck, and having re- peatedly been found on the first day of the disease. In one of the writer's cases 25,000 cells were counted within 4 hours after the begin- ning of the first symptom, a sharp chill, while liieder and Laehr found a marked increase within 6 hours of the chill. Theoretically the leu- cocytosis should be preceded by a period of hypoleucocytosis, but this period has never been observed clinically, except in fatal cases with prolonged hypoleucocytosis. The maximum increase is reached usually just before the crisis (Hayem, Klein, Bieganski), but has been observed on the first day of the disease, or on succeeding days. AVhen the leucocytes increase slowly, they usually diminish slowly, and the disease defervesces by lysis. Peculiar cases of marked severity in which there was no dis- tinct leucocytosis until the temperature began to fall, are reported by Bieganski and Turk. During the high febrile movement there is usually little alteration in the leucocytosis, but extension of the lesion to other lobes, or to ad- joining serous membranes, may cause irregular rises in the count. Yet in a case in which bronchial breathing passed successively up one side of the chest and down the other, the writer found rather uniform and high leucocytosis. A few hours before or after crisis the leucocytes begin to diminish rapidly, sometimes falling from a high to a normal figure witiiin twenty-four hours, after which there are commonly some slight oscil- lations. During lysis the reduction of leucocytes usually keeps pace with the temperature, but Limbeck's, Laehr's, and Billings' ' charts show some marked upward curves of the leucocytes after lysis. Such rises are most often due to complications. (Kuhnau.') In fatal cases there is often a continuous increase (Pick), but some- times the leucocytes are high at first, but steadily diminish as the pa- tient grows worse (Bieganski, Rieder), while in many fatal cases there is no leucocytosis. Hayem believed that in fatal cases without leuco- cytosis, the lesion is " less exudative," but subsequent reports have shown this view to be inapplicable to the majority of cases. Alcohol- ism, virulent infection, and old age are more important factors. In LEUCOCYTES. 251 four such cases of the writer's the lesion was of the usual type. In one of these, although the exudate was considerable, the marrow of the ribs and vertebrae failed to show the usual hyperplasia. Pseudo-crises usually have no effect upon the leucocytosis (Limbeck, Klein, Laehr, Bieganski), but Turk and others have shown that this is not an invariable rule. The grade of leucocytosis is usually considerable, and is affected by several factors. Hay em found in mild cases 6,000—12,000 leuco- cytes, in moderately severe cases 18,000-20,000, and in very severe cases a maximum of 36,- 000. In children the in- Fig. 21. crease is usually greater than in adults. (Monti, BerggruUjFelsenthal.) Se- vere uncomplicated cases often show 50,000 cells, while some writers have found purulent complica- tions with many of the high leucocytoses, especiall}' when the temperature is relatively low. (Smith.) The highest figures appear to be 115,000, recorded by Laehr, in a peculiar case with delayed resolution. The fever and the leucocy- tosis often run parallel, but there are numerous in- dividual exceptions to this general rule, the most sig- nificant of which are the fatal cases, showing norma 1 or diminished leucocytes. The extent of the exudate has considerably more in- fluence upon the leuco- cytes, as first shown by Limbeck. Although con- -Temperature. . ° Leucocytosis. trary conclusions have been course of leucocytosis in pneumonia. (After Laehr. ) reached by many writers reviewing short series of cases, it nevertheless remains true that the leuco- cytosis in general bears a rather close relation to the extent of the exu- date. In the writer's cases, especially in those which came to autopsy, this rule was readily verified, although its application to individual cases proved unsafe. Thus in 63 cases in which one lobe was afiected, the average number of leucocytes was 20,000, in 24 cases with the two lobes involved, 22,700, in 12 cases with three lobes, 25,000, in one 1 17000 "t _j_ 1 ledbo .1 II 1 11 .^ M 1 1 , 15000 \ i 1 1 "■ ^ \ 1 1 11 '• t \ n t \ ' \i t \ 1 1 / uooo „ _ T V T J i T 1 . II r~ T tj T ll 5:^ -1 , ' 1 } J 1 12000 ' n J 1 (1 J. \\ 1 1 1 h "F-^ 1 1 u 1 k \ ' ' j I L 1 -J 1 1 1 \ 1 \ \ 1 1 10000 1 1 \\ 1 1 n 1 1 jl 1 1 t) \ il V \ / A / V \, / \ \ \ \ 7000 Ut ^= ^ J 1 252 PNEUMONIA. DIPHTHERIA. case with four lobes, 27,000, and in one in which there was bronchial breathing over the entire back of chest, 32,000. In 10 cases with lesions extending to the pleura (empyema), pericardium and peritoneum, the average was 17,000, a tendency toward hypoleucocytosis beginning to appear. The degree of systemic reaction to the disease chiefly determines the grade of leucocytosis. This view, first clearly stated by Rieder, has been fully verified by many observers, and embodies the true meaning of leucocytosis in infectious diseases. In 47 cases marked by vigorous systemic reaction (temperature 105°, full pulse, sthenic condition), the writer found an average of 30,000 leucocytes ; in 27 moderate cases (temperature below 105°, symptoms less severe), the average was 20,000 ; in 27 cases with deficient reaction, 9,000, while in 12 asthenic cases the leucocytes were subnormal. The significance of hypoleucocytosis, observed in many fatal cases, has been demonstrated in various experimental infectious like those studied by Tschistowitch, who found that rabbits die without leucocytosis when inocu- lated with virulent cultures of the pueumococcus, but with attenuated cul- tures there is leucocytosis, and the animals usually recover. Types of Leucocytes in Pneumonia. — At the height of nearly all well- marked leucocytoses in pneumonia the polynuclear neutrophile cells form 80-95 percent of the cells. Turk counted as high as 96.5 per- cent of these cells, and they arc frequently above 90 percent. This high proportion may be seen when the leucocytosis is slight or absent, but is usually most marked when the white cells are very numerous. At the same time there is a marked reduction, relative or often absolute, of lymphocytes to 2—4 percent. The large mononuclear cells usually persist in considerable numbers (Turk), may sometimes be distinctly increased (Klein), and are never entirely absent (Jez). The eosinophile cells at the height of the leucocytosis are always very scanty and frequently cannot be found after very prolonged search. (Zappert, Turk.) Instead of polynuclear leucocytosis tliere may be well-marked lym- phocytosis, as in a case reported by Cabot, in which in a chikl of 6 years there were 66 percent of lymphocytes among 94,600 white cells. Stienon refers to similar cases, and the writer and others have observed such inflammatory lymphocytoses in diphtheria. During defervescence, the polynuclear cells diminish rapidly, usually to a high normal proportion, at which they are apt to remain several days. The lymphocytes at the same time increase in proportion, and often in numbers, reaching their normal figures also after a few days. The large mononuclear leucocytes are usually over-abundant at this stage, reaching 16 percent in one of Turk's cases, and constituting the " large-celled lymphocytosis " described by Klein at this period. Eosinophile cells usually reappear in scanty numbers on the day before defervescence, or rarely a day earlier. (Turk.) Distinct post-critical eosinophilia occurs in a moderate proportion of cases (Zappert, Bie- ganski), but not in all (Turk). PLATE X. Fig. 1. ;r : fe: 4Y \ • - • • • Mast Cells. (Ehrlich's Dahlia-stain.) Fig. 2. ^•^ ^ Glycogenic Degeneration of Leucocytes (Pneunnoniai. (Iodine in Mucilage of Acacia.) LEUCOCYTES. 253 In severe cases without leucocytosis, the polynuclear cells are usually relatively increased (Rieder), but may be normal (Billings). Very often in such cases there is a moderate leucocytosis during or after defervescence. (Laehr.) Degenerativ,e Changes in Leucocytes in Pneumonia. — Klein described numerous "leucocyte-shadows" in severe pneumonia. They are seen in other infectious diseases, especially diphtheria, and are to be classed with the degenerative changes in tissue cells which mark the Fig. 22. i'-^ 1.3 1 4 I's ije I'r 1 8 19 20 21 22 ., i * 1 / ' \ I ' / \ \ / \ ; » 1 ' \ \ 1 » \ 1 1 ' 1 1 1 !. l-\ __ ■*■ 20000 \i j\ 1 1l \ 1 \ ^ 1 / Y \ / / 1 / / K 1 \ n — , 1 A / ^ / 1 / h 1/ w. \i '/ \ t 1/ \ l\' n" \ j s--- y'X / \ 1 /' . ] I \/i \ / y Temperature. Leucocytosis. Precritical leucocytosis in pneumonia. (After Limbeck.) status infectiosus. Glycogen granules in the leucocytes have been found in considerable abundance by Livierato and others, usually in propor- tion to the severity of the disease and the height of the leucocytosis. Myelocytes Avere found by Turk in nearly all cases. They were most abundant about the time of defervescence, and once reached a propor- tion of 11.9 percent among 8,800 cells. Peculiar lymphocytes of large or small size, with single or double nuclei, staining densely and with a brownish tinge by the triacid mix- 254 PNEUMONIA. DIPHTHERIA. ture, are described by Turk as of frequent occurrence, and as resulting from abnormal stimulation of blood-forming tissues. Applications in Diagnosis and Prognosis. — The very numerous situa- tions in which the examination of the blood is of great value in the diagnosis of pneumonia have been pointed out by numerous writers. It is especially in such acute uncomplicated infections that the results of examination of the blood are to be relied upon, but even here a word of warning is needed against too implicit confidence in this or any other isolated clinical sign, ichUe it is especial/ 1/ iinicise to transfer to other fields in blood analysis the rather positive rules ichich may usually be applied to the changes in the blood in pneumonia. Barring mild cases, and those fatal ones in which there is no increase of white cells, leucocytosis is an invariable accompaniment of pneu- monia, and its absence is, therefore, very strong negative evidence against pneumonia. When present, leucocytosis excludes, with somewhat less certainty, a considerable class of diseases which are not usually accompanied by leucocytosis. These are principally typhoid and typhus fevers, ma- laria, acute tuberculous pneumonia, uncomplicated influenza, and the catarrhal pneumonia of influenza. (Rieder.) In prognosis a slight leucocytosis indicates a mild case, while in se- vere cases a low proportion of leucocytes is an extremely unfavorable prognostic sign. This fact was first pointed out by Kikodse and v. Jaksch, although such cases had previously been reported by Halla and Hayem. Yet absence of leucocytosis appears not to be so unfavor- able as was at first supposed. Out of 57 severe cases without leuco- cytosis, reported by Halla, Hayem,^ Laehr, Ewing, Sadler, Bieganski, Zappert, Turk, Billings, Cabot, only 44 died, although all of them were unusually severe cases. The persistence of leucocytosis may be of value in distinguishing pseudo-crises, but Bieganski, Turk, and others have seen the leuco- cytes fall during a pseudo-crisis. Continuous or increasing leucocytosis after defervescence often indi- cates a relapse or complication. The reappearance of eosinophile cells indicates that the lesion has passed its acme. Bacteriology of the Blood in Pneumonia. — That infection with the pneumococcus may often take the form of a general blood infection is indicated by the clinical course of virulent infections in man and animals and by the frequent occurrence of the micro-organism in the blood of the cadaver. Similar indications are found in the report by Netter and Levy of finding the pneumococcus in the blood of a dead fetus whose mother had died of septicemia, and by the discovery of the same germ in the milk of a nursing woman who was attacked by pneu- monia. (Bozolo.) Numerous bacteriological studies of the blood during life have shown also that in a moderate proportion of severe eases the pneumococcus may be obtained in culture by the usual procedure. DIPHTHERIA. 255 Sittmann obtained the largest proportion of positive results, 6 out of 16 cases examined, four of Avhich were fatal and gave signs during life of gen- eral septicemia. Similar cases of pneumococcus septicemia are reported hy Belfanti, Netter, Marchiafava and Bignami,Goldscheider, Grawitz, Bacciochi, and others, and it has become evident that in this group of infections the pneumococcus is usuallj' to be found in the blood during life. Other success- ful results were obtained only a few hours before death from pneumonia, and the presence of the pneumococcus in the blood must be regarded as an ante- mortem invasion. (Boulay.) Ill pure cases of lobar pneumonia, cultures made a considerable time before death have shown that the pneumococcus is rarely present in the blood under such circumstances. Thus Kraus and Kuhnau ^ had only two positive results in 21 cases, and both were fatal, while most of the abov^e writers have reported many more negative than positive results. Kohn was more fortunate, finding the pneumococcus in 9 out of 32 cases. Seven of the nine were fatal. James and Tuttle had negative results in 12 cases, 10 of which were fatal. From these studies, it may be concluded that when pneumonia leads to metastatic inflammations, the pneumococcus is frequently found in the blood ; that in some cases the blood is invaded just before death ; and that in uncomplicated pneumonia the pneumococcus is rarely found in the blood during the progress of the lesion and that when it is pres- ent, the disease is usually fatal. DIPHTHERIA. Red Cells. — The concentration of the blood demonstrated to a slight degree in pneumonia is even more distinct in the highly febrile stages of diphtheria. Grawitz refers the tendency to a lymphogogic action of the diphtheria toxin which he has observed clinically and experimen- tally to cause an increase in the specific gravity of the blood. Although Bouchut foimd an average of 4.3 million cells in 93 cases, a distinct re- duction, yet many of his observations were made late in the disease, and individual cases often showed slight polycythemia. Cuffer re- ported extreme polycythemia, 7.2-7.8 millions in three cases. Later observers, especially Morse and Billings, have closely followed the changes in the red cells, finding high normal proportions, 5.1-5.6 millions during the first week. In the second week Morse found even higher numbers, once 6.8 millions, while Billings, although finding an average loss of half a million cells, encountered some cases with distinct polycythemia. With the fall in temperature slight anemia appears in many cases, but not in all. Bouchut and Billings observed losses of 2 million cells in several cases. Morphological changes in the red cells are not marked, but when anemia develops there is slight polychromasia and the usual deficiency of Hb. Nucleated red cells were noted by Engel. The Hb is slightly diminished in the average case, but may be considerably reduced in cases which have lost many red cells. Billings found an average loss of 10 percent in patients not receiving antitoxine. 256 PNEUMONIA. DIPHTHERIA. The Leucocytes. — The early observers, Bouchut, Cuffer, Pee, Halla, Limbeck, and Riecler, fully established the presence of leucocytosis in diphtheria, but only Bouchut attempted to deduce any rules gov- erning its occurrence. He observed an increase in all cases, usually proportionate to the severity of the disease, but in septicemic cases the increase was less marked and the prognosis fatal. In view of the recent observations of many myelocytes or an extreme proportion of lymphocytes in diphtheria, Bouchut's statement that the blood of diph- theria may be " leukemic " is not far from the truth. More detailed studies were contributed later by Gabritschewsky, Morse, Ewing,'' Billings,^ Engel, Schlesinger, and File. The leucocytosis of diphtheria resembles in most respects that of pneumonia. It begins very early when the disease develops suddenly, and reaches its highest point within one or two days, or begins more slowly and increases steadily for several days, or until death. It probably begins earlier and reaches its acme sooner in more refractory individuals, with whom tiie prognosis is comparatively good. The writer observed two cases loithout leucocytosis until the fourth and sixth days. In one of these the condition was probably referable to prolonged toxic hypoleucocytosis, and the patient died later. Not in- frequently the white cells begin to rise only after defervescence, while long persistent leucocytosis is also of rather common occurrence. In favorable cases the leucocytes usually diminish steadily after the first few days. In fafal case.'< the leucocytes may steadily increase or may steadily decrease during the days before the fatal issue, or there may be no leucocytosis. The height of thp: leucocytosis is usually considerable, 25,- 000-30,000 cells being frequently present in severe cases. The ma- jority of fatal cases at some period show between 25,000-o0,000 cells, hence the conclusion of Gabritschewsky that a high and progressive leucocytosis is a bad prognostic sign. Mild attacks, especially in adults, may not show any increase. The writer found 72,000 cells in one case with lymphocytosis, while in a peculiar case marked by hemor- rhagic eruption Felsenthal counted 148,000 — possibly an agonal hyperleucocytosis. The grade of leucocytosis is little affected by the fever, but in many cases is proportional to the extent and depth of the membrane. It may be distinctly affected by pneumonia, but the writer failed to find any uniform lymphocytosis when the cervical nodes were greatly enlarged, nor did the character of the infection, whether pure or mixed, appear to influence the leucocytes, (Cf. File.) In general, the leucocytosis of diphtheria is less marked than that of pneumonia with equal temperature, and less than that of non-diph- theritic angina of equal extent. (File.) Types of Leucocytes in Diphtheria. — With well-marked leucocytosis the polynuclear neutrophile cells are usually much increased in num- bers and proportions, but they very seldom reach the very high ratios seen in pneumonia, probably on account of the greater tendency of children to lymphocytosis. Lympliocytes, probably for the same rea- EFFECTS OF ANTITOXINE UPON BLOOD IN DIPHTHERIA. 257 son, are relatively more abundant. A striking lymphocytosis occurred at the height of the disease in two of the writer's cases, in one of which the lymphocytes, large and small, formed 60 percent of 72,000 leu- cocytes, and in another, 62 percent of 22,500 cells. In neither was there marked enlargement of cervical lymph nodes. Of quite diifer- ent significance is the relative lymphocytosis without increase in total numbers which may be seen in the earlier stages of the disease, and was noted by Billings in a fatal case without leucocytosis, and by Rieder and File during convalescence. Eosinophile cells are usually reduced in proportion? ,ften persist in low normal ratio, and fre- quently cannot be fon . at all. They are relatively more numerous than in pneumonia. Engel made the interesting observation that myelocytes, both neutro- phile and eosinophile, are frequently present in the blood of diphtheria, and that when these cells form over 2 percent, the prognosis is very unfavorable. In cases which recovered the myelocytes never formed over 1.5 percent, while in 7 fatal cases they ran from 3.6 percent to 14.6 percent. It should be added that the numbers of myelocytes varied greatly from day to day, and that 8 of 32 patients died with few or no myelocytes in the blood. Degenerative changes in the leucocytes in diphtheria are often well marked. The " leucocyte-shadows " of Klein are seen in all severe infections and may become very numerous in fatal septic cases. Even before the cells become fragmented many of them, principally the poly nuclear forms, show deficiency of chromatin and of neutrophile granules. The writer, examining the blood in .6-percent salt solution tinged with gentian violet, found a striking deficiency of chromatin in a case in which the leucocytosis disappeared before death. In the other septic and fatal cases also it was plainly apparent, although not demonstrable in less severe infections. This change in the leucocytes, which may be seen in other infectious diseases, and especially in acute leukemia, must be classed with the general cellular degenerations which mark the status infectiosus. Its significance in diphtheria has been fully discussed by Gabritschewsky, while Fil6 believes that these pale staining cells are necrotic. The writer believes that, with careful technique, variations in the staining capacity of leucocytes may be made of value in prognosis. An increased acidophile tendency of the neutrophile granules ap- peared to be demonstrable in some of the writer's fatal cases. Kan- thack has also noted a similar change in the staining tendency of the pseudo-eosinophile granules in rabbits which had received small injec- tions of microbic poisons. Effects of Antitoxine Upon the Blood in Diphtheria. — The red cells show no distinct or uniform effects from the injections of anti- toxine, although in some healthy subjects there is a moderate reduction lasting a few hours. On the other hand the use of antitoxine, by limiting the progress of the infection, tends to prevent further destruc- tion of blood cells. 17 258 PNEUMONIA. DIPHTHERIA. Within half an hour after the injection of antitoxine, the leucocytes, especially the polynuclear forms, if previously abundant, show a marked diminution, and in most cases, although the leucocytosis returns after 24-48 hours, it seldom reaches its previous grade. The writer noted the reduction of leucocytes after antitoxine in all but two fatal cases, while Schlesinger found it in all of his examina- tions. It has also been noted by File. In the writer's cases the loss was from 4,000 to 15,000, but Schlesinger observed a loss of 20,000 within seven hours after the injection. In some fatal cases the use of antitoxine is followed immediately by distinct hyperleucocytosis. (The writer, Gundobin.) The most favorable cases show a steady diminu- tion of white cells after the injection, but in many fatal cases this re- duction is interrupted after the third to fourth day. In critical cases with many pale staining leucocytes the writer found an improvement in the staining quality of many polynuclear cells shortly after the in- jection of antitoxine. In some fatal cases this change could not be detected. CHAPTER XII. THE EXANTHEMATA. VARIOLA. Red Cells. — During the febrile period Hayem found very little change in the red cells, while in the stage of suppuration polycythemia was sometimes to be noted. Immediately after defervescence, when the destruction of blood cells through the septic process was no longer masked by the concentrating influence of fever, the red cells showed a sudden reduction. In a case of confluent smallpox there was a loss of 2 million cells. In a fatal case before death there were 4.6 million cells, while in another, markedly hemorrhagic, there were only 2 millions. Pick, from his study of 42 cases, found very little change in the red cells throughout the entire course of the disease. He failed to find any severe anemia, although many of his cases were hemorrhagic and confluent. In severe cases which recovered the red cells remained normal before the eruption, almost invariably increased during the suppurative stage, even in hemorrhagic cases, and after defervescence or before death were rarely found much below their original figure. The usual condition of the blood after smallpox appears to be, there- fore, a mild form of chlorotic anemia, while under some rather unusual conditions this anemia may become severe. Leucocytes. — Although leucocytosis was demonstrated in the erup- tive stage of variola by Brouardel, Hayem, Halla, and Pee, the de- tailed knowledge of the subject has been contributed in a remarkable study by R. Pick. Mild cases of varioloid with scanty eruption ran their course, under Pick's observation, with normal or subnormal numbers of leucocytes, but when suppuration occurred, the leucocytes were increased, reaching in one case 20,000. Variola, also, even in high febrile periods of the early stages of the most severe cases, failed to excite leucocytosis, up to the day of death, showing that this infection when uncomplicated does not tend to excite leu- cocytosis. Secondary infection with pyogenic bacteria, however, in- creased the leucocytes after the usual manner, and in nearly all of Pick's cases there was well-marked leucocytosis beginning with the appearance of the vesicles, increasing as the exudate became purulent and reaching its height when suppuration became most active. In cases which recovered the leucocytes returned slowly to the nor- mal as suppuration subsided. Abscesses frequently interrupted recov- 260 THE EXANTHEMATA. ery and caused marked exacerbations of the leucocytosis. In seven of eight cases dying from pneumonia there was a rapid and steady de- crease of leucocytes although suppuration continued. The grade of leucocytosis was usually proportionate to the severity of the septic process. Biological Examinations of the Blood in Variola. — In 1887 G. Pfeiffer called attention to the presence in vaccine lymph of small unicellular ame- boid bodies, and the following year he fouud these same bodies in human lymph. These observations were soon confirmed by Rieck and later by Ogata, who classed the bodies as Protozoa, order Gregarinidee. Guarnieri, in 1892, believed that he had cultivated the same bodies in the epithelial cells of the rabbit's cornea, and although Ferroni and Massari claimed that Guarnieri's parasites were degenerating nuclear products, sub- sequent observers have succeeded in verifying Guarnieri's conclusions. In 1894 Pfeiffer reported the discovery of amebse in the blood of smallpox patients, describing them as free in the plasma, about one -fourth the size of red cells, possessing one or more nuclei, and projecting pseudopodia. Weber and Doehle have described in the blood of measles, scarlet fever, and smallpox, several forms of ameboid bodies which they believe to be de- . velopmental stages of a parasitic protozoon. In 1897 Reed found pigmented ameboid bodies in the blood of vaccinated monkeys and children, and in the blood of variola, but he found the same bodies in noi-mal subjects, and he could not demonstrate any nucleus. Kruse, reviewing the evidence in 1897, ascribed great significance to Guar- nieri's bodies in the epithelial cells of the cornea, believing that they repre- sent the first stage in the solution of the etiology of variola, but he did not grant equal importance to the bodies found in the blood. Widal and Bezancon are reported to have found streptococci in the blood of six cases of variola. VACCINIA. Sobotka examined the blood of 43 children after vaccination, mak- ing the following observations : The red cells and Hb showed no constant variations, and usually remained normal. Vaccination always caused a leucocytosis, beginning usually on the third or fourth day after inoculation and gradually falling till the seventh or eighth day when the leucocytes were frequently subnormal. From the tenth or twelfth day a secondary leucocytosis regularly ap- peared, lasting two to six days, and showing close relation to the ac- tivity of the virus and the number of the pustules. The height of the first leucocytosis varied between 12,000 and 23,- 000; that of the second between 10,000 and 17,500; while in the intermediate period the cells fell as low as 3,500. In uncomplicated cases the lowest figures usually corresponded to the highest tempera- ture, the leucocytosis preceding by several days the local and general manifestations of vaccinia. VARICELLA. Engel ^ found 67 percent of polynuclear cells, with absence of eosins, in a child, at the height of the pustular stage. Three days later, when SCARLET FEVER. 261 most of the pustules were healed, the polynuclear cells had fallen to 47 percent, and the eosins risen to 16 percent. The total numbers of leucocytes were not stated, but the above proportions indicate a mod- erate leucocytosis. SCARLET FEVER. Hayem gave the first systematic account of the blood changes in scarlatina, finding a loss of about one million red cells after deferves- cence, and a moderate leucocytosis in average cases, much increased by severe angina or by suppurative processes. Somewhat isolated obser- vations were reported by many of the earlier blood analysts, Halla, Pee, Pick, Peinert, Sadler, while considerable series of cases were studied by Pieder and Kotschetkoff. The ichole blood suffers in a slight degree the usual effects of fever. The formation of fibrin is usually increased especially when there are extensive angina or suppurative complications. (Hayem.) Heubner noted heraoglobinemia in one case, representing the septicemic type of the disease. Red Cells. — The gradual loss of red cells noted by Hayem was fully verified by Kotschetkoff who found a reduction to 3 millions or lower in nearly all cases. The regeneration of the blood was slow and com- plete only after 6 weeks or longer. On the other hand Zappert found less than 4 million cells in only one of six cases, and very slight anemia was noted in a considerable number of cases examined at varying periods by Leichenstern, Arnheim, Pee, Pick, Reinert, Sadler, and Felsenthal. Leucocytes. — Kotschetkoff classified the leuoocytoses in three groups, the mild cases, showing between 10,000 and 20,000 white cells, the moderately severe cases, with 20,000-30,000 cells, and very severe and usually fatal cases with a leucocytosis of 30,000-40,000 cells, while in some rapidly fatal cases over 40,000 leucocytes were found. Yet Rieder's 10 observations never gave more than 25,000 cells, and usually less than 20,000, although some of his cases were complicated with pneumonia and croupous pharyngitis, and were fatal. Felsen- thal's 6 cases in children were of moderate severity and showed 18,- 000-30,000 cells. The leucocytosis begins 1—2 days before the appearance of the rash, reaches its height with or shortly after the full development of the eruption, and while in some cases rapidly falling with the eruption, usually continues for 4—5 days longer, and very often persists for days or weeks after the temperature has become normal. The grade of leucocytosis seems in general to correspond with the severity of the disease, especially of the angina, but not with the height of the temperature. Complications ^ such as lymphadenitis, otitis, nephritis, usually have little effect on the leucocytes (Kotschetkoff), l)ut Pee observed two cases in which the leucocytosis increased when the lymph nodes began to swell late in the disease. Pneumonia caused but slight increase in some of Rieder's cases. 262 THE EXANTHEMATA. Types of Leucocytes. — The percentage of polynuclear cells is in all cases much increased, varying from 85-98 percent, according to the intensity of the disease, reaching the highest point on the second day of the exanthem, and thereafter slowly declining. In fatal cases the proportion of polynuclear cells falls but slightly, or soon regains or passes the original figure. The eosinophile cells may show characteristic variations. In all but very severe cases they are normal or subnormal at first, steadily increase after 2—3 days, reaching a maximum of 8—15 percent in the second or third weeks, and thereafter declining slowly, reach the normal figure about the sixth week. In fatal cases the eosins may disappear early in the disease. The lymphocytes are at first diminished, but later rise to normal pro- portions. The above rules deduced by Kotschetkoff, while probably represent- ing the average case, are not always applicable. Thus Weiss found no eosins in one case at the height of the exanthem. Rille observed marked eosinophilia in a fatal case ; Bensaude observed as high as 20 percent of eosins in one instance ; Klein reports lymphocytosis during convalescence ; and many writers (Rieder, Turk) have described a high persistent leucocytosis, especially in those cases followed by nephritis or other complications. Turk has called attention to the remarkable change which the leucocytes undergo about the fifth day of the disease, when the polynuclear cells rapidly diminish and eosins and lympho- cytes rapidly increase. This " secondary leucocytosis " he likens to the somewhat similar phenomenon seen in variola. Bacteriological. — Raskin found fitreptococciis pyogenes in the circu- lating blood in 2 of 64 cases, while all other culture was sterile. Negative results were obtained in 2 cases by Sittmann and in 4 cases by Kraus. MEASLES. The red cells have been found in the great majority of cases to suffer little or no change, but a loss of Hb is usually demonstrable after def- ervescence. In adults, uncomplicated measles never causes leucocytosis but is characterized rather by hypoleucocytosis, reaching in one of Rieder's cases 2,700 cells. From 4,000 to 6,000 cells are commonly seen. This fact was first noted by Pee, and has been confirmed by Pick, Rieder, Rille, Felsenthal, and others. Normal or slightly reduced numbers of white cells are found at the onset of the disease. At the height of the exanthem they are usually at their lowest figure (Pee, Turk), and return to the normal within a few days or a week after defervescence. When the bronchitis is severe there may be a moderate leucocj^tosis, Hayem finding 10,000-14,000 cells in such cases occurring in children. Rieder observed slight leu- cocytosis in a case complicated by catarrhal pneumonia. Cabot ob- served 9,000 cells in two cases, one hemorrhagic. The writer found GERMAN MEASLES. 263 no leucocytosis in three cases occurring in malarious subjects. The malarial parasites reappeared, with chills, during convalescence. The proportions of the various leucocytes show no distinctly ab- normal variations. Turk found a rather high percentage of poly- nuclear cells during the fever, with diminution of small lymphocytes. Pee, Klein, and Turk noted an excess of large mononuclear cells. The eosins are usually in low normal proportions during the early fe- brile period, but tend to diminish as the eruption declines. Yet Turk found nearly 5 percent during the second week of the disease. Bacteriological examination of the blood was negative in 10 cases examined by Barbier. Weber claims to have found in the blood of measles a protozoon, which he has also seen in variola. Applications in Diagnosis. — Typical cases of measles and scarlet fever may sometimes be distinguished from each other in their early stages by the examination of the blood. Yet as Turk says, the blood in measles strongly resembles that of a mild scarlatina, as both show a nearly normal number of leucocytes and normal proportions of eosins. Yet equally severe constitutional disturbance should give, on the second to third days, leucocytosis, if scarlatinal, normal or diminished leucocytes, if from measles. In German measles there was no leucocytosis in 2 cases men- tioned by Cabot. CHAPTER XIII. TYPHOID FEVER. The whole blood commonly sufiers concentration in the early stages of the disease as a result of the febrile process, while at any period profuse diarrhea and repeated hemorrhages considerably reduce the total bulk. This concentration is so marked and constant that the deep red appearance of the blood drop in typhoid fever is a very char- acteristic differential sign between this disease and pernicious malaria. Fibrin formation is usually deficient and may not be demonstrable at all, but with inflammatory complications this element may reappear and become abundant. (Hayem, Turk.) Red Cells. — During the febrile period the red cells usually show a slight and gradual decline. Yet the initial concentration of the blood often yields moderate polycythemia during the first two weeks or longer, so that the slight anemia is not to be noticed till the fever and diarrhea subside. (Sadler, Felsenthal, Thayer.) Even when distinct polycythemia does not result the same factors tend to keep up the pro- portion of red cells, so that in the vast majority of cases of typhoid fever the red cells fall within normal limits, and the presence of slight anemia can be demonstrated only by following the case from the first. Accordingly, nearly all observers have reported over 4 million red cells, and many over 5 millions during the febrile period. The Hb suffers to a greater extent, and in spite of the concentra- tion of the blood most febrile cases register between 70 and 80 per- cent of Hb. Severe diarrhea ©r long hemorrhages may at any time cause marked oscillations in the red cells and Hb. After, and often before defervescence, the blood may show distinct anemia, which with rapid defervescence may appear suddenly. (Arnheim, Zaslein, Boeck- man, Laache.) During convalescence the anemia slowly disappears, the Hb being; restored much later than the red cells. The grade of anemia observed after typhoid fever varies greatly with the character of the disease. Uncomplicated cases may show no dis- tinct reduction of cells and very slight loss of Hb. In a considerable proportion of severe cases the red cells fall below 4 millions and the Hb below 70 percent. Kohler in 29 cases found an average loss of 19.4 percent of Hb in men, and 20.4 percent in women, while the average of red cells for men was 4.03 millions, and for women 3.58 millions. Hayem, Thayer, Henry, and others have reported severe grades of pernicious anemia as sequelae of typhoid fever. In one of Henry's cases there were only 804,000 red cells. Morphological changes are usually not marked, but in severe LEUCOCYTES. 265 cases Turk found megalocytes, small microcytes, and polychromasia, in comparative abundance. Nucleated red cells may appear after hem- orrhage. Leucocytes. Numbers. — In the first week uncomplicated cases nearly always show a normal number of leucocytes, but severe angina, bronchitis, enteritis, etc., may yield a moderate polynuclear leucocy- tosis. Klein and Aporti regarded this initial leucocytosis as of fre- quent occurrence, but most other observers have failed to encounter it. The writer has seen some examples. In the second week the leucocytes usually show a distinct reduction, especially of the polynuclear forms, but the number of leucocytes found at this time is by no means constant. In some cases the lowest figures of the disease are reached in the second week (Turk, Chetagurow), while in others there is no distinct reduction at this time (v. Jaksch,^ Sadler). In the third and fourth weeks the leucocytes usually continue to diminish until the acme of the disease is reached (Turk), after which they slowly increase. In many cases the lowest figures are reached at this period (Rieder, Turk, Thayer), but sometimes the reduction con- tinues till after defervescence (Klein). A relation between the leucocytes and the fever seldom exists nor does the size of the spleen appear to influence the number or type of the leucocytes. (Turk.) That the reduction of the white cells is de- pendent on the action of the typhoid toxine is indicated by the further losses which commonly go with unfavorable turns in the disease. (Rieder, Jez, Turk, Nagaeli.) It is thei^ejore a general rule that the more severe the typhoid intoxication the lower is the count of leucocytes. The reduction is seldom below 2,000 cells, but 1,000-2,000 were reported by Hayem, Limbeck and Rieder, Cabot, and Kohler. In Thayer's cases the lowest weekly average was 5,877, and was obtained in the fourth week, but the majority of cases at some periods show be- tween 4,000 and 6,000 cells, and many fall below 4,000. Leucocytosis during the course of typhoid fever is by no means uncommon. Usually a severe diarrhea from extensive ulcers, or hem- orrhage, or pneumenia, or other exudative or suppurative complica- tions, may be found to explain the increase. Yet several reported cases with leucocytosis have shown no such complications (Aporti), and a short experience at the bedside seldom fails to bring to light some moderate leucocytoses without apparent cause. On the other hand leucocytosis may fail in the presence of pneumonia or other markedly exudative complications. (Turk, the writer, Cabot.) Kohler found 6,200 and 4,200 leucocytes in cases complicated by severe bronchitis. After hemorrhage in one case the white cells remained at 2,600, in tM^o others there was an increase of about 4,000 cells. Complicating pneumonia once reduced the leucocytes from 2,600 to 1,000, at another time they remained at 6,300, and twice 11,200 and 11,800 cells were counted. (Kohler.) Cold baths have long been known to cause temporary massing of 266 TYPHOID FEVER. leucocytes in peripheral capillaries. (Winternitz.) lu 20 cases Thayer ^ found an average increase of 5,346 cells, and a maximum increase of 17 000. With the return of peripheral circulation the excess of cells disappears. Perforation usually causes poly nuclear leucocytosis but there may be no effect upon the leucocytes, or the percentage of polynuclear cells may rise without increase in their total number. Types of Leucocytes. — Daring the first week of the disease the per- sistence of a normal proportion of lymphocytes without increase of neu- trophile cells is a very characteristic feature of typhoid blood. Bar- ring the occasional occurrence of initial polynuclear leucocytosis, from the end of the first week there is a progressive increase of the lympho- cytes and diminution of the neutrophile cells. At first the lympho- cytes do not pass beyond high normal limits, but during the third, fourth, and fifth weeks, or later, they usually occur in distinct excess. While mononuclear cells seldom fall below 30 percent at any stage of the disease a proportion of 40-60 percent is then seen. Among these cells the proportion of small lymphocytes is often striking. In some cases there is a very marked absolute lymphocytosis. In one of the writer's cases at Montauk, the blood resembled that of lymphatic leu- kemia, a resemblance rendered still more striking by the enormous size of the mesenteric nodes at autopsy. With the increase of mononuclear cells the numbers of polynuclear cells gradually fall, reaching 50 percent in most cases and occasionally a much lower figure. (35 percent, Jez ; 20 percent, Klein.) The excess of lymphocytes usually persists during the first weeks of convalescence and may not reach its acme until this time, while Ouskow found that normal relations were not restored till the tenth or eleventh week. Nagaeli describes the development during convalescence, especially in children, of a well-marked lymphocytosis with a moderate increase of neutrophile cells. In children this lym- phocytosis was most marked 2—3 months after the fever, while in adults it was less marked and disappeared by the end of the second month. Eosins are usually absent or very scarce during the febrile period, but reappear shortly before (Nagaeli), during, or after defervescence. Aporti found as high as 18 percent during the intermittent pyrexia at the end of the disease, but they are usually much less abundant. Degenerative changes in the leucocytes occur as usual, Jez reporting large numbers of pale leucocyte-shadows in severe cases. Glycogen was found in the leucocytes in increased quantity between the twelfth and twentieth days by Livierato. Applications in Diagnosis. — The morphological examination of the blood is often of great assistance in the diagnosis of typhoid fever and the diseases which simulate it. The writer knows of no clearer illus- tration of this fact than those which he reported' in 1893, and which have been paralleled from numerous later observations and by many writers both before and since. Suppurative processes, if active, may usually but not always, be dis- tinguished from typhoid fever by the presence of polynuclear leucocy- BACTERIOLOGICAL. 267 tosis. Yet it shpuld be remembered that slow suppuration, or the mere presence of pus the secretion of which has ceased, are frequently unaccompanied by leucocy tosis or are even marked by relative lym- phocytosis. Thus the writer found 50 percent of lymphocytes among 11,000 cells in a case of large abscess of the liver with mild typhoidal symptoms, and 45 percent of lymphocytes among 7,000 cells with the chest full of pus. With increasing experience therefore, the writer believes that most careful observers will recognize the partial justice of Grawitz's claim that in difficult cases the blood examination here often fails to be of service. Among the conditions which, by the almost invariable presence of leucocytosis, may be distinguished from typhoid fever, are bacterial en- docarditis, suppurative appendicitis, and pneumonia. Malaria of suf- ficient gravity to simulate typhoid fever can nearly always be distin- guished by the anemic appearance of the blood drop as compared with the deep red concentrated blood of typhoid fever. An astonishing change in the blood was often noted at Montauk when typhoid fever developed in subjects of malarial cachexia, and in no instance were malarial parasites found in such concentrated blood. In acute par- oxysms the discovery of the parasite is usually possible, but chronic malaria without parasites in the blood may simulate typhoid fever. Bstween miliary tuberculosis, typhoid fever, and some forms of meningitis, Widal's test is required and the enumeration of leucocytes is of little value. From an extensive study of the leucocytes in typhoid fever Nagaeli draws the following conclusions regarding jjrognosls. The prognosis is favorable : (1) When eosins are present at the height of the disease, or reappear in the second or third stages of the febrile period. (2) When lymphocytes begin to increase after the severest toxemia is past. (3) The diminution of neutrophile cells is slight only in more favorable cases. Unfavorable signs are : (1) Very small numbers of all varieties of leucocytes. (2) Failure of leucocytosis with complications. Bacteriolog"ical. — In the hope of developing a method of early diagnosis of typhoid fever many biological studies of the blood have been undertaken. Aspiration of the spleen has given successful results in the majority of cases (Chantemesse and Widal, Redtenbacher, Lu- catello, Neisser, Bruschettini), but this dangerous procedure cannot be generally adopted. Yet the results of Silvestrini who found the typhoid bacillus in the aspirated blood of the spleen in four cases in which there were no characteristic intestinal lesions are most suggestive. In blood drawn from the skin or rose sjjots, negative results were ob- tained by Chantemesse, Widal, Janowski, Merkel, Seitz, Lucatello, Urban, and small proportions of positive results have been secured by Rutimeyer, Wiltschour, Frankel, Grawitz, Menas, and jNIenzer. About 50 percent of successful tests were made by Sudakoff and Thiemich, while Neuhaus^ has recently succeeded in cultivating the typhoid bacillus from the rose spots in 13 of 14 cases. In the circulating blood during life the typhoid bacillus has been ob- 268 TYPHOID FEVER. tained in one case each by Pasquale, Thiemich, Bloch, and Kraus, and twice by Ettlinger and by Stern. Kuhnau secured 1 1 positive cultures n 41 cases, examining blood drawn from the basilic vein, at the acme of the disease. Of 38 cases examined by James and Tuttle there were 3 positive results, one only proving fatal. On the other hand, entirely negative results in about 40 cases are re- ported by Neuhaus, Ettlinger, Frankel, Klein, Bloch, and by many others. That the disease may assume the character of a mixed septicemia is indicated by a considerable number of reports of the presence in the blood of other bacteria, especially Staphylococcus aureus. (Loisson, Sitt- mann, Kraus.) From the above synopsis of the work in this field, it appears that bacteriological blood-analysis in typhoid fever can have only a very limited clinical application. "While the results obtained from the aspi- rated blood of the spleen are usually successful, and Silvestrini's re- port of finding the Bacillus typhosus in the spleen when characteristic intestinal lesions were absent is very suggestive, yet this procedure is too dangerous for general use. In the examination of the blood drawn from the rose spots negative results thus far outnumber the positive, although Neuhaus' recent re- port is encouraging. Kuhnau's rather numerous successful cultures from considerable quantities of the circulating blood are the most favorable yet recorded. While not widely applicable in diagnosis the bacteriological ex- amination of the blood has, however, greatly extended our knowledge of the morbid processes concerned in typhoid fever. Bibliography. Pneumonia, Diphtheria, Typhoid, Exanthems. Albu. Virchow's Archiv, Bd. 149, p. 405. Aporti. XI. Internat. Cong., Rome, III., p. 381. Arnheim. Jahrbuch f. Kinderheilk., Bd. 13, p. 293. Bacciochi. Lo Sperimentale, 1893, Nos. 16-17. Barbier. Semaine Med., 1897, p. 37. Belfanti. La Rif. Med., 1890, I., p._338. Bensaude. Cited by Besancon, Archiv Gen., 1900, p. 491. Bieganski. Deut. Archiv klin. Med., Bd. 53, p. 433. Billings. » Johns Hopkins Bulletin, 1894, p. 105. 2 N. Y. Med. Record, Vol. 49. p. 577. Bloch. Johns Hopkins Bulletin, 1897, p. 119. Boeckman. Deut. Archiv klin. Med., Ed. 29, p. 481. Bollinger. Miinch. med. Woch., 1895, No. 32. Bouchut. Gaz. d. Hop., 1879, p. 153. Boulay. Des. infec. apneumococ, Paris, 1891. Bozzolo. La Rif. Med., 1889, No. 45. Brouardel. Gaz. med. de Paris, 1874, p. 141. Bruscheitini. La Rif . Med., 1892, No. 34. Chantemesse, Widal. Gaz. hebdom., 1887, p. 146. Chetagurow. Abstr. Virchow's Archiv, Bd. 126, p. 187. Ouffer. Rev. mens, de Med., 1878, p. 519. Doehle. Cent. f. Path., 1892, p. 150. Engel. » Deut. med. Woch., 1897, pp. 118, 137. "XV. Cong. f. inn. Med. Ettlinger. Cited by Bloch. BIBLIOGRAPHY. 269 Ewing. IN. Y. Med. Jour., Vol. 58, 1893, p. 713. «Ibid., 1895, Vol. 62, p. 161. Felsenthal. Archiv f. Kinderheilk., Bd. 15, p. 78. Ferroni, Massari. La Kif. Med., 1893, II., p. 602. File. Lo Sperimentale, 1896, p. 284. Frankel. Deut. med. Woch., 1886, No. 1. Oabritschewsky. Annal. de I'lnstitut Pasteur, 1894, p. 673. Ooldscheider. Deut. med. Woch., 1892, No. 14. Gundobin. St. Petersburg med. Woch., 1897, Lit. No. 7, p. 37. Halla. Zeit. f. Heilk., 1883, p. 198. Hayem. "Archiv gen. de Med., Mar., 1884. Henry. Cited by Cabot. Heubner. Deut. Archiv klin. Med., Bd. 23. V. Jaksch. »Zeit. f. klin. Med., Bd. 23, p. 207. "Prag. med. Woch., 1890. James, Tuttle. Presbyterian Hospital Reports, III., p. 44. Janowski. Cent. f. Bact., Bd. 5, No. 20. Jez. Cited by Turk. Kanthack, Hardy. Jour, of Physiol., Vol. 17, p. 91. Kikodse. Cited, Cent. f. Path., 1893, No. 3. Klein. Cited, Baumgarten's Jahresb., 1894, p. 266. * Volkman' s Vortrage, 1893, No. 87. Kohler. Deut. Archiv klin. Med., Bd. 60, p. 221. Kohn. Deut. med. Woch., 1897, p. 136. Kotschetkoff. Ref. Cent. f. Path., 1892, No. 11. Kraus. Zeit. f. Heilk., Bd. 17, p. 117. Zeit. f. Heilk., Bd. 17, p. 138. Kruse. Pflugge, Microorganismen, 1897. Kuhnau. » Zeit. f. klin. Med., Bd. 28, p. 534. ^Zeit. f. Hygiene, Bd. 25, p. 544. Laehr. Berl. klin. Woch., 1893, Nos. 36-37. Limbeck. 'Zeit. f. Heilk., Bd. 10, p. 392. Livierato. Deut. Arch. klin. Med., Bd. 53, p. 303. Loison. Eev. de Med., 1893, No. 4. Lucatello. Baumgarten's Jahresb., 1886, p. 176. Marchiafava, Bignami. La Rif. Med., 1891, Nos. 251, 252. Menus. Cited by Kuhnau. Memer. Diss. Berlin, 1892. Merkel. Cent. f. klin. Med., 1887, No. 22. Monti, Berggrun. 'Archiv f. Kinderheilk., Bd. 17, p. 1. 3Iorse. Boston Med. Surg. Jour., Vol. 132, p. 228. Nagaeli. Deut. Archiv klin. Med., Bd. 67, p. 279. Neisser. Zeit. f. klin. Med., Bd. 23, p. 93. Netter, Levy. Cited by Sittmann. Neuhaus. 'Berl. klin. Woch., 1886, Nos. 6, 24. ^Zeit. f. Hygiene, Bd. 30, p. 498. Ogata. Mitt. a. d. Kaiserl. Japanish Universitat, Tokio, 1892, p. 403. Ouskoiv. Cited by Turk. Pasquale. Baumgarten's Jahresb., 1891, p. 249. Pee. Diss. Berlin, 1890. Pfeiffer. Monatsheft. f. prak. Dermat., 1887, No. 10. Handb. d. spec. Therapie, 1894, p. 229. Pick. Prag. med. Woch., 1890, No. 24. B. Pick. Archiv f. Derm. u. Syph., 1893, p. 63. Piorry. Hematopathol., Leipsic, 1839. Boskin. Ref. Cent. f. Bact., 1889, p. 286. Bedtenbacher. Zeit. f. klin. Med., Bd. 19, p. 305. ^ Beed. Jour, of exper. Med., 1897, p. 515. Bieck. Rundschau a. d. Gebiet. d. Thiermedecin, 1887, p. 77. Rieder. Beitriige zur Kenntniss d. Leucocytose, Leipsic, 1892. Bille. Archiv f. Derm. u. Syph., 1892, p. 1028. Butimeyer. Cent. f. klin. Med., 1887, No. 9. Sadler. Fort. d. Med., 1892, Suppl. Heft. Schlesinger. Archiv f. Kinderheilk., Bd. 19, p. 378. Seitz. Stud. u. Typhusaetiol., Miinchen, 1886. Silvestrini. Settimana Med., 1897, Nos. 45, 46. Sittmann. Deut. Archiv klin. Med., Bd. 53, p. 323. Smith. N. Y. Med. Jour., Vol. 49, p. 649. Sobutka. Zeit. f. Heilk., Bd. 14, p. 411. Sorensen. Cited by Reinert, Diss. Copenhagen, 1876. Soudakoff. Cited by Kuhnau. 270 TYPHOID FEVER. Stern. Cent. f. inn. Med., 1896, p. 1249. Stieiion. Jour, de Med. (Brussels) Annales, 1895. Thayer. ^ Johns Hopkins Hospital Reports, Vol. 4, p. 82. ^ Johns Hopkins Bul- letin, 1893, p. 37. Thiemich. Baumgarten's Jahresb., 1894, p. 266. Tschistowilsch. Annal. de 1' Institut Pasteur, 1891. Archiv des Sci. Biol., St. Peters- burg, 1893. Tumas. Deut. Archiv f. klin. Med. , Bd. 41, p. 323. Turk. iKlin. Blutuntersuch. Wien., 1898, p. 96. Urban. Wien. med. Woch., 1897, Nos. 32-35. Weber. Cent. f. Bact., Bd. 21, No. 6. Weiss. Jahrb. f. Kinderheilk., Bd. 35. Widal, Bezaneon. Cent. f. Path., 1896, p. 569. Wiltschour. Cited, Cent. f. Bact., 1890, p. 279. Winternitz. Cent, f klin. Med., Bd. 14, p. 177. Znppert. Zeit. f. klin. Med., Bd. 23. Zaslein. Diss. Basel, 1881. CHAPTER XIV. WIDAL'S TEST. The history of the development of Widal's test from the discovery of the phenomenon of Pfeitfer in 1894 through its first application to the diagnosis of typhoid fever by Pfeiffer and Kolle, and Gruber and Durham, in 1896, up to its more practical modification by Widal, has now been very fully reviewed in numerous monographs and treatises on bacteriology and need not here be repeated. It may only be said that while the scientific basis of the test had been firmly laid by previous studies, up to the time of Widal's communication there had been no indication that the specific reaction could be obtained from small quantities of serum and at the beginning of the disease, while the use of dried blood and the introduction of the test into municipal laboratories, thus rendered possible, was first accomplished by John- ston of Montreal. Description of the Reaction, — When to a few cubic centimeters of blood serum of a patient suffering from typhoid fever is added a few drops of an actively motile culture of Bacillus typhosus, the bacteria are precipitated in flakes, leaving the supernatant fluid clear. When the sediment is examined microscopically it is found that the bacilli have lost their motility and are agglutinated in masses, while with very intense reactions they may be broken up into granules or even completely dissolved. When the test is performed with a drop of dis- solved blood inoculated by means of a platinum loop with a small por- tion of broth culture, all stages of the reaction may be followed under the microscope. With very intense reactions the agglutination is found to be complete as soon as a cover-glass can be adjusted over the drop and the specimen placed under focus. The bacilli are motionless, they are nearly all gathered in characteristic clumps, which stretch in a network irregu- larly through the field, and very soon, in some cases, evidences of sub- division or solution of the bacteria may be detected. If the 7'eacfion is less complete, the clumps are less compact and a few motile bacilli may be found between the masses. With high dilutions of the blood, all stages of the reaction may be followed. The first change noted is a partial loss of motility. The bacilli cling to one another after struggling to free themselves, but gradually entangling their neighbors. In the course of ten or fifteen minutes the process results in the gradual clumping of all bacilli, or, according to the strength of the reaction, a variable number of actively motile bacilli may remain between the loose masses. 272 WIDAL'S TEST. When the reaction is present in traces only, much longer time is re- quired for the appearance of small clumps, but these reactions must be discarded as unreliable. All experienced observers have insisted that distinct islands of clumped bacilli must be present to constitute a posi- tive reaction, and Fischer and others maintain that all bacilli must be rendered quiescent before the result can be regarded as 'positive. PsEUDO-REACTioxs. — While partial reactions occur under many conditions when the technique has been faultless, certain eflPects simu- lating the true agglutination may appear in specimens containing not a trace of the specific principle on which the test depends. In old cultures many loose clumps of bacteria may be transferred by the loop from the culture tube and be mistaken for an effect produced by the serum. Normal blood or serum, in somewhat concentrated form, exerts a marked influence upon the motile typhoid bacillus, causing it to lose its motility after a time, but failing to develop typical clumps. If the specimen has been allowed to dry many bacteria adhere to slide or cover- glass or gather at the thickening edges of the specimen and sometimes these collect in motionless groups. Such clumps always lack the re- ticular arrangement seen in the true reaction. A considerable variety of confusing appearances will meet the beginner, resulting from the formation of fibrin and the incomplete solution of masses of red cells. Technics : Methods of Obtaining the Body-Fluid. — Although specific reactions have been obtained from many secretions, excretions, and other fluids of the body, only the blood, its serum expressed from the clot, or the fluid gathering in blisters, yield sufficiently constant results to be available in clinical diagnosis. The Whole Blood. — When the examiner can reach the bedside there is no better method than to employ the whole blood diluted ia known proportion by distilled water. In order to obtain an accurate dilution a medicine dropper may be graduated in one and ten or twenty drop marks, as suggested by Cabot, but such dilution cannot be very accurate. The writer has long used for this purpose the leucocyte pipette of the Thoma-Zeiss heraatocytometer, which gives an exact dilution of 1-20, or 1-50, or more, and which is itself a very conve- nient instrument for handling the blood. Levy and many others have devised special tubes adapted to this purpose. The chief objection to this method is the failure of the blood to completely dissolve in dilu- tion of 1-20, or 1-50, of water, but no other method can furnish more exact dilution, which is essential in accurate work. This difficulty may be avoided by laying the pipette on the side until the corpuscles settle. Blood dried on a glass slide and dissolved by mixing with water is now used in probably 90 percent of all examinations. One or two drops are touched to a clean glass slide and dried in the air. Such a specimen protected from moisture may be kept for days, transported for long distances, and being easily obtained, even by the patient him- self, furnishes by far the most practical method of obtaining the neces- sary material for the test. Its great disadvantage is the impossibility DESCRIPTION OF THE REACTION. 273 of securing an exact dilution of the blood when it is redissolved. This difficulty, while theoretically insurmountable, proves to be of compara- tively little moment in practice, since an experienced worker can learn to control the dilution by the color of the drop. Also, when imper- fectly dissolved the detritus of red cells may obscure the field, although it should not fail to be distinguished from masses of bacilli. Some observers have found less powerful and constant effects from dried blood than from serum (Widal, Delepine, Johnston), but for diagnostic purposes Park finds the two methods of nearly equal value. Nevertheless, for accurate studies of variations in the reaction, the use of dried blood is inapplicable, although it will probably always remain in extensive employment for simple diagnostic purposes. Serum. 1. The Bulbed Tube. — A capillary tube with central bulb is filled by capillary attraction from the expressed drop, the ends are fused, and when the blood coagulates a few drops of serum are exuded. A great many modifications suited to individual purposes have been suggested for this method. It offers considerable technical difficulties, but furnishes serum without delay. 2. The Blister. — Serum may be obtained by means of a Spanish fly blister in from 6-18 hours and in quantity sufficient for extended ex- amination, without involving much pain or inconvenience. The fluid may be drawn in a sterile test-tube and afterwards divided in meas- ured quantities, or it may be sealed in a capillary tube. For detailed study blister-serum is superior to all other fluids. For measuring serum the *' mixer" of the hematocytometer is well adapted, and the test may be performed in the hanging drop or in the test-tube. The Culture. — Considerable difference in the response of different cultures of bacilli have been demonstrated by Kolle, Achard and Ben- saude, Durham, v. de Velde, and Ziemke. Although Foerster found very little difference in a considerable series of cultures, most ob- servers agree that attenuated cultures are more often agglutinated by normal serum in considerable dilution (1-10 ; 1-15). It appears that for each culture there is a limit beyond which normal serum is posi- tively inert, while at the same dilution typhoid serum exerts a specific action. (Foerster.) While any pure culture may be used, the best results have been obtained in New York City by a so-called " Pfeiffer specimen" imported by Park. A stock culture of this particular growth is kept in sealed tubes of nutrient agar in the ice chest and re- planted every few months. From this growth the specimen employed in testing is obtained in broth culture grown for 24 hours at a temper- ature of 35° C. In such broth cultures the bacilli are completely iso- lated and actively motile, so that the slightest change in motility or tendency to clump can be detected by comparison with a control specimen. Dead Bacilli. — Widal first observed that cultures of typhosus which had been killed by heat (57-60° C, 30-45 minutes) had lost little or none of their capacity to agglutinate, and this fact has been abundantly attested by others. Many have therefore preferred to use 18 274 WIDAL'S TEST. a dead culture killed by heat, or by 10-percent formalin, or by thymol, or by corrosive sublimate 1,000-5,000 (Park), thereby dispensing with the trouble of preparing a fresh culture. Such an expedient, however, can hardly meet with general adoption, on account of the difficulty of keeping the dead bacilli well isolated, and because the gradual loss of motility is a desirable and reassuring feature of the test in mild reactions, which are just the ones in which uncertain clinical signs render a positive test most desirable. Attenuated cultures have been recommended by Johnston, who found that in using dried blood pseudo-reactions were not infrequent with active virulent cultures, while broth cultures planted from old agar stock were less susceptible and did not respond to healthy or non- typhoid sera. Others have failed to find this advantage in attenuated cultures, which have not come into prominent use with either dry blood or serum. Degree of Dilution of Serum and Time Required. — Although it has been believed that the serum reaction in typhoid fever depends on the presence in the blood of a peculiar substance which exerts a specific action on the typhoid bacillus, yet it has been found that in a consider- able proportion of cases healthy serum, or that of patients suffering from other diseases, contains substances which exert a very similar in- fluence upon typhoid cultures. In typhoid fever, however, the sei'uin may usually be diluted 20 to 50 times, or often very much furthen- {1,000- 2,000, Widal ^) without destroying the reaction, while in other diseases the reactions usually disappear with dilutions of 1—10, or almost certainly with dilutions of 1—20. The element of time is also of great importance in distinguishing the specific typhoid reaction from that produced by other sera. While a distinct reaction often occurs instantly with typhoid serum, blood of other diseases almost never acts immediately but requires one-half to one hour or longer to disclose its effects. Consequently, in order to demonstrate the specific typhoid reaction it is necessary to increase the dilution and to limit the time. " The results obtained in the New York City Health Department laboratories, and elsewhere, have shown that in a certain proportion of cases not typhoidal, there occurs a delayed moderate reaction in a 1-10 dilution of serum or blood ; but very rarely, if ever, excepting in ty- phoid fever, does a complete reaction with this dilution occur within five minutes. When dried blood is used, the slight tendency of non- typhoid blood in 1-10 dilution to produce agglutination is increased by the presence of fibrinous clumps and perhaps by other substances de- rived from disintegrating red cells. From many cases examined by Fraenkel, Stein, Foerster, Scholtz, ourselves, and others, it has been found that in dilution of 1—20 or more, a decided quick reaction is never produced by any febrile disease other than typhoid infection, in which it often occurs in dilution of 1—50^ (Park.) Yet even this dilution appears insufficient to eliminate all sources of error, and there has been a constant tendency among experienced OCCURRENCE OF THE REACTION. 275 workers to increase the grade of dilution. Grunbaura placed it at 1—30 ; Simon, at 1—30 or more ; Stern, after a large experience, at 1-40 ; and Mewius, at 1-60. From these authoritative opinions it becomes necessary to prescribe a dilution of at least 1-20, better 1-30, while the limit of time should not be greatly extended. Even these limits will be found insufficient to avoid all possible error, and in doubtful cases one should use a dilu- tion of 1—40 or 1—60, increase the time to one or two hours, and re- quire a distinct result. , Occurrence of the Reaction. — Statistics vary considerably regard- ing the dates of occurrence and the constancy of the reaction in typhoid fever. Cabot collected over 3,000 cases of supposed typhoid fever, of which 95 percent gave the reaction at some period of their course, while in 2,500 control cases, not typhoid, 2 percent gave positive reactions. Park reports positive results in a large series. During the 1st week, in 20 percent of the cases. " " 2d " " 60 " " " " " 3d " " 80 " " " " " " 4th " " 90 " " " " " 2d month," 75 " " " " In 88 percent of hospital cases in which repeated examinations could be made, a definite reaction was obtained at some period of the disease. The earliest date of appearance of a positive reaction has not been and can not well be determined in the human subject, but in animals inoculated with dead typhoid bacilli the reaction appears between the third and eighth days. In man a positive reaction has been found by Johnston and Taggart, and by Fraenkel, on the second day, but the re- actions were not sharp. Levy found the first distinct reaction on the sixth day in the human subject inoculated with dead typhoid bacilli. It may be found before the appearance of rose spots, splenic tumor, or the diazo reaction in the urine. During the course of the disease the reaction may continue without interruption, or may be absent one day while present the next. It may be absent entirely in mild cases, may appear only in a relapse (Lichtheim, Breuer, Thoinet), may disappear entirely after a few days (Elsberg), or may persist for months or even a year. Incomplete reactions have been reported after many years. (Stewart.) It not infrequently appears for the first time during con- valescence or may exhibit a sudden increase at this time. According to Widal, in the majority of cases the reaction disappears by the fif- teenth to thirtieth day of convalescence, but Stewart believes that it persists at least one year in 50 percent of the cases, for two years in 25 percent, and for ten years in 5 percent. His conclusions can only apply to very indistinct reactions. Relation of the Reaction to Other Features of the Disease. — In some cases the reaction is most intense at the height of the disease, as in cases boserved by Jemma. Widal and Sicard, who found that in the active 276 WWALS TEST. stage a dilution of 1-60 or 1-80 does not usually prevent the reaction, noted a marked weakening as convalescence proceeded, some cases fail- iuo- to react with dilution of 1—10. They noted a reaction in one case with a dilution of 1,000, or 12,000. Foerster, who found a reaction in dilutions varying from 1-60 to 1,000, or 5,000 could not detect any relation between the intensity of the reaction and the severity of the disease. Pfeiffer and Kolle, and Foerster have shown that the agglu- tinating power has no relation to the bactericidal activity of the serum. Effect of the Typhoid Serum Upon the Colon Bacillus. — Although Fraenkel in a large series of eases reported that typhoid serum has no marked power to agglutinate the colon bacillus, and recommended the use of this serum in the separation of colon from typhoid bacilli, this claim has not been verified. Stern and Biberstein found, very shortly, five cases of typhoid fever of which the blood serum agglutinated the colon bacillus even more powerfully than the typhoid, and many other observers, including Park, Rodet, Courmont, Ustvedt, Beusaude, and Kuhnau, have had similar experiences in many cases. Baumgarten, therefore, concludes that this method can no longer be accepted as positive proof of the identity of Bacillus coli communis. It neverthe- less remains true that a germ which on cultural characters falls be- tween the tvphoid and colon groups, is almost certainly coli communij<, if it is not agglutinated by a well-tested and active typhoid serum. The possibility that infection by the colon bacillus is intermingled with that of Bacillus tj/phosiis in enteric fever has been abundantly con- sidered by several writers. The safest method of identifying the colon bacillus is not by the serum of a case of typhoid fever but by the serum of an animal which has been moculated with a pure culture of BojciUus typhoms. Reactions in Conditions Other than Typhoid Fever. — In view of the fact that positive reactions have been obtiiined both with healthy serum and in diseases other than typhoid fever, when a dilution of 1-10 was used, many observers have recommended that higher dilutions only be employed. Thus Schultz obtained complete reactions in 11 of 100 cases of various febrile diseases with a dilution of 1—10 ; in 7 cases, with a dilution of 1—15 ; and in 3, with 1—20 ; while a faint response followed a dilution of 1—25 in a single instance. The time limit, howev'er, was 1—2 hours. In single cases complete reactions with di- lutions of 1-10 or more have now been reported in so many cases that it is no longer possible to refer the results to faulty technique. Some of these conditions are as follows : Septicemia (Ferrand), malaria (Bloch, Villier, Catrin, and the writer), pneumonia (Kasel and Mann), tuberculosis (Wesbrook, Jez, d'Espine), otitis (Stern), influenza (Wes- brook), typhus (Park), meningitis (v. Oordt), normal serum (Kuhnau). In many of these cases the dilutions were from 1—10 to 1—30, or more, and the clumping was prompt and complete. Negative results in undoubted cases of typhoid fever are reported by many observers. Stewart, using dry blood in 538 cases, found typhoid lesions at au- topsy in five which failed at three periods during the disease to give a positive reaction. Thompson also reports 6 negative results in 163 cases, while a slight reaction was obtained in 12 percent of various conditions other than typhoid fever. Ustvedt encountered a peculiar epidemic of typhoid fever among soldiers in which 4 of 15 cases failed to yield a reaction. Total failure of reaction, under full pre- cautions, is also reported in single cases by Artaud, Haushalter, Schu- macher, Fisher, and others. Value in Diagnosis. — The difficulty in determining the value of Widal's test in diagnosis arises principally from the divergent opinions regarding the essential features of a complete reaction. Although the belief in the specific quality of AVidal's reaction has been abandoned, the test remains an extremely valuable diagnostic procedure under several conditions. It must be granted at once that absolute certainty cannot be ascribed to any test unless the dilution has been very high, certainly not less than 1 to 60, with a time limit of not more than 15—30 minutes. A positive result with much less dilution (1—30) must stand as al- most certain evidence that the disease is typhoid fever. Such results are usually obtainable, however, only after the disease is well estab- lished and its clinical symptoms distinct. Accordingly, the chief value of the test lies in its capacity to distinguish, during the height of the illness, certain conditions such as acute gastritis, tuberculosis, menin- gitis, and pneumonia, which may simulate typhoid fever. The application of the test in the ear/y diagno-ns of typhoid fever has, in the experience of most observers, proved disajipointing. It is rarely to be found until other signs render the diagnosis very probable, when the blood test may often be added, as any other isolated clinical symptom, to the evidence for or against typhoid fever. The writer believes, however, that the combination of an indistinct serum reaction, diminution of fibrin, absence of leucocytosis, and presence of relative or absolute lymphocytosis, can almost never be demonstrated in the early stages of any obscure febrile disease except typhoid fever. The morphological examination of the blood is, therefore, a valuable adjimct and control in the application of Widal's test, and the writer believes that in doubtful cases it should never be omitted. Negative results are of very moderate import until the third or fourth week, and unless often repeated, while some reported cases, often fatal, have failed at any time to yield the reaction. Summary. — The most reliable method for clinical purposes is the use of fresh blood diluted with water in a graduated pipette, and in- oculated as a hanging drop preparation. Blood or blister serum should be employed for more extended study, while the use of dried blood, sometimes necessary, gives very slightly less reliable results. The culture should be grown in broth at 36° C, 24 hours old, actively motile, and of moderate virulence. A recognized brand should be obtained if possible, and the behavior of all newly isolated cultures should be tested before use. 278 WIDAL'S TEST. A positive reaotion consists iu the complete immobilization of all bacilli and the compact clumping of the great majority. A time-limit of five minutes is all that is necessary, with dilutions of 1 to 20, in the great majority of positive reactions. With higher dilu- tion (1 to 40) the time may be extended to 30 minutes, and with dilu- tion of 1 to 60, to 2 hours. (Me wins.) A dilution of at least 1 to 20 is required in all positive reactions, with which, however, occasional errors will occur. Dilutions of 1 to 30, or 1 to 60, are to be strongly recommended, for the higher the di- lution the more certain are all positive results. A negative result is of little import unless obtained during the height of the disease (third to fifth weeks) and on repeated examina- tions. In the great majority of cases the reaction disappears within a few weeks after convalescence, and sev^eral authentic cases, usually fatal, have not given a positive reaction at any time. The agglutinative power of typhoid serum is usually slight and transient in mild cases and marked and persistent in severe ones, but bears no constant relation to the gravity of the disease, and gives no certain prognostic indications. Bibliography. Widal's Test. Aehard, Bensaude. Compt. Rend. Soc. Biol., 1896, 5p. 940. Artaud. Presse Med., 1898, No. 18, p. 106. Bensaiide. L' Agglutination des Microbes, Paris, 1897. Biberstein. Zeit. f. Hygiene, Bd. 27, p. 847. Biggs, Park. Amer. .Jour. Med. Sci., Vol. 113, p. 274. Block. Jour. Amer. .ALed. Aasoc, 1897, Vol. 29, p. 7. Breuer. Berl. klin. Woch., 1896, p. 1037. Catrin. Soc. des Hop., 1896, p. ()98. Courmont. Compt. Rend. Soc. Biol., 1898, p. 756. Also, Sero-prognostic d. 1. fiev. tvph., Paris, 1897. JDdepine. Brit. Med. Jour., 1897, I., p. 1894. Durham. Lancet, 1896, II., p. 1746. Elsberg. N. Y. Med. Record, Vol. 51, p. 510. d. Espine. Rev. med. de la Suisse Roinunde, 1898, p. 113. Fei-rand. Bull. Soc. d. Hop., 1897, p. 104. Fischer. Zeit. f. Hygiene, Bd. 32, p. 407. Foerster. Zeit. f. Hvgiene, Bd. 24, p. 500. Fraenkd. Munch. liied. Woch., 1897, p. 107. Gruber, Durham. Munch, med. Woch., 1896, pp. 206, 285. Orunbaum. Brit. Med. Jour., 1897, II., p. 1852. Haushalter. Presse Med., 1896, p. 505. Jemma. Cent. f. inn. Med., 1897, p. 65. Jez. Wien. med. Woch., 1897, p. 98 ; 1898, p. 890. Johnston. N. Y. Med. Jour., 1896, Vol. 64, p. 573. Johnston, Tagqarl. Brit. Med. Jour., 1896, II., p. 1629. Kolle. Deut. med. Woch., 1897, p. 132. Kosel, Mann. Munch, med. Woch., 1899, p. 581. Kuhnau. Berl. klin. Woch., 1897, p. 397. Levy. Miinch. med. Woch., 1897, p. 1435. Lichtheim. Cited by Foerster. Mewius. Zeit. f. Hygiene, Bd. 32, p. 422. V. Oordt. Miinch. med. Woch., 1897, p. 327. Pfei^fer. Zeit. f. Hygiene, Bd. 18, p. 1. Pfeiffer, Kolle. Deut. med. Woch., 1896, p. 185. BIBLIOGRAPHY. 279 Bodet. Compt. Rend. Soc. Biol., 1898, p. 756. Scholtz. Hvg. Rundschau, Bd. 8, p. 417. Schumacher ' Zeit. f. Hygiene, 1899, Bd. 30, p. 364. t Simon. Cited, Baumgarten's Jahresber., 1898, p. 336. Stern. Cent. f. Bact., 1898, p. 673. Stewart. Amer. Public Health Assoc, Vol. 23, p. 151. Thoinot. Seinaine Med., 1896, p. 504. Thompson. Brit. Med. Jour., 1897, II., p. 1775. Ustvedt. Cited, Baumgarten's Jahresber., 1898, p. 342. V. d. Veldes. Acad. med. de Belgique, 1897, Mar. 27. Villiers. Presse Med., 1896, p. 54. Wesbrook. Ref. Cent. f. Bact., 1898, p. 713. Widal. iSoc. med. des Hop., 1896, p. 561. ^Presse Med., 1897, I., p. c. Ziemke. Deut. med. Woeh., 1897, p. 234. « 1- 3 CHAPTER XV. MISCELLANEOUS INFECTIOUS DISEASES. SEPTICEMIA, PYEMIA, OSTEOMYELITIS. Red Cells. — In no other diseases do the red cells suffer destruction so constantly and to such an extent as in the toxemia of diffuse inflam- mation caused by the common pyogenic bacteria. This fact was very early noted, Braidwood finding shrinkage of the red cells and absence of rouleaux, Mannassein remarking on their reduced diameter, and Quincke and Patrigeon observing extreme loss of Hb in cases of py- emia. Hayera and Toenissen placed the average loss of red cells in ordinary septic fever at 200,000 to 1,000,000 per week, while a continuous dimi- nution was found to persist as long as suppuration continued. That the loss of red cells is often very much more ra})id than Hayem sup- posed is shown by Grawitz's remarkable case, in which fatal puerperal sepsis combined with considerable hemorrhage reduced the red cells in about twenty-four hours to 300,000. Various forms of acute septicemia not infrequently reduce the red cells below two millions, but none appear to act more violently than does puerperal or uterine sepsis. Hayem ^ found only 1.45 million cells, 20 percent Hb, in a recent case of puerperal sepsis. Cabot reports 1,8 million cells from a "suppurating fibroid." The writer found L6 million cells, and 20 percent Hb, in a septic form of endometritis, not puerperal. When the suppurative process is localized, the red cells are usually found to be only slightly affected, but with the first appearance of the " septic " condition the reduction of red cells promptly begins, Roscher finding evidences of diminished concentration of the blood within the first few hours. In empyema, suppuration in wounds, pelvic ab- scesses, appendicitis, peritonitis, etc., the red cells are seldom markedly reduced, while the usual effects of fever are present and polycythemia is often found. The Hb is, as usual, more susceptible than the red cells, and its loss is usually out of proportion to that of the red cells, the Hb-index being invariably low. Timotjewsky found that nucleated red cells appear promptly in the blood of dogs after moderate injections of pyogenic bacteria, in one in- stance reaching the enormous proportion of 25,698 per cmm. Turk found considerable numbers of nucleated red cells in two cases of sep- ticemia from cystitis. In cases of intense puerperal sepsis the writer CHEMISTRY. 281 has usually been able to find a few nucleated red cells, but they have never been very numerous. In chronic suppuration the red cells tend to diminish as long as the discharge continues. In a case of chronic empyema of one year's duration the writer found 1.8 million cells and 25 percent of Hb, but an old pelvic abscess discharging a small quantity of pus for two years had induced very slight anemia. Morphological changes in the red cells in septicemia are usually present in moderate degree. Most cases show a pure type of secondary chlorotic anemia with marked loss of Hb. The reduced diameter of the anemic cells, previously noted, was distinct in the writer's case of septic endometritis, but did not approach the grade seen in some forms of primary disease of the blood. In severe cases many cells show granular degeneration, but polychromasia has, in the writer's experi- ence, not been prominent. Marked differences in the size and shape of the cells do not appear, as a rule, till after the second or third week of a severe septic process. Leucocytes. — A considerable number of cases of asthenic septicemia run their course without leucocytosis, or with distinct reduction of white cells. Such cases have been reported by Limbeck, Krebs, Cabot, and Turk, and proving invariably fatal, their significance is similar to cases of pneumonia, diphtheria, etc., with hypoleucocytosis. In Turk's case, and in one observed by the writer, while the leucocytes were sub-normal, the proportion of polynuclear cells was very high. The great majority of cases, however, show pronounced leucocytosis, which is usually in proportion to the severity of the disease. Yet among the reported cases showing leucocytosis are some, like Rieder's, in which very severe processes either failed at any time to cause marked increase of white cells or else the examination of the blood was made during the ante-mortem decline in their numbers. Leuco- cytosis in septic processes has, therefore, the same significance as in pneumonia and diphtheria. Regarding the more minute factors determining the grade of leuco- cytosis little is definitely known, and the examination of the blood must be interpreted according to the circumstances in each case. While the rule that suppuration induces leucocytosis is almost invaria- ble, it must be remembered that leucocytosis may promptly disappear when the exudation ceases, and that suppurations involving mucous surfaces may induce ve7'y slight leucocytosis. The polynuclear leucocytes are almost always in high proportion in cases of sepsis, but Klein has recently described a case of hemorrhagic septicemia^ with 76 percent of eosins in the pleural exudate and 40 percent in the blood. Chemistry. — The loss of Hb has already been mentioned as one of the chief alterations in the blood of septicemia. Its solution in the plasma and the occurrence of hemoglobinemia is observed in severe cases with rapid destruction of blood. In cases of puerperal sepsis, or pyemia in infants, the solution of Hb may reach a very extreme grade. 282 MISCELLANEOUS INFECTIOUS DISEASES. In a puerperal case the writer found the viscera discolored, and so great was the deposit of pigment that the liver and spleen closely re- sembled, m gross and microscopical appearance, the organs of pernicious malaria. The increased globulicidal action of the serum in these cases, to which the solution of red cells must be referred, has been described by Maragliano and others. An increased tendency toward crystallization of the Hb has been noted by Bond who, in examining fresh specimens, found a rich de- posit of such crystals about the edge of the cover-glass. Roscher noted in severe cases a very rapid lowering of the specific gravity of the blood, beginning a few hours after the initial symptoms. In fatal cases the dry residue of the whole blood fell to 15 percent, while in favorable cases the loss of solids was usually much less. This rapid and extreme loss of albumens Roscher and Grawitz regard as a valuable prognostic sign. The serum also usually showed a pro- nounced loss of albumen, proportionate to the severity of the septic process, its dry residue falling from 10.5 percent to a minimum of 6.25 percent. Yet in a very severe case the dry residue of the serum was 13.1 percent of its weight, which Grawitz explains as resulting from the solution of Hb, In various forms of pyogenic infection Livierato found an excess of glycogen in the blood, even when the temperature was normal. Gold- berger and Weiss also describe in the leucocytes during abscess- for- mation a reaction to iodine either in the form of a diffuse stain, or a granular deposit. They also find many brownish staining extra-cellu- lar granules after fractures, and believe that the somewhat obscure " iodine reaction " of the blood may be made of value in the diagnosis of abscess, fractures, etc. Bacteriology. — The frequency with which specific bacteria can be isolated from the circulating blood in cases of septic infection, in spite of the vast amount of study devoted to the subject, still remains a matter of doubt. There are, on the one hand, a large number of studies of the blood in pyemia, septicemia, puerperal fever, osteomyelitis, etc., reporting the discovery of bacteria in a large proportion of cases, while in many others, reported by competent observers, very uniformly negative re- sults were obtained. On reviewing many of these reports, it appears that the majority of them must be set aside as unreliable. Rosenbach, Garrod, Raskin, Brunner, Blum, Czerniewski, Eisel- berg, Cautu, Roux and Lannois, Bommers, Stern and Hirschler, and others, all used blood squeezed from the finger, and when they at- tempted to demonstrate the same germ in the viscera after death, were not always successful. Many of their successful cultures were made only a few hours before death. AVhen accurate details are given, as by Czerniewsky, it appears that contaminations by clearly non-patho- genic germs frequently occurred. On the other hand, Brieger, drawing blood from a vein by a steri- APPENDICITIS. 283 lized syringe, obtained negative results in five cases of puerperal sepsis. Sanger, by the same method, in three cases of osteomyelitis found Staph ijlocoGciis aureus or albus in the blood, but cultures from an in- flamed joint gave only Streptococcus pyogenes. Likewise Kraus re- ports six positive results in ten cases of puerperal sepsis, but one was examined post-mortem, and two showed only Staphylococcus albus. Cannon, who secured a number of successful cultures from the blood in puerperal sepsis, osteomyelitis, wound infection, etc., took some of his specimens from a puncture of the skin, and Petruschky, who usually obtained some bacteria, drew the blood through the "disinfected" skin by means of a wet cup. Sittmann, accepting as genuine the results of previous investigators, and adding nine positive cases of his own, concludes that in septico-pyemia the bacterial agent is always to be found in the blood, and that its presence does not insure a fatal issue. Yet later observers, as Kuhnau, would not accept the conclusions of Sittmann and do not admit much significance in the isolation from the blood of Staphylococcus albus or the colon bacillus. Kuhnau examined the blood repeatedly in 23 cases of septico-pyemia, 19 of which were fatal. In 2, Streptococcus pyogenes was obtained, in one Staphylococcus aureus. Several cultures gave non-pathogenic germs, among them Staphylococcus albus, which was obtained from the blood but not from the abscesses after death. The only successful cultures were made shortly before death, or in one case during a chill. The results of Kuhuau's study cannot fail to raise the suspicion that previous investigators have been too lax in technique, or have not fully identified their species. From the review of the above studies the writer draws the following conclusions : 1. The only reliable method of obtaining blood for bacteriological examination during life, is to draw it from a vein, in quantity not less than 5 cc, through the thoroughly sterilized or, better, the incised skin, by means of a sterile syringe or aspirator. 2. In the great majority of cases of local or general septic infection, septico-pyemia, septicemia, pyemia, diffuse suppuration, osteomyelitis, etc., bacteria are present in the circulating blood only for short periods and at infrequent intervals, most frequently during chills. 3. A few hours before death from septic infection, various bacteria, some of which may not be active in the original process, make their way into the circulation. 4. In a very moderate number of cases of septic infection, especially those which are not attended by local abscess formation, the bacterial agent may be isolated from the blood during the progress of the dis- ease. APPENDICITIS. Special interest attaches to the examination of the blood in appendi- citis, and Cabot, in a study of 72 cases, has pointed out conditions in 284 MISCELLANEOUS INFECTIOUS DISEASES. which the information thus gained may be of great value in diagnosis. Quoting largely from his observations it appears that : A catarrhal appendicitis is rarely accompanied by leucocytosis, one case showing 14,000 cells. The majority of more severe cases, on the second to fourth days show leucocytosis, from 15,000 to 24,000. More marked leucocytosis (maximum 52,000) was always associated witii large abscesses or peritonitis. Twelve of 15 cases with less than 15,- 000 cells recovered without operation or would probably have done S(» as no pus was found by the surgeon. Four cases with general purulent peritonitis failed to show leucocytosis, and other cases with peritonitis gave slight leucocytosis (14,800, 16,000). Other cases illustrated the fact that when pus ceases to be exuded the leucocytosis may subside. After successful operations the leucocytosis usually declined, but was subject to considerable oscillations, not always referable to collections of pus. It therefore appears that the interpretation to be placed upon the examination of the blood in appendicitis is extremely variable. With leucocytosis ranging between 14,000 and 16,000 cells, the condition may be : (1) Catarrhal appendicitis, (2) perforative inflammation with abscess, (3) large abscess, (4) general peritonitis. Moreover, the character of the leucocvtosis in those dano^erous forms of the disease which pursue a mild course for a time but suddenly develop peritonitis from perforation is not yet clear. It appears that the blood, as well as most other symptoms, equally fails to give warning of this outcome. Nor can the question of operation in frank cases be decided from the blood examination, as one of Cabot's cases with 33,000 leucocytes re- covered without interference. In differential diagnosis a positive conclusion seems to be warranted only in those cases which show well-marked leucocytosis, from the presence of which it is possible to rule out nearly all forms of typhoid fever and most but not all cases of fecal impaction. Hubbard has offered some very practical comments on the limitations of blood ex- aminations in this field. ABSCESS FORMATION. Active suppuration in a confined space is almost invariably accom- panied by distinct leucocytosis. Cabot has remarked on the high leu- cocytosis usually observed with very limited suppurative foci in sub- cutaneous, submucous, or interstitial connective tissues, and reports very considerable increase of white cells, with furuncle, carbuncle, and abscess of vulva, vagina, middle ear, uterus. Fallopian tube, ovary, lung, brain, parotid gland, neck, and in pectoral, psoas, and perinephritic regions. The writer's experience in a considerable variety of such conditions fully accords with these reports, leucocytosis having been present in the active stages of all such processes, but subsiding slowly or rapidly after operation or after the exudative process had ceased. An important exception to the rule was found in abscess of the liver with muco-purulent exudate. (See Liver.) ACUTE RHEUMATISM. 285 ERYSIPELAS. A slight elimination of red cells and Hb was observed in severe cases by Malassez, Hayem, and Reinert. In mild cases the fibrin is not affected, but in severe forms of the infection its formation is much increased. (Turk.) Leucocytosis occurs in nearly all cases, but Rieder reported one in- stance of facial erysipelas with a temperature of 39° C. and 6,800 white cells. Turk observed one case of considerable extent with tem- perature 38.5-40.2° and leucocytes between 7,000 and 8,900. Later the temperature rose to 40.5° and the leucocytes to 10,100. Zappert, also, observed two febrile cases with subnormal numbers of white cells (5,500-6,500). Hayem found between 7,000-8,000 leucocytes in very mild cases, but 12,000-20,000 when the rash was extensive. Most of the cases of Pee, Rieder, Ehrlich, Limbeck, and Cabot, gave leucocytosis of moderate grade but in proportion to the severity of the disease. Higher leucocytosis was observed by Halla in uncomplicated facial erysipelas and in a case complicated by pneumonia (23,000). Suppuration raised the count to 39,600 in a case of Reinert's, and to 59,400 in one of Epstein's. When the total increase is considerable the polynuclear cells are much increased in proportion, but with slight increase the polynuclear cells are usually unaffected. Eosins are usually diminished or absent. (Zappert, Turk.) Negative bacteriological examinations of the blood in erysipelas are reported by Petruschky in three cases and by Kraus in one case. ACUTE RHEUMATISM. A much increased formation of fibrin has been noted by many ob- servers, Halla, Hayem, Berggrun, Turk. Red Cells. — It is a general clinical rule that patients who pass through an attack of acute rheumatism become distinctly anemic. In many cases of ordinary severity the loss of red cells is slight, while the Hb falls to 65-80 percent. In more severe cases Hayem seldom found more than 4 million cells. Sorensen found an average of 4.16 millions in 8 cases. \\\ prolonged and relapsing cases the red cells not infrequently fall slightly below 4 millions. Yet in few of the reported studies are the examinations sufficiently extensive to show that such anemia is referable to the acute disease. Turk, while admitting the usual presence of anemia even during the febrile period, saw a distinct increase in red cells beginning with defervescence, and in Cabot's cases it does not appear either that there was any uniform loss of red cells or that the cases of long stand- ing were invariably the more anemic. It seems, therefore, that the severity of post-rheumatic anemia has been overestimated. The Hb suffers more severely and much more constantly than the red cells. Cabot's average in 31 cases was 67 percent, and a few cases 286 MISCELLANEOUS INFECTIOUS DISEASES. fell below 60 percent, while Turk found between 60 and SO percent. Prompt use of salicylates prevented the loss of Hb in a case of Lichten- stern's, and apparently also in one of Cabot's. In convalescence the restoration of Hb remains considerably behind that of the red cells. Leucocytes. — In mild cases without exudation there is usually no distinct leucocytosis. (Pee, Rieder.) When fever and swelling of joints exist there is almost always an increase to 10,000 or 15,000 cells. Turk insists that when the leucocytes reach 20,000 or more there are nearly always complications, such as pleurisy, pericarditis, or pneu- monia. Hayem also found 17,000-18,000 cells in moderately severe attacks, and as high as 25,000, only in extremely severe and cerebral types of the disease. Cabot reports 21,000-31,000 cells in 6 cases, one complicated by acute endocarditis, while in the others no complica- tion was mentioned. The writer, from the examination of 40 cases, in 1893, can support Turk's statement, having found signs of pneumonia, or pericarditis, or hyperpyrexia, whenever the leucocytes rose above 20,000. The pneumonic signs however were not always those of com- plete consolidation. With defervescence the leucocytes promptly fall to normal, and in relapses are much less affected than by initial attacks. (Turk.) Types of Leucocytes. — When the white cells do not greatly exceed 10,000, the proportions of the various forms are not much disturbed. With distinct leucocytosis the proportion of polynuclear cells rises to a considerable height. Eosins are absent only in the early stages ; later, in spite of fever and exudation they are always present in mod- erate numbers, while after defervescence most cases show a distinct tendency toward eosinophilia. (Turk.) In one of Turk's cases there were 13.8 percent of eosins, and in another 8.33 percent, shortly after defervescence. This observer believes that a high proportion of eosino- phile cells during the febrile period is a good prognostic sign, and oc- curs principally in self-limiting cases. Bacteriology. — In several studies, reviewed by Sittmann, various pathogenic bacteria were obtained from the circulating blood, but these cases have later been classed as examples of septicemia and not of articular rheumatism. The principal bacteriological studies of the blood of acute rheumatism are those of Sittmann, Singer, Kraus, and Kuhnau. Sittmann obtained negative results from repeated cultures in 5 cases, and Kraus in 12 cases. Singer conducted an elaborate study of the blood and urine in 60 cases, but while he found Staphylococcus a/6M.s in several instances, this germ was probably a contamination from the skin. In one case Streptococcus pyogenes was isolated, but the history shows the patient to have been suffering from hemorrhagic septicemia and endocarditis. Even more conclusive were the totally negative results obtained by Kuhnau in 67 cases representing all types of the disease and many complications. There seems therefore to be no longer any room to doubt that the blood in acute articular rheumatism is sterile. INFLAMMATIONS OF SEROUS MEMBRANES. 287 TONSILLITIS. Follicular tonsillitis usually causes moderate leucocytosis, seldom rising above 15,000 (Halla, Pick, Pee), yet in mild cases it may be absent (Cabot). In phlegmonous tonsillitis more marked leucocytosis, reacliing 20,000 cells or more, usually occurs. (Pee, Eieder.) The leucocytosis is generally more marked than in diphtheria with equal constitutional disturbances. (Pee.) WHOOPING-COUGH. Meunier has observed, in 30 cases of pertussis in children of various ages, a pronounced or extreme leucocytosis which "/«/• exceeds the in- crease found in any other afebrile disease of the resjyiratory passages J^ The grade of leucocytosis varies with the age of the patient, being most marked in children under four years, and usually reaching a slightly lower figure in children from four to seven years. It appears in the catarrhal stage before the characteristic cough, when its demon- stration may be made of diagnostic value, and disappears slowly with the improvement of the disease. It is little influenced by complica- tions. The average leucocytosis was 27,800, but in several cases 40,- 000 cells were found, and in one 51,150. The lowest figure obtained during an active period was 15,500. The proportion of lymphocytes was always high, averaging 53.8 in a portion of the cases, while the polynuclear cells maintained a low average (39 percent), and the eosins were scarce or normal in number. Meunier refers the lymphocytosis to hyperplasia of bronchial lymph nodes. De Amicis and Pacchioni also report marked leucocytosis in whoop- ing-cough, beginning in the first days of the disease, reaching its height in the spasmodic stage, and being sometimes prolonged after the cessa- tion of the typical cough. Cabot found 12,600 white cells, and 78 per- cent Hb in a girl of 6 years. INFLAMMATIONS OF SEROUS MEMBRANES. PLEURISY, PERICARDITIS, PERITONITIS. Serous inflammations of the large serous membranes usually cause a slight increase of leucocytes in the blood during the acute febrile stages. Thus, Hayem found between 7,500 and 12,000 leucocytes in acute pleurisy, liieder 11,000-13,000, and some of Cabot's early cases regis- tered as high as 15,000. Higher leucocytosis belongs to the more severe inflammations, with temperatures reaching 101—104° F. In two cases of sero-fibrinous pleurisy Limbeck found, with a temperature of 38-8° C, 18,000-19,000 leucocytes, and with a temperature of 41.5° C, 22,000 cells. The grade of leucocytosis excited by fibrinous pleu- risy, pericarditis, etc., has apparently not been determined. The in- flammation stands in an intermediate position in point of intensity. 288 MISCELLANEOUS INFECTIOUS DISEASES. With purulent processes the leucocytosis is usually much higher. In 1893 the writer found no diiference in the leucocytosis of pneuraonia from that of idiopathic empyema and purulent pericarditis. All the cases of pericarditis were complicated by pneumonia or rheumatism and gave very high counts (maximum 60,000). The writer has seen one case of very rapidly fatal empyema, yielding Streptococcus pyogenes in cultures, in which there was scanty purulent exudate and slight and diminishing leucocytosis. In purulent peritonitis, also, leucocytosis is not infrequently absent. After the exudative process has ceased the leucocytosis subsides, very rapidly in the case of serous exudates and usually more slowly with purulent processes. The majority of serous effusions are there- fore encountered, as in Halla's and Cabot's cases, when the leucocytes are normal in number, nor is it rare to find extensive empyema with no excess of white cells in the blood. Tuberculous inflammations of serous membranes, when pure and un- complicated, are seldom accompanied by leucocytosis. Limbeck reports 4,600 cells in a case of tuberculous peritonitis. In serous pleurisy it is seldom possible to distinguish the quiescent stages of simple serous effusions from beginning tuberculous exudates, and the examination of the blood cannot prove of much value in diagnosis. In a case of empyema in which tubercle bacilli were extremely numerous, the writer found moderate intermittent leucocytosis (maxi- mum, 18,000). GONORRHEA. In acute gonorrheal urethritis the red cells remain unaffected. A moderate leucocytosis is usually observed in severe febrile cases, and is markedly increased by many complications, including cystitis, epididy- mitis, orchitis, etc. The poly nuclear cells are in the usual excess, while eosins, although sometimes very abundant in the urethral discharge (Bettman), usually remain within moderate limits (.5 to 11.5 percent) in the blood (Vor- bach). Bettman, however, while finding that there is no relation between the eosins of the discharge and those in the blood, believes eosinophilia is usually present in gonorrhea, especially in posterior urethritis. He found 25 percent of eosins in the blood in a case of gonorrheal epididymitis. In gonorrheal rheumatism the changes in the blood are similar to those of the idiopathic joint affections. The gonococcus has several times been isolated from the blood in gonori'heal endocarditis (q. v.). YELLOW FEVER. From Jones' interesting observations about yellow fever, it would appear that anemia is infrequent, that fibrin formation is deficient, that EPIDEMIC INFLUENZA. 289 the globulicidal action of the serum is very greatly increased, and that cholemia is responsible for some of the changes observed in the red cells before and after shedding. " When a drop of blood from a yellow fever patient falls upon blotting paper, a dark brownish ring due to diffusion of bile-stained plasma spreads about the central mass of blood cells." (Jones.) Pothier examined in some detail the blood of 154 cases. The red cells never fell below 4.28 millions and in one fatal case were normal. The Hb suffered a considerable loss during the course of the disease, registering between 50 and 72 percent. In the above fatal case 90 percent of Hb was found. The restoration of the Hb was slow. Morphological changes in the red cells were usually absent, but the presence of normoblasts was noted in a few specimens. The leucocytes fell between 4,660 and 20,000. The polynuclear cells were found in high proportions in some of the slides from these cases, examined by Cabot, while in others their proportions were normal. Eosins were very scarce and myelocytes were found in one case. The writer examined dry specimens of the blood of two rapidly fatal cases of yellow fever brought from Rio Janeiro by Dr. J. M. Masury, and noted slight hemoglobinemia, and hypoleucocytosis. Pothier tested the action, of the serum of some of his patients on cul- tures of Sanarelli's bacillus, but the results did not indicate the presence of any specific power in the blood. TYPHUS FEVER. In 4 cases the writer fouud between 5,000 and 9,000 leucocytes. The patients were adults, the examinations were made during the high fever of the early period of the disease, and at least two of the patients died. Tumas followed one fatal case during a three weeks' course, observ- ing a steady reduction of red cells from 4.4 to 3.1 millions, of Hb from 80-50 percent, and hypoleucocytosis, 9,600-1,600. EPIDEMIC INFLUENZA. The scanty reports of the condition of the blood in epidemic influ- enza indicate that the disease in uncomplicated form fails to cause leu- cocytosis. This fact, while in accordance with the catarrhal character of the essential lesions, is somewhat difficult to reconcile with acute in- fectious nature of the disease. Grippe appears to be the only bacterial disease, beginning acutely with marked chill, which fails to induce leu- cocytosis. The absence of leucocytosis in grippe was first demonstrated by Rieder" who, in 7 cases, found that the white cells at the acme of the disease were reduced in number (7,000-2,800). In the catarrhal pneu- monia which complicates the disease and which is distinguished by signs of incomplete consolidation, he found little or no leucocytosis, 19 290 MISCELLANEOUS INFECTIOUS DISEASES. maximum, 13,000. In lobar pneumonia following grippe the usual leucocytosis was observed. In 1893 the writer found no difference in the leucocytosis of lobar pneumonia following grippe from that of pri- mary pneumonia. Cabot reports the examination of the blood in 67 cases, most of which showed normal or reduced numbers of white cells, and when there was slight leucocytosis (maximum 14,000) complica- tions were usually present. Bacteriological Examination. — In a considerable proportion of cases Canon, Klein,^ Hirschfeldt, Bruschettini, and others, have ob- tained from the blood cultures of what they believed to be PfeiflPer's bacillus. Pfeiifer and Kruse, and others were unable to isolate the bacillus of influenza from the blood, and denied the identity of some of the germs obtained from the blood by others. Recently Kuhnau has added 12 negative results, and there seems little reason to doubt that the bacillus of influenza has no special capacity to invade the blood stream. TETANUS. Cabot reports 70 percent of Hb, 11,900 leucocytes, and persistence of eosinophile cells, in a fatal case of tetanus treated by antitoxine. BUBONIC PLAGUE. Aoyama in 1895 examined the blood of many cases, reporting four in full. The red cells varied from 4.4 to 8.1 million, the latter in a case which recovered, and in the report of which no cause of polycy- themia is apparent. The Hb was normal. The leucocytosis also was reported to have reached a very high grade, 110,000-200,000 in three cases, and 20,000 in a convalescent case. Polynuclear cells made up the bulk of the increase, but large and small lymphocytes were numer- ous, and eosins scarce. The Austrian Pest-Commission reported only moderate leucocytosis in the majority of cases. (Wien. klin. Woch., 1897, p. 465.) Zabolotny studied the reaction of the blood serum on cultures of the bacillus. During the first week no effect was observed even with a dilution of 1-6 ; in the second week reactions were obtained with dilutions of 1-10; in the third week, with 1-25 ; in the fourth week, with 1-50. The reaction was most marked in the severer cases, and the blood serum of the cadaver was inert. MALTA FEVER. The hemorrhages which mark the severer forms of this malady com- monly give rise to severe anemia. Bruce places the average number of red cells at 3.5 millions, and reports the absence of leucocytosis and occasional presence of free pigment in the blood. In a prolonged case without hemorrhages, Musser and Sailer found 5.05 million red cells, 60 percent Hb, 11,564 leucocytes, all varieties being in normal pro- portions. BIBLIOGRAPHY. 291 Wright and Smith found that the blood serum of cases of Malta fever agglutinates cultures of Bacillus melitensis, and diagnoses of the disease have been made by this method by Miisser and Sailer, Cox, and others. A dilution of serum 1—50 should give a prompt reaction, while typhoid serum has no eifect. ACTINOMYCOSIS. The writer found 21,500 leucocytes in a case of pulmonary acti- nomycosis, clinically resembling acute phthisis, and Cabot reports a case of actinomycosis of liver with 3,700 white cells. The markedly purulent character of the exudate excited by the ray fungus explains the grade and character of the leucocytosis. Bier- freund found marked chlorotic anemia in actinomycosis, the Hb regis- tering 30-50 percent. GLANDERS. Cabot reports 11,600-13,600 leucocytes in a fatal case of very acute glanders. Duval isolated the bacillus of glanders from the blood dur- ing life, and Noniewitch reports that in fatal cases in horses, the bacilli can be found in the circulating blood, usually wdthin the leucocytes. ANTHRAX. Anthrax in some lower animals frequently develops into a pro- nounced bacteriemia early in the disease. Man, however, usually ex- hibits only a moderate grade of susceptibility to the infection and the majority of cases recover after excision of local foci of infection. Al- though it is probable that in the severe septicemic forms of the disease anthrax bacilli may multiply in the general blood stream early in the disease, the literature appears to contain very few reports of competent bacteriological analyses of the blood in such conditions. The rather frequent impression that the diagnosis of anthrax can be made by morphological examination of a drop of blood from the general circu- lation is almost entirely without foundation, as the great majority of cultures of the blood, even in severe cases, have been found sterile. Only in the late stages of some septicemic cases in man do the bacteria invade the circulation and yield positive cultures of the blood. Such a case is that of Blumer and Young, who report successful cultures, and claim to have identified the germ in blood-smears during life, but the time before death was not stated. Ante-mortem invasion of the blood by anthrax in man is apparently not much if any more abundant than with some other infections, and even post-mortem cultures are frequently negative. Detailed clinical examinations of the red cells and leucocytes in the blood of anthrax are still wanting. Bibliography. Miscellaneous Infectious Diseases. Aoyama. Mittheil. a. d. kaiserl. Japanisch. Univers., 1895, Ed. III., No. 2. Bein. Zeit. f. klin. Med., Bd. 17, p. 545. 292 MISCELLANEOUS INFECTIOUS DISEASES. Bettman. Archiv f. Derm, et Syph., Bd. 49, p. 227. Bierfreimd. Arcliiv f. C'hirurg., Bd. 41. Blum. Miinch. med. Woch., 1893, p. 297. Blumer, Young. Johns Hopkins Bulletin, 1885, p. 127. Bommers. Deut. med. Woch., 1893, p. 552. Bond. Lancet, 1887, II., pp! 509, 557. Braidwood. On Pyemia, etc., London, 1868. Brieger. Charite-Annalen, 1886, p. 198. Bruce. Lancet, 1892, II., p. 1101. Annal. de I'lnstitut Easteur, 1893. Brunner. Wien. klin. Woch., 1891, p. 391. Bruschettini. Kit'. Medica, 1893, No. 186. Cannoa. Deut. Zeit. f. Chirurg., Bd. 33, p. 571. Canrm. Virchow's Archiv, Bd. 131, p. 401. Ointu. Kif. Med., 1892, II., p. 243. Cox. Philadelphia Med. Jour., Vol. 4, p. 491. Czerniewsky. Archiv f. Gynecol., 1888, p. 73. Be Amicis, Pacchioni. Ref. in Amer. Jour. Med. Sci., Vol. 119, p. 599. Duval. Archiv de Med. exper., 1896, p. 361. Ehiiich. Charite-Annalen, Bd. 12, p. 288. Eiselberg. Wien. klin. Woch., 1890, p. 731. Epstein. Allg. Wien. med. Zeit., 1889, No. 43. Friedrich. Arbeit, a. d. kaiserl. Gesundheitsampte, 1890, p. 250. Garrod. Fort. d. Med., 1885, p. 165. Ooldberger, Weiss. Wien. klin. Woch., 1897, p. 601. Halla. Zeit. f. Heilk., Bd. 4. Hayem. ' Du Sang. ^ Med. Moderne, 1897. Hirschfeldt. Baumgarten's Jahresber., 1893, p. 206. Hubbard. Boston Med. and Surg. Jour., Vol. 142, p. 409. Jones. Jour. Amer. Med. Assoc, Vol. 24, p. 402. Klein. iCent. f. inn. Med., 1899, p. 97. ^Baumgarten's Jahresber., 1893, p. 206. Kraus. Zeit. f. Heilk., Bd. 17, p. 117. Krebs. Diss. Berlin, 1893. Kruse. Cent. f. Bact., Bd. 7, p. 657. Kuhnau. Zeit. f. Hygiene, Bd. 25, p. 492. Livierato. Deut. Archiv f. inn. Med., Bd. 53, p. 303. MalasAcz. Archiv de Physiol., 1874, p. 32. Manassein. U. d. Verand. in d. Dimens. d. rot. Blutkorp., Tubingen, 1872. Maragliano. Cong. f. inn. Med., 1892. Meunier. Compt. Rend. Soc. Biol., 1898, p. 103. Musser, Sailer. Philadelphia Med. Jour., 1898, p. 1408. Nonieviitch. Baumgarten's Jahresber., 1897, \). 415. Pairigeon. Recherch. s. 1. nombre d. glob, rouges., Paris, 1877. Pee. Diss. Berlin, 1890. Petruschky. Zeit. f. Hygiene, Bd. 17, p. 59. Pfeiffer. Deut. med. Woch., 1893, p. 816. Pothier. Jour. Amer. Med. Assoc, Vol. 30, p. 884. Quincke. Virchow's Archiv, Bd. 54, p. 537. Raskin. Cent. f. Bact., 1889, pp. 286, 433. Beineri. Zahlung d. Blutkorp, etc Rieder. ' Leucocytose. ^Miinch. med. Woch., 1892, p. 511. Roscher. Blut. bei septisch. Fieber, Diss. Berlin, 1894. Rosenbach. Microorg. bei Wundinfection, Wiesbaden, 1884. Roux, Lannois. Rev. de Med., 1890, p. 1011. Sanger. Deut. med. Woch., 1889, p. 148. Singer. Aetiol. d. acut. Gelenkrheumat. Siitmann. Deut. Archiv f. klin. Med., Bd. 53, p. 323. Spronek. Semaine Med., 1898, p. 393. Stem, Hirscfder. Wien. med. Presse, 1888, p. 1102. Timofj'wsky. Cent. f. Path., 1895, p. 108. Toenissen. Inaug. Diss. Erlangen, 1881. 'Turk. Klin. Blutuntersuchungen Wien., 1897. Vorbach. Diss. Wurzburg, 1895. Wright, Smith. Lancet, 1897, I., p. 656. Zabolotny. Compt. Rend. Soc. Biol., 1897, p. 520. CHAPTER XVI. SYPHILIS, TUBERCULOSIS, LEPEOSY. SYPHILIS. Very accurate conclusions regarding the state of the blood in syphilis were reached by Becquerel and Rodier, who stated that a moderate grade of anemia was to be found in the great majority of cases, that as long as the disease progressed without complications the blood maintained a high or normal standard, and that if the course of the disease was pro- longed, there was a loss of red cells. They noted also that the abuse of mercury might lead to the same changes. The resemblance of the anemia of syphilis to that of chlorosis was also early recognized, especially by Ricord, and this belief was later supported by analyses of the blood by Grassi, Wilbouche witch, Keyes, Laache, Malassez,^ Gaillard, and others, who found in the secondary and tertiary stages of most cases a loss of red cells and Hb. Keyes was the first to point out that mer- cury in small doses, while curing the disease, increases the red cells. The grade of anemia observed was usually moderate, the majority of cases showing slight loss of red cells, which however sometimes fell to 3-4 millions. The Hb was usually found diminished more than the cells, especially by Lezius who claimed that the only essential lesion of the blood in syphilis is a loss of Hb. Other patients appeared to have normal blood, and Sorensen in 20 cases failed to detect any distinct anemia. On the other hand, Muller reported cases in which all the typical lesions oi pernicious anemia were present, and in one of his cases the red cells fell to 428,100. Other such cases are reported by Ponfick, Kjerner, Klein, Laache, Fisischella (cited by Dominici). It was thus established that syphilis in some stages and in some patients may entirely fail to reduce the red cells, while in others grave pernicious anemia must be charged to its action. The later studies have been concerned with the more detailed course of the blood changes in the disease, and have been contributed principally by Anc, Lezius, Bieganski, Konried, Rille, Loos, Justus, and Riess. From these contributions it appears that the blood suffers in a some- what uniform degree in the several stages of uncomplicated syphilis, but unfavorable conditions and abuse of mercury may greatly aggravate the changes observed. Primary Stage. — During the first four to seven weeks after infec- tion the red cells do not diminish perceptibly in number (Lezius, Konried, et al.) unless there is fever or some other disturbing factor, when a moderate decrease may be noted from the first (Stoukovenkoff). 294 SYPHILIS, TUBERCULOSIS, LEPROSY. No doubt, as Hayem says, these disturbing factors are frequently present, and it still remains uncertain whether syphilis alone affects the number of red cells during the first few weeks. The majority of observations indicate that it does not, but Riess came to a contrary conclusion, while Bieganski found uniform polycythemia in the early stages of syphilis. The Hb is almost invariably diminished from the first, a loss of 15-30 percent being commonly noted before the appearance of second- aries. (Konried.) Secondary State. — There is uniform agreement among very numer- ous observers that with the outbreak of secondary symptoms the red cells begin to fall rapidly, reaching in untreated cases as low as 2 millions or less. (Konried.) In 10 cases Wilbouchewitch found an average decrease of 229,000 cells daily. At the same time the Hb continues to diminish and may fall to 55-25 percent within a few weeks or months. Riess denies that the Hb suffers particularly at the outbreak of the eruption. With the appearance of fever and new eruptions a further and more marked loss of cells and Hb has been observed. (Stoukovenkoff.) Under unfavorable hygienic conditions, and in feeble and especially in young subjects, the anemia in this stage may be unusually severe. In untreated cases the disappearance of the eruption is not followed by any immediate improvement in the blood. Tertiary Stage. — In untreated cases there can be no doubt that the anemia may progress till the pernicious type is established, but treat- ment usually limits the impoverishment of the blood so that only moderate grades of anemia are commonly observed. Konried found an average of 4 million red cells in ten cases, and from 50-80 percent of Hb in 22 cases suffering from gummatous lesions. In several cases showing advanced tertiary lesions (gummata) at autopsy the writer has found different grades and types of secondary chlorotic and pernicious anemia. In one the spleen was much enlarged and contained gummata, while the blood showed the lesion of secondary pernicious anemia, with a tendency toward the microcytic type, and with low Hb-index. In other cases the abundant megalocytes with increased Hb closely resembled those of primary pernicious anemia. It is especially in infants that syphilis induces the grave types of anemia. Eifects of Mercury Upon the Blood in Syphilis. — The effect of proper doses of mercury in arresting the progress of the anemia of syphilis was clearly stated by Wilbouchewitch, who reported an aver- age gain of 102,000 cells daily in 10 cases, while Keyes and many later observers have fully verified these results. That prolonged use of mercury may of itself lead to marked anemia resembling that of syphilis was early noted by Becquerel and Rodier, and was also dem- onstrated both in man and animals by Wilbouchewitch, and later by Bieganski, Hayem, Lezius, Anc, Schlesiuger, and Jelleneff. The ex- tent to which mercurial treatment may be carried without diminishing the red cells or Hb has been placed at 24 days, by Gaillard ; at 25-35 I JUSTUS' TEST IN THE BLOOD OF SYPHILIS. 295 inunctions, by Konried ; at 16 injections (gram .5, 1 percent Hg-ben- zoate) by Jelleneff ; and at 140—150 milligrams of bichloride, or 77 milligrams of benzoate, injected in increasing doses, by Lindstrom. Daring this period most observers agree that both red cells and Hb increase from the first, and many instances of moderate polycythemia are recorded. At the same time the eruption commonly disappears. If treatment is continued beyond this point, especially if the patient is salivated, the cells and Hb steadily and sometimes rapidly decline and well-marked mercurial anemia is established. Jawein, however, failed to observe any anemia during prolonged courses of inunctions. The Leucocytes in Syphilis. — Leucocytosis was early observed in syphilis and its usual connection with the hyperplasia of lymph nodes was one of the facts that led Virchow to locate the origin of the lymphocytes in these structures. More detailed study of the behavior of the leucocytes was made by Wilbouchewitch, Biganski, JelleneflP, Konried, and Rille, who have shown that the leucocytosis of syphilis is connected principally with the eruptions and- the anemia of the disease. In the primary stage, before the appearance of the eruption, the leucocytes are usually normal (Rille), but Jelleneif found that an in- crease of white cells usually precedes the development of anemia, and Konried's cases showed a slight leucocytosis (maximum 16,400) in the first weeks of the primary stage. In the secondary stage, with the appearance of eruptions, anemia, and hyperplasia of lymph nodes, the leucocytes are nearly always in- creased. In Konried's cases of untreated, secondary syphilis the white cells were never under 10,000, while the maximum figure was 17,500. Riess, however, found leucocytosis absent in many cases, but observed an increase to 20,000 in some instances, while the excess of lympho- cytes was usually very distinct (maximum 68 percent). The adminis- tration of mercury commonly reduces the white cells, which become normal with the disappearance of the eruption and the anemia. Jel- leneif found the leucocytosis to be more nearly proportionate to the extent of the eruption than to the size of the lymph nodes, and noted leucocytosis in the absence of external signs of lymphoid hyperplasia. The increase affects principally the small and large lymphocytes, but the eosinophile cells may also be increased, especially in those cases with marked papular exanthems. (Rille.) Zappert found a slight increase of eosins, 4.91 percent in one early case, but normal numbers in seven others. The proportion of lymphocytes diminishes as the patient improves, falling after each inunction. (Riess.) In very se- vere cases there may be progressive polynuclear leucocytosis. Myelo- cytes have been found by Rille and others in both secondary and tertiary syphilis. In tertiary stages with gummatous lesions the leucocytosis usually persists, but lymphocytosis, though sometimes distinct, is less constant. Konried found 8,500-17,710 leucocytes in nine cases. Justus' Test in the Blood of Syphilis. — By a series of observations 296 SYPHILIS, TUBERCULOSIS, LEPROSY. in over 300 cases Justus came to the conclusion that in the blood of florid syphilis after injections of moderate doses of various preparations of mercury, or after inunctions, there is- a period of a few hours or days during which the Hb is considerably reduced (10-20 percent). After a certain period, varying with the general condition of the pa- tient and the severity of the symptoms, the Hb begins to increase. In some cases the diminution continued only one day and in others it was repeated after each of three or four injections. The reaction was noted in all stages of the disease after the primary swelling of lymph nodes, Fig. 23. XI XII 16 1 18 19 20 21 22 23 24 25 2 5 27 28 29 30 1 2 3 4 5 6 7 8 9 10 11 12 13 U 15 2 \ / ■110 — , Vll / 9 V / / / / . II in V J y , VI ; / / / / / / ' 100 8 ■7 6 1 / 95 / i. 3 2 ' 1 HB. Je90 1 Course of changes in Hb in Justus' test. (Roman numerals indicate intravenous injections of sub- limate.) and in hereditary syphilis, but could not be obtained " during or shortly before the symptoms of the disease begin to disappear." In the serum, within two minutes after the intravenous injection of bichloride, Justus found the spectrum of oxyhemoglobin. In blood smears immediately after the injection many pale degenerating red cells were found, which disappeared within a few minutes. It thus appeared that the mercury had caused an immediate solution of red cells in the plasma. The specific gravity of the blood M'as not studied except in one case in which it showed irregular fluctuations. CONGENITAL SYPHILIS. 297 The value of Justus' test has not been fully determined. Cabot reported positive results iu seven cases of syphilis and negative in 32 control cases, but obtained a positive reaction in a case of chlorosis. Jones recently applied the test in 35 cases of syphilis in various stages, and in 18 controls, estimating the Hb usually by the specific gravity method. None of the control cases responded to the test. Of 17 cases of secondary syphilis 13 responded, and four failed. Of eight cases of chancre with adenitis, two responded to the test by Fleischl's method, but failed when tested by the specific gravity method. Six others Avere not tested by Fleischl's instrument. The conclusion was reached that a positive reaction has considerable diag- nostic value, while a negative result is of less significance. Jones' conclu- sions would be more reliable had he used the Fleischl instrument through- out. Further studies are required before the value of Justus' test can be shown, but on general grounds it would seem that the test v^ould prove unreliable in many forms of anemia, when the globulicidal action of mercury might dissolve red cells. CONGENITAL SYPHILIS. Syphilitic infants invariably suffer from anemia. In the mildest cases, Loos found over 5 million red cells and 65-75 percent of Hb. Schiff also found the usual polycythemia of infants in mild cases of congenital syphilis. In the majority of cases there is considerable loss of red cells and the Hb falls to a low figure, minimum 21 percent. (Loos.) In the severer forms the blood may show the changes of sec- ondary or progressive pernicious anemia, examples of which have been described by Loos, Luzet, Monti, Berggrun, and others. Whatever the grade of anemia, the changes are influenced by the special tendencies of infants' blood. Marked differences in the size of the cells are early established, nucleated red cells are frequently seen in abundance, and white cells, especially lymphocytes, are often present in greatly increased numbers. Loos noted an extreme degree of polychromasia in the large nucleated red cells in congenital syphilis. The leucocytes did not fall below 12,000 in any of Loos' 16 cases, and in one fatal case rose to 58,000. Baginsky, and Monti and Berggrun have also reported excessive leucocytosis in similar cases, while the former author speaks of a distinct leukemic tendency to be found in the blood of congenital syphilis. The lymphocytes usually fall Avithin normal limits for this age. Eosinophile cells are sometimes increased, especially when the eruption is extensive. (Loos.) The presence of a few myelocytes in the severe cases makes the resemblance of these cases of syphilitic anemia to v. Jaksch's anemia often very close, while the excess of large lymphocytes may lead to confusion of the disease with lymphatic leukemia of Frankel's type. Congenital syphilis is recognized by Demelin, as one of the common causes of melena neonatorum. In such a case, dying four days after birth from abdominal hemorrhage, the writer found in the blood the typical characters of idiopathic pernicious anemia, especially in the 298 SYPHILIS, TUBERCULOSIS, LEPROSY. large number of raegaloblasts. The marrow was everywhere lymphoid, and showed pronounced megaloblastic changes. TUBERCULOSIS. The fact that the blood of consumptives may often fail to show any changes comparable with the pallor of the skin and the emaciation of these subjects was noted in the first studies of the blood by Andral and Gavarret, Becquerel and Rodier, and others. The discrepancy between blood and facial appearance was fully recognized by Laache who, find- ing an average of 4.4 million cells in 14 cases, stated that phthisis of itself usually does not tend toward marked anemia. Similar results in the hands of Sorensen, Oppenheimer, Gnezda, Barbacci, and Reinert, confirmed this opinion in a considerable group of cases, although well- marked chlorotic anemia was occasionally encountered. In the Massa- chusetts General Hospital series, Cabot found 41 out of 60 patients with more than 4 million red cells, and 69 of 80 cases with more than 50 percent of Hb. Yet Malassez had in 1874 reported well-marked anemia in some early cases, had seen the red cells fall over half a million in the course of one week, and found less than a million cells in an advanced case with diarrhea. In nearly all of their series of cases Laker and Feno- glio found low percentages of Hb, and the specific gravity was found markedly reduced by Schmaltz and Peiper. In 1893 current views were presented by v. Noorden who stated that in pulmonary tuberculosis a loss of more than 20 percent of red cells or Hb is seldom observed unless from complications such as hem- orrhage, suppuration, or amyloid degeneration. The subject was at this time thoroughly reviewed by Grawitz ' and Strauer, who found three distinct periods in the changes of the blood in phthisis. 1. With beginning apical lesions, they generally found marked chlorotic anemia with loss of red cells, irregular leucocytosis, and reduc- tion in dry residue and specific gravity. In robust subjects however the blood during this stage was sometimes normal. 2. In pale emaciated subjects with chronic phthisis and cavities, but without much fever, the blood commonly did not vary from the normal standard, the red cells numbering 5 millions or more, the leucocytes 5,000-10,000, the dry residue of blood and serum being moderately low, but the specific gravity higher than the dry residue would indicate. 3. When hectic fever supervened, usually from suppuration in cavi- ties, and the patient began to suffer from septicemia, the blood became rapidly impoverished, the red cells were then much reduced, the leucocytes frequently much increased, and the dry residue and specific gravity falling to a low point. The intense septicemia of rapidly advancing phthisis was found to yield some very severe grades of anemia, in one of which the red cells fell to 700,000. Similar cases have been reported by Malassez ^ and TUBERCULOSIS. 299 Limbeck/ both of which suffered from marked diarrhea. In Lim- beck's case there was also tuberculous peritonitis, and the blood showed marked poikilocytosis and increased Hb-index. In the attempt to correlate his own and Robert's divergent results Dehio described two different types of the disease, the marantic and the anemic cases. In the former group the patients showed a marked tendency to emaciate while their blood retained a good standard, while in the others all grades of chlorotic anemia might be observed. It would seem that the classifications of both Dehio and Grawitz are based on accurate clinical study. The following conclusions may be drawn regarding the general course of blood changes in tubercu- losis. 1. The primari/ anemia of tuberculosis is seen not only in diseases of the lungs but quite as frequently in chronic tuberculosis of other tissues, especially of the lymph nodes and bones, constituting the very numerous group of cases of '' lymphatic anemia." Moreover, it has recently been placed beyond doubt that the chronic anemia of Hodg- kin's disease is in a large proportion of cases the result of tuberculosis of lymph nodes. The peculiar character of these cases of secondary tuberculous ane- mia has been described by Laker, Neubert, Wiskeman, Vierordt, Bier- freund. Brown, Dane, and many others, and they constitute one of the largest groups of cases encountered in routine blood examinations. In appearance the patients clearly resemble cases of chlorosis and the blood shows a slightly or considerably reduced number of red cells, more marked loss of Hb, and usually slight relative or absolute lym- phocytosis. Yet this description does not apply to all cases of early tuberculosis or phthisis, as some of them from the first appear anemic and yet show normal cells and Hb. 2. The second group in ivhich the blood is but slightly altered in quality, although cavities have formed in the lungs and the patient is pale and emaciated, includes the majority of cases of chronic phthisis. One of the most striking examples seen by the writer occurred in a patient suflPering from subacute tuberculous empyema from which the pus contained very large numbers of tubercle bacilli. But the same qual- ity of blood is sometimes seen with early apical lesions and moderate emaciation, while considerable anemia may exist in moderately ad- vanced cases of phthisis who are not suifering from distinct hectic fever. The cause of the concentration of the blood, or oligemia, in the average phthisical subject, for such must be the condition, is probably to be found, as indicated by Heidenhain, Gartner and Romer, and Grawitz,^ in the specific lymphogogic action of the toxines of the tubercle bacillus, by which there is established a continuous excess in the balance of fluids which leave the tissues through the lymphatics. This view is strongly supported by the resemblance which exists be- tween the tuberculous and the typhoidal processes, both showing a pe- culiar relation to lymphatic structures and both tending to concentrate the blood. Identical effects upon the blood follow also the experi- mental injection of tuberculin and typho-toxin. 300 SYPHILIS, TUBERCULOSIS, LEPROSY. By this means the considerable destruction of blood which results from the chronic toxemia, hemorrhages, and malnutrition of the dis- ease, is obscured, and only becomes visible at autopsies on these sub- jects when the shrunken appearance of the tissues finds a parallel in the diminished total volume of blood which is often apparent to the naked eye. In the inajority of cases of well-advanced phthisis, therefore, approximately normal blood indicates considerable absorption of the tox- ines of the tubercle bacillus. Prolonged and profuse night sweats, and severe diarrhea, doubtless have a similar effect, which, however, is usually overbalanced by the destruction of blood which results from the associated septicemia. Normal blood is found in many phthisical patients who do not sweat or suffer from diarrhea. Yet in some ex- treme cases one or both of these factors may very well prove to be the chief influence in the concentration of the blood. 3. The third group of cases includes those who are anemic from the first or who become anemic in the terminal stage of the disease, in either instance from the destruction of blood which occurs in all se- vere septicemias. Since the severe hectic fever of phthisis is largely referable to mixed infection, the condition established is not very dif- ferent from the ordinary type of pyogenic sepsis. The destruction of cells may be quite rapid. Malassez observed losses of 730,000 cells in one month, and 760,000 in three weeks in cases without hemor- rhage. Here belong the acute cases with grave anemia described by Malassez, Limbeck, and Grawitz, from which it appears that the blood in tuberculosis may develop much the same characters as in pernicious anemia. Effects of Complications. — Hemoptysis causes impoverishment of the blood in proportion to the extent of the hemorrhage. Malassez observed a reduction of 940,000 cells from small hemorrhages in a period of eight days. Amyloid degeneration is seen with advanced and anemic cases, but its effects on the blood have not been specially studied. Stenosis of the larynx, and diabetes, have been found with concentrated blood. (Grawitz.) Regeneration of the Blood in Phthisis. — Prompt regeneration of the blood after hemoptyses has been observed by Malassez, but such improvement must depend largely upon the general condition of the patient. Laker came to the conclusion that if the Hb did not steadily improve after operation on tuberculous foci, it might safely be con- cluded that all of the disease had not been removed. In seven of Bierfreund's cases and in two of Brown's, this rule held good, and three times a steady decline in Hb preceded for several weeks the develop- ment of symptoms of general tuberculosis {vide infra). In phthisis, as well as in other tuberculous processes, great caution must be used in judging of the patient's impi'ovement from an increase in red cells or Hb. The writer has seen the Hb and red cells increase while the patient teas rapidly losing flesh, the lesions advancing, and the total quantity of blood doubtless falling. In several of Bierfreund's tases the Hb steadily increased while the patient was developing general cuberculosis. THE LEUCOCYTES IN TUBERCULOSIS. 301 Morphological change.'', in the red cells are rather less marked than in most other types of secondary anemia, owing probably to the con- servative eiFect of a plasma of high gravity. When anemia exists it is usually of the simple chlorotic type with relatively high Hb-index. In his case of grave anemia Limbeck described extreme poikilocytosis, but nucleated red cells were apparently absent. In two very anemic acute cases lasting 4 and 5 weeks, the writer found under two million red cells, but there were no distinct megalocytes and no nucleated red cells. Cabot finds that nucleated red cells are usually absent in tuberculosis, even after hemorrhage, and contrasts this fact with their abundance in carcinoma. Degenerative changes in the red cells in tuberculosis are not marked. In 13 cases of uncomplicated phthisis Grawitz^ found no signs of granular degeneration. The Leucocytes in Tuberculosis. — In the majority of cases of uncomplicated tuberculosis the leucocytes remain within normal limits or are distinctly deficient in number. This rule has been es- tablished principally by later observers, since the earlier studies cited by Rieder, Reinert, and others, indicated that tuberculous inflammation usually excites leucocytosis. Yet Halla noted that leucocytosis was usually found only in ad- vanced cases with fever, and Rieder distinguished the fresh cases from chronic febrile tuberculosis by the absence of any increase of white cells in the former. It is perhaps a still more uniform rule that leuco- cytosis is found almost exclusively in the febrile and anemic cases, but it is absent also in many febrile periods. In pulmonary tuberculosis, an increase of leucocytes is usually re- ferable to suppurating cavities, advancing pneumonia, severe anemia, or hemoptysis. Suppurating cavities, when of recent formation, usually raise the leucocytes distinctly, 15,000 cells being frequently seen, and when the lung is softening the excess may be much greater. The leucocytosis usually persists while the fever continues. With old cavities the expectoration may be abundant, but the leucocytes usually fall to normal. In the pneumonia of various types which complicates phthisis, the leucocytes are usually much increased, but considerable areas of diifuse tuberculous pneumonia may be found with acute miliary tuberculosis, which have failed to increase the leucocytes during life. A lobar pneu- monia causes the usual leucocytosis, but the writer has seen both lungs consolidated and riddled with small cavities, in a case lasting five weeks, yet the leucocytes were never found above 12,000. Tixe absence of leuco- cytosis in these cases of acute phthisis which resemble pneumonia may often be of value in diagnosis. Similarly in a case of subacute empyema in which the tubercle bacillus was largely concerned, the leucocytes were found not to exceed 14,000 during an acute febrile period. The leucocytosis of tuberculous cachexia is moderate, usually between 10,000 and 15,000. After hemoptysis the leucocytes are increased in proportion to the hemorrhage, but the increase is transient. Acute general miliary tuberculosis offers the best illustration of the 302 SYPHILIS, TUBERCULOSIS, LEPROSY. failure of pure tuberculous inflammation to induce leucocytosis. Rieder, Limbeck, Warthin, and Cabot find rather a diminution of leucocytes, often to an extreme degree and, as Warthin has shown, during a pro- lono-ed period. Minor suppurative complications seem to have no capacity to raise the white cells, but in some reported cases a high per- centage of polynuclear cells was noted among the scanty leucocytes. (Warthin.) Tuberculous inflammations of serous membranes, when uncompli- cated by pyogenic infection, follow the type of pure tuberculous in- flammation elsewhere and fail to increase the leucocytes. In many cases however there appears to be mixed infection, and the eifusion, in- stead of remaining serous or sero-sanguinolent, becomes sero-purulent and purulent. In such cases of pleurisy with sero-purulent or puru- lent exudate the writer has found a moderate increase of leucocytes, usually in proportion to the height of the temperature. While it is undoubtedly true that the tubercle bacillus may of itself produce pus the fact remains that it is seldom or never seen alone in purulent exu- dates. The slight leucocytosis that accompanies cases of purulent tu- berculous inflammations of serous membranes is perhaps a further in- dication that these cases represent mixed infections. Tuberculous Meningitis. — In the majority of reported cases of this important localization of the disease leucocytosis has been absent, as seen by Limbeck,^ Pick, Rieder, Sorensen. Yet in five of 7 cases reported by Cabot and in one by Ziemke, there was distinct leucocy- tosis, from 14,700-34,300, while in one of Rieder's cases there were 14,400 white cells. It is difficult to reconcile these conflicting results. Some of the leucocytoses may have been ante-mortem phenomena, but not all. Ziemke's case was one of advanced general tuberculosis. The writer has seen cases with leucocytosis, but complicating terminal pneu- monia was in each instance found at autopsy. Morphology of Leucocytes in Tuberculosis. — In most cases of chronic tuberculosis, when the leucocytes are normal or diminished, the mono- nuclear cells are relatively in excess, but in some instances the usual rule is reversed, as in Warthin's case, and the polynuclear cells are in excess. Distinct lymphocytosis is seen in many chronic cases, espe- cially those with large lymph nodes (lymphatic anemia). Lymphatic leukemia sometimes develops in tuberculous subjects, but the relation of the two diseases is uncertain. When distinct leucocytosis occurs, excess of polynuclear cells is usual, and pyogenic infections, etc., are indicated. Neusser believes that eosinophilia indicates relative immunity to tuber- culosis. General experience seems to show tbat eosins are frequently pres- ent, usually in normal numbers, in the blood in chronic tuberculosis, and sometimes absent in the acute cases, or when there is fever and leucocytosis. Zappert found none in 2 of 5 cases of febrile tuberculosis. They are com- monly increased, sometimes to an extreme degree, in the afebrile period fol- lowing injections of tuberculin. (Zappert.) Tuberculosis of Bones and Joints. — Valuable studies in this field have been contributed by Brown in 72, and Dane in 41 cases, of tuber- culosis of hip and spine. BACTERIOLOGICAL EXAMINATION OF THE BLOOD. 303 From these studies it appears that the above conditions are very frequently accompanied by moderate or marked leucocytosis (maximum 41,000), which is usually referable to secondary infection and the for- mation of pus. Yet Brown observed several cases with leucocytosis but without formation of abscess, and believes that leucocytosis may be referable alone to increased activity of the tuberculous process. A frequent cause of leucocytosis, often of marked degree, was secondary infection of the sinus after operation. The leucocytosis was usually but not always associated with fever. Both authors agreed that when leucocytosis was absent in cases with abscess, secondary infection had not occurred, since the pus was sterile. They did not apparentl}^ con- sider the possibility that a micro-organism once present may disappear, or that leucocytosis may disappear when pus ceases to form. Myelocytes have been reported in scanty numbers in a few cases. Chemistry. — Specal chemical alterations of the blood in tubercu- losis have not been demonstrated, but the usual changes of secondary anemia have been reported in a few cases. Hammarschlag found a low gravity only in cachectic cases. Moderate reductions only were ob- served by Devoto and Scholkoff, but Schmaltz found a gravity of 1.036 in an advanced case of phthisis, and Grawitz's case of grave anemia gave 1.032. The gravity is much lower, in proportion to the loss of cells, than in pernicious anemia. Dieballa found a density of 1.039 in a case of phthisis with 2.4 million cells, and in one of pernicious anemia with only 840,000 cells. Freund found in the blood of tuberculous subjects a body which he regarded as cellulose, which Nischimura has likewise identified, while others have found the same substance in tuberculous tissues. Bacteriological Examination of the Blood in Tuberculosis. — Even before the discovery of the tubercle bacillus Weigert had pointed out the considerable involvement which the vessels may show in acute miliary tuberculosis. Weichselbaum, in 1884, was able with prolonged searching to find numbers of tubercle bacilli in stained specimens of the clotted blood of the great vessels in three cases of acute general tuberculosis. In the same year Meisels and Lustig both succeeded in finding tubercle bacilli in the blood intra vitam, the speci- mens having been squeezed from the finger tip during a period of ris- ing pyrexia. In the blood of the spleen the bacilli were much more numerous and Meisels recommended aspiration of the spleen as an aid in the diagywsis of obscure cases. Rutiraeyer also succeeded in finding many bacilli in the aspirated splenic blood of one of two cases examined shortly before death. In a case which eventually recovered. Sticker, after considerable search, found a few bacilli in the finger blood taken on the tenth day of the illness. With the introduction of tuherculin, Liebmann claimed to have found tubercle bacilli in 56 of 141 blood specimens, most abundantly about 24 hours after the injection of tuberculin. Grave doubt was, however, thrown upon all previous work in this field by the negative results obtained by Gutt- 304 SYPHILIS, TUBERCULOSIS, LEPROSY. mann and by Kossel, who called attention to the danger of contamination in such specimens. This method of diagnosis has not been followed up in later years, although Kronig, in 1894, recommended the staining of centri- fugedlaked blood in doubtful cases of miliary tuberculosis. In the blood of tuberculous cadavers various micro-organisms have been isolated by Pasquale, Petruschky, Canon, Welch and Nuttall, and others. Cultures of the blood during life have been made, by reliable methods, by Sittmann, Kraus, Kuhnau, Hewelke, White, Mich- aelis and Meyer, and Hirschlaif. Of their 79 cases of advanced and usually febrile phthisis, there were 23 positive and 56 negative results. Staphylococcus pyogenes aur- eus or albus was isolated in 20 cases. Streptococcus pyogenes and a diplo- coccus in 1 case each, and the tubercle bacillus was once obtained from inoculation, by Kuhnau. The pyogenic germs were sometimes found together. Jakowsky and Petruschky were rather more successful, drawing the blood through the skin, but this method is unreliable. There can be little doubt therefore that the blood in the late stages of phthisis suffers bacterial invasion as in other forms of septicemia. Kuhnau's one positive result from 12 inoculations of the blood of se- vere febrile cases does not encourage further search for the tubercle bacillus in the blood of tuberculous subjects. LEPROSY. ^^Winiarski, in 1892, studied the blood in 17 cases illustrating various phases of this disease. ,In some cases he found polycythemia, 6.38- 6.09 million red cells, which may perhaps be referred to local stasis, as the hands appeared cyanosed. Few of the patients showed any marked loss of cells, the average being 5.05 millions for 9 men, and 4.3 millions for 8 women. Winiarski therefore concluded that the milder cases of one or two years' standing do not suffer from any dis- tinct anemia. Yet in 2 advanced cases with extensive lesions, the red cells numbered 2.3 and 1.9 millions, and the latter patient presented many signs of pernicious anemia. In these cases the anemia was re- ferred to extensive ulceration. The Hb suffered even less than the cells, usually registering 80-118 percent in all but the two very anemic cases. The Hb-index was uniformly high, and in one case with 1.9 million cells and 64 percent Hb, the index reached 1.7. An increased diameter of the red cells was noted in this and other cases. The leucocytes were normal or subnormal in number, and the lympho- cytes were in relative excess (maximum 47 percent), except when sup- puration occurred. Brown, more recently, reported his observations in 16 cases, 8 of which appeared to be quiescent and presented normal blood, while only one showed severe anemia. He claimed to have found the bacillus in the leucocytes of the circulating blood in 9 cases, 8 of which showed the tubercular type of lesion. BIBLIOGRAPHY. 305 Streker, also, examined the blood of 5 " very anemic " cases of leprosy and found bacilli both free in the plasma and inclosed in leuco- cytes. The blood was drawn with antiseptic precautions from a deep incision through normal skin. These observations are quite in accord with the reports of Joseph regarding the large deposits of Bacillus leprce in the spleen. Spronck claims that the serum of leprous subjects, in dilutions be- tween 1-60 and 1-1,000, agglutinates fresh living cultures of the ba- cillus of Hansen, and he recommends the use of this test in diagnosis. Bibliography. Tuberculosis, Syphilis, Leprosy. Anc. Monatshefte f. Prak. Dermat., Bd. 12, p. 266. Virchow-Hirsch. Jahresber., 1892, II., p. 537. Andral, Oavarret. Annal. de Chem. et de Physiol., 1840. Baginsky. Lehrb. d. Kinderkrank., Berlin, 1889, p. 287. Barbacci. Cent. f. med. Wissen., 1887, No. 35. Becquerel, Bodier. Cited by Wilbouchewitsch, Arcliiv de Physiol, norm, et path. 1874, p. 509. Also, Memoir Acad, des Sci., 1844. Bieganski. Archiv f. Dermat. u. Syph., 1892, p. 43. Bierfreund. Langenbeck's Archiv, Bd. 41, p. 1. Brown. Occidental Med. Times, 1897, p. 537. Also, Trans. California Med. Soc, 1897, p. 168. Cabot. Boston Med. Surg. Jour., 1899, Vol. 140, p. 323. Canon. Deut. Zeit. f. Chirurg., Bd. 37, p. 571. Bane. Boston Med. Surg. Jour., 1896, Vol. 134, p. 559. Dehio. St. Petersburg med. Woch., 1891, p. 1, cited by Limbeck. Demelin. Mai. de I'Enfance (Grancher), II., p. 135. Bevoto. Zeit. f. Heilk., Bd. XL, p. 176. Dieballa. Deut. Archiv klin. Med., Bd. 59, p. 308. Bmiinici. Presse Med., 1898, I., p. 468. Evdng. N. Y. Med. Jour., 1893, Vol. 58, p. 713. Fenoglio. Oesterreich med. Jahrb., 1882, p. 635. Fischl. Zeit. f. Heilk., Bd. 13, p. 291. Freund. Wien. med. Jahrb., Bd. 1, p. 335. Oaillard. Gaz. des Hop., 1885, No. 74. Gartner, Homer. Wien. klin. Woch., 1892, No. 2. Gnezda. Inaug. Diss. Berlin, 1886. Gross;'. L'Union Med., 1859. Gratvitz. ^ Deut. med. Woch., 1893, No. 51. ^Zeit. f. klin. Med., Bd. 21, p. 459. 'Berl. klin. Woch., 1900.^ Guttmann. Cited by Kossel. Hammarschlag. Zeit. f. klin. Med., Bd. 21, p. 475. Hayem. Du Sang., p. 920. Heidenhain. Pfliiger's Archiv, Bd. 49, p. 209. Hewelke. (Abstr.) Cent. f. Bact., Bd. 19, p. 563. Hirschlaff. Deut. med. Woch., 1897, p. 766. Jakowsky. Cited by Sittmann. Jawein. Cited by Lindstrom. Jellenef. Annal. de Dermat., 1892, p. 924. Jones. N. Y. Med. Jour., 1900, Vol. 71, p. 513. Joseph. Archiv f. Dermat. u. Syph., Bd. 43, p. 359. Justus. Virchow's Archiv, Bd. 140, p. 91. Ibid., Bd. 140, p. 533. Keyes. Amer. Jour. Med. Sci., 1876, Vol. 71, p. 17. Koberi. St. Petersburg med. Woch., 1889, No. 32. Konried. Wien. klin. Woch., 1893, p. 341. Kossel. Berl. klin. Woch., 1891, p. 302. Kraus. Zeit. f. Heilk., Bd. 17, p. 117. Kronig. Deut. med. Woch., 1894, V. B, p. 42. Kuhnau. Zeit. f. Hygiene, Bd. 25, p. 492. 20 306 SYPHILIS, TUBERCULOSIS, LEPROSY. Laache. Die Anemie, 1883, p. 63. Laker. Wien. med. Woch., 1886, No. 18. Lezius. Blutver. bei Syphilitischen, Diss. Dorpat, 1889. Liebvmn. Berl. klin. Woch., 1891, p. 393. Limbeck, i Grundriss, etc., p. 336. « Cited by Kieder, Leucocytose. Lindstrom. Presse Med., 1898, T. 13, p. 267. Loos. Wien. klin. Woch., 1892, p. 291. Lustig. Wien. med. Woch., 1884, p. 1187. Malassez. lArchiv de physiol. norm, et path., 1886. "Prog. Med., 1874, p. 562. Meisels. Wien. med. Woch., 1884, p. 1187. Michaelis, Meyer. Charite-Annalen, Bd. 22, p. 150. Muller. Charite-Annalen, Bd. 14, p. 253. Nischimura. Archiv f. Hygiene, Bd. 21, p. 52. Oppenheimer. Deut. med. Woch., 1889, No. 42. Pasquale. Ziegler's Beitrage, Bd. 12. Peiper. Cent. f. klin. Med., 1891, p. 217. Petruschky. Zeit. f. Hygiene, Bd. 17, p. 59. Pick. Cited by Eieder. Reinert. Zahlung d. Blutkorp, p. 189. Ricord. Schmidt's Jahrb., Bd. 45, p. 45. Rieder. Leucocytose, p. 126. Rille. Wien. klin. Woch., 1893, p. 155. Rutinieyer. Cent. f. klin. Med., 1885, p. 353. Schiff. Pest. med. chir. Presse, 1892, p. 48. Schlesinger. Archiv f. exper. Path., Bd. 13. Schmaltz. Deut. Archiv klin. Med., Bd. 47, p. 145. Scholkoff. Diss. Bern, 1892. Sittmann. Deut. Archiv klin. Med., Bd. 53, p. 323. Sorensen. Cited by Reinert. Sticker. Cent. f. klin. Med., 1885, p. 441. Stoukovenkoff. Annal. de Dermat., 1892, p. 924. Strauer. Zeit. f. klin. Med., Bd. 24, p. 295. Slreker. Munch, med. Woch., 1897, p. 1103. Vierordt. Cited by Reinert. Warthin. Medical News, 1896, Vol. 68, p. 89. Weichselbaum. Wien. med. Woch., 1884, p. 364. Welch, Nuttall. Hyg. Rundschau, Bd. II., p. 927. White. Jour, of Exper. Med., Vol. IV., p. 425. Winiarski. St. Petersburg med. Woch., 1892, p. 365. Wiskeman. Zeit. f. Biol., 1873, p. 434. Zappert. Zeit. f. klin. Med., Bd. 23, p. 227. Ziemke. Deut. med. Woch., 1897, Apr. 8, cited by Cabot. PAET IV. CONSTITUTIONAL DISEASES. CHAPTER XVII. HEMOERHAGIC DISEASES AND THE HEMOREHAGIC DIATHESIS. Hemorrhages of greater or less extent occur iu many general dis- eases, in which the loss of blood is not referable to any specific predispo- sition on the part of the individual but to peculiar conditions arising in the course of the disease. Such hemorrhages are regarded as symptomatic. The chief causes of symptomatic hemorrhages are : 1. Infections, as in many severe infectious diseases such as septi- cemia, scarlatina, smallpox, measles, etc. 2. Mechanical, as in partly asphyxiated infants, or other conditions of marked venous stasis. 3. Toxic and autotoxic, as in poisoning by the so-called blood- poisons, or in jaundice, anemia, etc. 4. Nervous, as in rare forms of hysteria. In the hemorrhagic diathesis probably all the above factors are at times unusually active but there is believed to be in addition a specific, preexisting, often hereditary change iu the blood which largely deter- mines the various manifestations of this diathesis. Modern bacterio- logical and chemical research has now considerably narrowed the scope of the hemorrhagic diathesis, and at present it is impossible to claim that any peculiar predisposition to hemorrhage exists except in cases of hemophilia. Yet it remains convenient to describe the various idio- pathic purpuras, scurvy, and hemophilia, as closely related conditions. PURPURA HEMORRHAGICA. WERLHOFF'S DISEASE. The etiology of this disease still remains obscure, although the evi- dence points more and more toward some form of infection as the es- sential element in the majority of cases. The prominence of hemor- rhage in the symptoms of many infectious diseases has prepared the way for an acceptance of this view. In adults, the extent of the hemorrhage attending well identified forms of infection is rarely so great as in pronounced cases of purpura hemorrhagica, in which often no bacterial origin has been demonstrable. In infants, however, in- 308 HEMORRHAGIC DISEASES AND HEMORRHAGIC DIATHESIS fections by the ordinary pyogenic bacteria have become recognized as a frequent source of fatal hemorrhages constituting for that age the typical picture of WerlhoiF's disease. There are now mauj'^ cases of purpura hemorrhagica on record in which bacteriological examination of the blood or viscera showed the presence of bacteria which were regarded as the cause of the disease. In some cases bacilli were isolated, which were not fully identified. Letzerich's Bacillus purpurse in many respects resembled Bacillus anthracis, and produced in ani- mals, and possibly in the investigator's own person, a hemorrhagic disease resembling Werlhoff 's. Streptococcus pyogenes has been isolated in typical cases by Hanot and Luzet, Widal and Therese, Guarnieri, and others. In the first of these the disease was transmitted from mother to fetus, the latter dying and yielding a pure culture from the blood. In infants, rapidly fatal hemorrhagic infections of this nature are not rare. Staphylococcus pyogenes aureus has been isolated by Lebreton, Litten, Fischl and Adler, Lewis, Silvestrini, and others. Fischl and Adler claim to have produced a fatal anemia in animals by inoculation with their coccus. Kolb obtained sterile cultures of the blood in 5 cases, but from the viscera of three of these, 3-4 hours after death, he isolated a diplobacillus which caused hemorrhagic septicemia in various animals. Auche obtained both the staph- ylococcus and streptococcus in a fatal case, and Levi obtained in another both Streptococcus pyogenes and the pneumococcus of Frankel. The Pneumococcus lanceolatus was isolated post-mortem by Claisse and by Claude. Bacillus pyocyaneus was obtained post-mortem from a case of melena neonatorum by Neumann, and Bacillus coli communis by Dansac, Legendre, and others. Hamilton and Yates report a rapidly fatal case showing, ten hours post-mortem. Bacillus aerogenes capsulatus. McLeod ob- served a case closely following Malta fever, and conditions described as pur- pura hemorrhagica have been reported as following pertussis, tuberculosis, congenital syphilis, etc. Negative results are reported in a fatal case by Denys ; in a purpuric complication of angina, by Legendre ; and in chronic cases by Marfan, and by Millard. In the majority of the above cases the bacterium was obtained post- mortem and from the viscera, or during life, from the lesions in the skin, and was rarely demonstrated in the circulating blood. Never- theless there can be but little doubt that a numerous group of cases of purpura hemorrhagica exists Avhich is referable to a variety of bacteria, as represented in the above reports. It seems necessary therefore to provisionally separate the cases of pur- pura hemorrhagica of probable infectious origin from those which in no respect resemble an infectious disease, and are therefore probably acute manifestations of hemophilia. Changes in the Blood. — In most of the febrile infectious cases oi pur- pura hemorrhagica the anemia is not excessive and appears to be sec- ondary to the infection and the loss of blood. In mild cases the red cells are slightly diminished in number but exhibit no other alterations. Carriere and Gilbert report cases with 3.35 million and 3.9 million red cells, many of which were microcytes. In more severe cases the red cells may be greatly diminished, in proportion to the extent of hemorrhage, but a microcytic type of red cell is usually prominent. The Hb-index is subnormal. Nucleated red cells occur when hemor- rhages are large or frequent. NON-INFECTIOUS IDIOPATHIC PUEPURA HEMORRHAGICA. 309 In a fatal case in an infant, yielding a pure culture of Streptococcus pyogenes at post-mortem, the writer found the blood to show all the characteristics of primary pernicious anemia. In 4 other cases of fatal hemorrhage in newborn infants at Sloane Maternity Hospital, the subjects had bled to death in the course of 3-5 days, but the blood showed only the characteristics of secondary anemia with leucocyto^is. Leucocytosis has been noted by the majority of observers. In one case Carriere and Gilbert found 126,000 leucocytes, 90 percent poly- nuclear, and 5.8 percent eosinophile. In the writer's cases in infants the leucocytosis reached as high as 56,000, but the eosins were scanty. A very marked degree of polychromatophilia was observed in two cases by Spietschka. Hayem' and Bensaude have described a peculiarity in the blood of purpura hemorrhagica which they claim to be pathognomonic. On allowing the blood to clot in a vessel it was found that after 24 hours the retraction of the coagulum is very feeble and fails to express the serum, as happens in normal blood. Associated with this feeble coagu- lation there is a marked reduction of blood plates (200,000 to 50,000). These two features of the blood were noted in 16 cases of purpura hemorrhagica, but in 152 examples of other diseases, some with pur- pura, the feeble clotting and loss of blood plates were never found to be combined. Non-infectious Idiopathic Purpura Hemorrhagica. — In addition to the cases of purpura which arise in the course of infectious diseases or after various infectious processes, in which hemorrhage constitutes the chief symptom of a cryptogenic infection, there are other cases of rapidly fatal anemia attended with severe and repeated losses of blood for which a diiferent etiology is indicated. Some of these cases are difficult to distinguish from pernicious anemia, but differ from that disease in the absence of megaloblastic changes in the blood or marrow, the great predominance of microcytes in the blood, and the prominence of hemorrhage among the symptoms. They differ from the infectious purpura in the more persistent hemor- rhage, the absence of leucocytosis, and the absence of general symp- toms of an infectious disease. Such cases have been reported by Ehrlich,^ Engel, and others, as pernicious anemia without characteristic changes in the blood or mar- row. The writer can find no evidence on which to claim that per- nicious anemia can exist without such changes in the marrow, and believes that these cases belong to the diseases referable to the hemor- rhagic diathesis. This view is supported not only by the great dis- similarity in the blood changes of the two conditions but by the ab- sence of any marked deposits of iron in the viscera. (Zalesky.) Of the various etiological factors to which fatal purpura has been attributed the non-infectious variety may be referred to any or all ex- cept the presence of bacteria. The underlying condition is probably identical with that of hemophilia of more typical course. Changes in tlie Blood. — The red cells are rapidly diminished in num- 310 HEMORRHAGIC DISEASES AND HEMORRHAGIC DIATHESIS. ber, reaching before death an extremely low figure. In a case of the writer's persistent epistaxis reduced them in three weeks to 456,000. In a case reported by Billings they numbered 483,000. In Engel's case there were over two millions. Contrary to the rule in secondary anemia following hemorrhage, nucleated red cells are extremely scarce or absent. The majority of the red cells are undersized and many are of oval shape. The leucocytes are normal or reduced in number and of those remaining a large proportion (90 percent, Engel ; 80 percent, Ehrlich ; 75 percent, Billings) are lymphocytes. Eosinophile cells are scarce or absent. HEMOPHILIA. Hemophilia is an extremely hereditary constitutional anomaly which predisposes the subject to persistent and fatal hemorrhages of traumatic or spontaneous origin. The peculiar law of heredity which transmits the condition to males, only through the female parents who are them- selves usually exempt, as fully set forth by Grandidier and Stahel, not only marks this malady as the clearest possible illustration of the hereditary transmission of disease but completely baffles the attempts to prove its dependence on any infectious agent. The essential lesion has been held by many to lie in an unnatural thinness and narrowness of the arteries (Virchow). In somewhat altered form this theory is maintained by Immerman and Oertel, who hold that there is a disproportion betioeen the bulk of blood and the capacity of the blood systeni, referable in part to hypoplasia of vessels (Immerman) or to hydremic plethora. Proceeding on this theory Cohen claims to have considerably lessened the tendency to hemorrhage in a pro- nounced bleeder by a prolonged course of diuresis and diaphoresis. In partial support of this mechanical theory are cited a few cases in which the heart was hypertrophied, and the vessels of small size, and showing fatty, hyaline, or granular degeneration. A locfd origin is perhaps indicated by the case cited by Stengel, in which the hemorrhages were observed to occur from cuts above the neck but never from those below. Hecklinghausen holds that a disorder of the nervous system is the essential cause of the hemorrhages, and Henoch believes that the local condition is one of paralytic dilatation of vessels, followed by diapede- sis. The condition has been seen in some very remarkable forms in neurotic and hysterical subjects. (Stengel.) Little success has followed numerous attempts to demonstrate essen- tial changes in the blood. Albertoni claims to have found dimin- ished resistance in the red cells. In morphology there is uniform agreement that the blood presents nothing peculiar. Decreased coagu- lability was held to be constant, by older writers (Grandidier, Lossen), and has later been observed by Schmidt and ManteufiFel, who found that the blood in hemophilia requires longer than usual to clot, Avhile its coagulation may be greatly hastened by adding a zymoplastic sub- stance (fibrin ferment). The fibrin itself is apparently not greatly SCURVY. 311 diminished, 5 percent having been found by Hey land, 2.6 percent by Gavoy and Ritter, and 4.3 percent by Otto (cited by Litten). '' Koch's theory of the infectious origin of hemophilia seems hardly worth discussing." (Litten.) Studies in pathological anatomy have also failed to show any con- stant lesion of essential importance. In most of the viscera there are the lesions which follow acute or chronic anemia. Buhl described an unusually rich network of capillaries in the skin of one case, but the tissue was the seat of a chronic inflammatory process. Birch-Hirsch- feld found normal vessels, slightly fatty heart muscle, and slight hyperplasia of the splenic stroma. Litten noted areas of granular de- generation in capillary endothelium, and irregularity in outline or widening of intercellular spaces between these cells, but suspected these changes to be artificial. In a series of cases hypoplasia of the heart furnished the basis of Virchow's theory. The blood changes are determined by the extent of the hemorrhages, but leucocytosis has been absent in the few cases reported. In general the blood resembles that of non-infectious purpura hemorrhagica. SCURVY. Etiology. — The most important fact in the etiology of scurvy is its intimate relation to improper diet. Aside from the almost exclusive occurrence of the disease in subjects whose food has been restricted in variety, it has been repeatedly demonstrated that the malady may be promptly cured by supplying some element previously deficient in the diet. Even a change in water supply has eradicated epidemics in garrisons. Scorbutic infants rapidly recover from grave stages of the disease after such apparently trivial assistance as a change in the milk. Yet it has never been fully determined just what chemical substance is concerned in the causation and cure of the disease. It has even been claimed that diet has no relation to the malady, but all such reports have thus far, on investigation, failed to be supported, although they have demonstrated that the dietetic error may be very trivial. Thus Seelaud, regarding a somewhat limited army ration as full mixed diet, came to regard the cases of scurvy under his care as of miasmatic origin. Accumulating evidence has shown that the most frequent source of the disease is a too exclusive diet of meat, especially if salted, and the absence of fresh vegetables, especially of potatoes, which are very rich in potassium. Guided by these observations various theories of origin of scurvy have been elaborated, only to be disproved by later study. Among these is the belief that the symptoms were caused by excess of NaCl in the blood, a condition long since shown to characterize most forms of anemia. That the essential element in the disease consists in a deficiency of potassium in the blood was held by Garrod, Liebig, and Hirscli, but a loss of potassium from the blood, with excess in the urine, while not constant in scurvy, is observed in other conditions. Similarly the opinion of Ralfe and Cantani, that a deficiency of 312 HEMORRHAGIC DISEASES AND HEMORRHAGIC DIATHESIS salts of vegetable organic acids and corresponding diminution in the alka- linity of the blood is the important factor, has failed to receive support. Nevertheless to those who have observed the immediate effects of correction of diet in severe cases, especially in children, this field of investigation appears most likely to offer the true explanation of the disease, for which the sole hope is an advance in our knowledge of the chemistry of digestion and absorption, and of the blood. On the other hand there are on record some very suggestive bacteri- ological studies of scurvy, pointing toward an infectious nature of the disease. The injection into animals of blood from scorbutic patients caused purpuric lesions and internal hemorrhages in the experiments of Murri,^ Petrone, and others. Babes failed to find bacteria in the blood of a series of cases and failed to produce significant lesions by injecting blood into animals, but from excised portions of the spongy gums he isolated a bacillus, regarded as identical with one previously described by Muller as common in the buccal cavity, and by injections of the comminuted tissue he produced, in two out of a series of rab- bits, a form of hemorrhagic septicemia. From the visceral lesions he recovered this same bacillus along with the ordinary bacillus of hemor- rhagic septicemia in rabbits. It does not appear that much significance can be attached to these results, although they are commonly quoted as strongly supporting the infectious theory of the origin of scurvy. Rosenell also recovered a somewhat similar bacillus from the viscera of a fatal case of scurvy, but inoculations into rabbits were negative. Entirely negative results were obtained by AVieruszky in a large num- ber of cases. Most of his cultures of the blood were sterile, but in 14 cases he found various germs to which he could not assign any relation to the disease. In recent years the bacteriological study of scurvy has not proved an inviting or accessible field to competent bacteriologists. A recent impox'tant study directed toward the chemistry of the blood and of digestion is that of Albertoni, who has shown that there is a marked loss or complete absence of free HCl in the gastric juice in scorbutic patients, that gastric digestion is deficient, that intestinal putrefaction is excessive, and that there are abundant evidences in the urine of absorption of toxic products, while the absorption of fats and carbohydrates is deficient. In the blood, Albertoni found a much greater reduction in iron (.3-.39 percent) than is seen in other anemias with equal numbers of red cells. The sodium and potassium varied between normal limits. The greenish yellow color of the serum and the excess of pigments in the urine were evidences of active destruction of blood cells. Changes in the Blood. — The morphological changes in the blood in scurvy resemble those of secondary anemia from hemorrhage. The red cells in cases of average severity number between 3 and 4 millions, but with severe hemorrhages the anemia may become very grave, as in Bouchut's case in which, after three weeks of persistent epistaxis, the red cells were reduced to 557,000. The reports of Us- kow, Hayem,^ Wieruschky, and Albertoni, show that the red cells vary HEMOCYTOL YSIS. 313 in number and size according to the length and severity of the disease. Litten saw many megalocytes and shrunken microcytes, evidently in severe cases. Albertoni found evidences of solution of red cells in the plasma, and an excessive number of pale cells or fragments of cells have been described by several observers. The Hb-index is low, White finding 20 percent Hb with 40 percent of red cells, while Bec- querel and Albertoni found the iron content much lower than in other anemias. Opitz and Duchek, however, reported a normal or increased proportion of iron. The proportions of sodium and potassium salts vary as in other anemias, and the claims of Ralfe and Cantani that the alkalinity of the blood is deficient have not been supported. On account of the frequency of inflammatory complications and hemorrhages the leucocytes are usually increased, Uskow finding as high as 47,000. Litten, however, observed no leucocytosis, and the writer, in two well-marked but uncomplicated cases, found no increase. In Barlow's disease, which is a form of scurvy with extensive sub- periosteal hemorrhages, the condition of the blood varies greatly, ac- cording to the general condition of the patient and the number and ex- tent of the hemorrhages. All grades of post-hemorrhagic anemia have been observed, up to Reinert's fatal case in which the Hb fell to 17 percent and the red cells to 976,000. The disease is often associated with rachitis. HEMOOYTOLYSIS. When the isotonic tension of the plasma falls below that of the red cells, the Hb may become dissolved in the circulating blood, and the shadow-like remnants of the stroma of the cells are distributed in the plasma. In other conditions, even without lowered isotonic tension of the plasma, the red cells may be split up into fragments. The purest examples of the former variety of destruction of red cells are perhaps seen after extensive infusions of salt solution, while the latter variety is observed especially from the effects of certain blood poisons, e. g., potassium chlorate. Usually the two processes are combined, as in the anemias and infectious diseases, while in some conditions the exact manner of the solution of Hb remains unexplained. The chemical processes in the cell leading to the solution of Hb are somewhat obscure, since this constituent is probably held in chemical combination with the stroma and not merely held in situ by a hypo- thetical cell membrane. Thus the most extreme form of hemoglobi- nemia (paroxysmal) may occur in blood of comparatively high isotonic tension, while the high grades of globulicidal activity of the plasma in infectious diseases are not always associated with hydremia. The morphological changes occurring in dissolving red cells are also not fully identified, although it appears more and more probable that many forms of degeneration previously described in red cells are the precursors of the complete solution of Hb. The polychromatic de- generation of Maragliano, that of Gabritschewsky, and the granular 314 HE3I0BBHAGIC DISEASES AND HEMORRHAGIC DIATHESIS degeneration of Grawitz, are probably morphological signs indicating approaching or progressive solution of the cell. From extensive study of the anemia following pyrodin poisoning Tallquist concludes that the presence of microeytes is a sign of hematolysis, while megalocytes indieate regeneration of the blood. The dissolved Hb imparts to the plasma, in the eosin-stained specimen, an abnormal reddish tinge, and discolors the serum. When present in moderate amount it may be gradually transformed by the liver into bile-pigment, which is occasionally reab- sorbed from the biliary passages, causing hematogenous jaundice. The presence of an excessive amount of Hb in the plasma is apparently the chief reason why it is sometimes excreted by the kidneys and appears in the urine as hemoglobin or methemoglobin. During its passage through the liver, kidney. Fig. 24. and other tissues, a con- siderable portion of the iron is retained in the demonstra- ble granular or diffuse form of hemosiderin. Under some conditions not fully understood, but believed by Thoma to consist essentially in the relative absence of oxygen, disinte- grating red cells are trans- formed largely into brown- ish granules or crystals of hematoidin. This change usually takes place without the intervention of cellular activity. (Thoma.) It fre- quently occurs post-mortem. Broken fragments of red cells and the stroma of dis- integrated corpuscles float in the plasma for a time and may be observed in the stained specimen, but are gradually englobed by phagocytic cells and lodged in various organs, principally in the spleen. The anemia of the infectious diseases, especially in grave septic conditions, is established rapidly and to a large extent through the di- rect solution of red cells by the globulicidal plasma. The condition is often attended during life by hemoglobinuria, and after death ex- tensive deposits of blood pigment are found in all viscera, especially in the spleen and liver. In malaria various forms of destruction of red cells occur, which will be considered later. Acute degeneration of red cells. Nitro-benzol poison- ing. (K. Ehrlich, LiNDENTH.\L. ) Subdivision of ccUs and appearance of basophilic granules in them. Special Conditions in Which Blood Cells Are Rapidly Destroyed. General Burns. — The remarkable effect of high temperatures in splitting up red cells may be readily observed by heating a fresh spec- SALT SOLUTION, SERUM, AND DEFIBRINATED BLOOD. 315 imen over a flame. Schmidt found that this effect occurs with tem- peratures above 50° C. The writer failed to find any traces of disin- tegrated red cells in a series of rabbits exposed for other purposes to a temperature of 45-46° C. In animals that had been scalded Wertheim found many subdivided red cells. In fatal cases of burns in the human subject similar eifects may be observed, but not in a degree sufficient to cause death, which must result from other factors. That hemoglobin may be dissolved in the plasma and appear in the urine after burns is stated by Hoppe-Seyler. Snake poison contains a toxic globulin or albumen which dissolves red cells and produces hemoglobinemia and hemoglobinuria. Fumes of arseniuretted hydrogen are very active, when inhaled, in dissolving red cells, producing hemoglobinemia, hemoglobinuria, with Hb-infarcts in the kidneys, and acute degeneration of the viscera. Poisoning" by toadstools, by the alkaloids of Quillaia saponaria (saponin, sapotoxin), of Solanum, and other motor-depressants, dis- solves red cells in the blood, and leads to hemoglobinuria. (Robert.) In a fatal case of poisoning by Guiacol, Wyss observed rapid de- struction of red cells, finding many degenerated, fragmented, shrunken red cells and megalocytes, on the second to third days. The white cells were considerably increased and the majority of these were lymphocytes. Nitrites. — Poisoning by nitroglycerin and by amyl nitrite is said to occur in the manufacture of these chemicals, and to be followed by methemoglobinemia. Nitrobenzol poisoning was studied in a fatal case by K. Ehrlich and Lindenthal. Ten hours after the initial symptoms the blood was chocokite colored, the serum brownish, and spectral analysis sliowed the presence of methemo- globiu, which disappeared by the eighth day. The red cells were rapidly reduced, 2,275,000 on the fifth day, and falling to 900,000 before death on the nineteenth day. Poikilocytosis appeared on the third day and soon reached a remarkable degree. Polychromatic and fragmented cells were abundant. Nucleated red cells were first seen on the third day, and there- after in very large numbers and of all sizes. On the ninth day the leuco- cytes, previously low, rose suddenly to 61,000, and the nucleated red cells were reported at 24,700. .Tudgiug from the authors' plates it was perhaps difficult to distinguish between karyorhexis in some nucleated cells and ex- treme forms of the granular degeneration of Grawitz. The Hb fell steadily to 40 percent, which, with 900,000 cells, was a remarkably high Hb-index. There were many myelocytes among the white cells, so that at one time the blood jiresented the appearance of leukemia. The bone marrow was not examined. These interesting observations recall the efforts of Bignami and Dionisi to produce experimental pernicious anemia bj' means of toluendia- min. The morphological characters of the blood described by Ehrlich and Liudenthal probably represent an extreme degree of the effects upon the blood of the entire group of anilin poisons. Transfusion of Salt Solution, Serum, and Defibrinated Blood. — Siegel and Schram both found no improvement in the regeneration of the blood from the transfusion of salt solution or of serum in animals after bleeding, and while it has since been shown that the regenera- 316 HEMORRHAGIC DISEASES AND HEMORRHAGIC DIATHESIS. tion is somewhat more rapid and complete after salt infusion, yet this procedure must be regarded as of more value as a means of saving life than as a stimulant to blood formation. The direct effects on the blood of infusion of defibrinated blood seem to be much more favorable. On the other hand, after the transfusion of blood, the above observers, and others, have found a rapid increase in the number of cells, and Bizzozero reported the same effect after transfusion of defibrinated blood in animals. Quincke was one of the first to note an increase in red cells in pernicious anemia as a result of transfusion of blood, and similar observations have been made by Ziemann in anemia and scurvy. It would seem, from the observations of Bizzozero and of Bareggi, that the red cells are quite resistant to the process of defibrina- tion and injection. Of the immediate effects of salt infusion upon the blood of the human subject there are a few reports at hand which in- dicate that it has considerable influence in lowering the number of red cells and increasing the leucocytes. Methemoglobinemia. — ^Various poisons not only dissolve red cells, but at once transform the hemoglobin into methemoglobin. The blood in such cases may exhibit a distinct chocolate color. Potassium chlorate is one of the most frequent forms of poison- ing which cause destruction of red cells with methemoglobinemia. According to v. Mering the KCIO3 acts directly upon the cells and is itself thereby reduced, while INIarchand, Falck, and others, hold that the serum is first altered. Brandenburg reports a fatal case in which the red cells fell in 6 days from 4.3 to 1.6 millions. Many of them were deformed. The blood and serum were chocolate colored and gave, during the first 5 days only, the spectrum of methemoglobin. There was marked leucocytosis (20,000). Sticker reports a severe case of poisoning after a vaginal douche by solution of chromic acid, the patient passing brownish urine. He re- fers to other fatal cases with similar signs. Pyrogallol and pyrogallic acid have caused death with symptoms of methemoglobinemia. (Grawitz.) Antifebrine has been found to produce extreme methemoglobinemia in animals (Lepine) and in man (Muller), but neither of these ob- servers could find any morphological changes in the blood of subjects dying from this cause. Antipyrine produces toxic symptoms similar to those of antifebrine, but reports on the examination of the blood in severe cases are want- ing. (See Falk, Muller.) Poisoning by CO. — A bright cherry red color of the blood is a char- acteristic sign in cases of " gas poisoning," and is referable to the presence of CO-Hb. The methods of diagnosis of this condition by spectral analysis have already been considered. (See p. 25.) Poisoning by Hydrocyanic Acid. — The bright red color of the blood in this condition is referable to the presence of cyan-methemoglobin (Robert), which gives a characteristic spectrum. The manner of death is believed by Geppert to be through inhibition of the metabolic POST-MALARIAL HEMOOLOBINEMIA. 317 processes in tissues, so that CO is not discharged and oxygen not ab- sorbed. Morphological changes in the blood have not been reported. Paroxysmal Hemoglobinuria. — This remarkable condition, con- sisting essentially in the rapid solution in the blood stream of enormous numbers of red cells, followed by the appearance of Hb in the urine, was recognized by Stewart more than a century ago (1794), and has been widely studied in its various phases ever since, Chvostek's monograph reviews the literature up to 1894. Etiology. — The most peculiar feature of this malady is the appar- ently trivial nature of the exciting causes of solution of the red cells. Slight exposure to cold, even of a part of the body, has been the most common exciting cause, many confirmed subjects being able to induce an attack at will. Other exciting conditions have consisted in mental excitement or nervous irritation, and muscular or nervous exhaustion. In recently malarious subjects Tomasselli and Koch have placed be- yond doubt that quinine is a frequent exciting agent. An infectious origin has been suggested by Layral and by Cima. The underlying conditions which predispose to the disease are known to include only constitutional syphilis (Boas, Gotze) and malaria (Legg), but in a con- siderable proportion of cases neither of these factors appears to exist. Pathogenesis. — That a vasomotor neurosis is prominently concerned in the solution of blood cells is strongly suggested by the frequency with which nervous influences have been known to precipitate the attack. This theory, first suggested by Dapper in 1868, has been uniformly maintained up to the present time by the majority of ob- servers. The further attempt to explain the destruction of blood by this means has as yet been unsuccessful. Chvostek concluded that as a result of malaria, syphilis, etc., the red cells possess diminished re- sistance, not to cold but to mechanical traumatism, and that when the circulation is disturbed by vasomotor contraction of peripheral vessels the fragile red cells become dissolved. He supported his view by showing, as Dapper had done, that the disturbance of circulation fol- lowing ligation of a finger may cause solution of red cells in the ab- sence of cold. With Pavy, McKenzie, Rosenbach, and others, he attributed to the kidney a special importance in the process. Murri ^ has offered evidence to show that the red cells are abnormally suscep- tible to the effects of cold, and that by paralytic dilatation of peripheral vessels they become exposed to cold and dissolved. But Rodet failed to find, by experiment, any lack of resistance of the red cells against cold, and numerous cases have been observed to follow muscular and mental over-exertion. (Fleischer, Strobing.) Ehrlich supposes that in the predisposed subjects, exposure to cold causes the development of " ferments " which dissolve the red cells. Post-malarial Hemoglobinemia. — The relation of hlackwater fevet^ to malaria has been the subject of much discussion, the opposing views in which are fully presented in the studies of Koch and of Plehn. From the observations, correlated and largely contributed by these ob- servers, it is evident that hemoglobinuria arises : 318 HEMORRHAGIC DISEASES AND HEMORRHAGIC DIATHESIS. 1 . As a direct result of the action of the malarial parasite upon the blood. 2. As an immediate effect of the administration of quinine on pa- tients predisposed to hemoglobinemia from present or preexisting ma- larial infection. 3. It is probable that hemoglobinuria results from simple exposure to cold or to other exciting agents, in subjects who have acquired the predisposition through previous malarial infection. Plehn fully demonstrates that blackwater fever may occur in patients suffering from malaria, without the administration of quinine, and with or without notable exposure to any of the ordinary exciting causes of the condition. Numerous cases collected or observed by Plehn leave this fact apparently beyond doubt. Plehn's report that a special variety of parasite is concerned in blackwater fever has not been confirmed, and it is probable that the author mistook the peculiar appearance of granular degeneration of red cells for minute parasites. The most extensive changes of this type observed by the writer were seen in severe malarial cachexia at Mon- tauk 1898, and in many instances the altered red cells presented a re- markable resemblance to corpuscles infected with young malarial para- sites. On the other hand Koch has recently drawn contrary conclusions from a thorough study of 41 cases. He finds that hemoglobinuria does not necessarily follow even when 80 percent of the red cells are infected with parasites. The great majority of his cases developed a few hours after the administration of quinine. With regard to the re- lation of malaria to blackwater fever he found that : (1) Parasites may be present in the blood, but there is no relation between their numbers and the severity of the hemoglobinuria. (2) The patient may have suffered from malaria some weeks or months previously. (3) In two fatal cases in the tropics he found neither parasites in the blood nor pigment in the viscera. Koch concludes that life in the tropics, and especially malarial in- fection, predispose certain individuals to hemoglobinuria, but that the attack is invariably excited by quinine, or occasionally by exposure to cold or heat, etc. His further claim that no cases are referable to malaria alone cannot be accepted in the face of the observations of Plehn and many others, of fatalities occurring in the course of ma- larial infections which have not been treated by quinine. Changes in the Blood. — The destruction of red cells very promptly reduces their numbers in the peripheral circulation, as shown by the reports of numerous observers. Bristowe and Copeman in a series of attacks induced in the same patient found a maximum loss of red cells of 129,000 to 824,000. In an idiopathic case of moderate severity Grawitz found a loss of 1. 13 millions following a single paroxysm. Us- ually the reduction in cells is not so marked, and Grawitz has explained this fact by showing that, from vasomotor spasm and increased diuresis during the attack, the blood is considerably concentrated, since the dry POST-MALARIAL HEMOOLOBINEMIA. 319 residue of the blood and serum is considerably increased after the seizure, notwithstanding the loss of cells. For the same reason and because the Hb is not immediately removed from the serum, the per- centage of Hb is not excessively low during or immediately after the paroxysm. Indeed Frazer found an increase of 10 percent of Hb about an hour after the beginning of an attack. Usually there is a loss of 5-10 percent. Ponfick claimed that at least one-sixtieth part of the Hb of the blood must be lost before it will appear in the urine. The appearance of the blood in the stained specimen taken shortly after the paroxysm usually gives evidence of active destruction of blood. The red cells may fail to form rouleaux, although having pos- sessed this property just before the attack. Many very pale, or shrunken, or fragmented cells have been observed in some cases, while in others the red cells were found to contain a normal amount of Hb. After repeated attacks megalocytes appear. On the other hand Koh- ler and Obermayer found no morphological changes in a case losing 650,000 red cells and 10 percent Hb. The presence of polychromatic cells or of granular degeneration of red cells has apparently not been re- ported, but in a case of blackwater fever, shortly before death, the writer found the latter form of degeneration in extreme degree. Frank noted in a paroxysmal case that the red cells in the fresh condition looked pale and brownish. The serum has been found distinctly tinged by the dissolved Hb. Bristowe and Copeman found pigment granules, crystals of Hb, and colorless crystals resembling Charcot's. The LEUCOCYTES were found by Frazer moderately increased (21,- 000), but the majority of observers have reported them as normal, or if above normal, as being little affected by the paroxysm. The blood plates have been reported as very much increased (Frazer, Mesnet), as seems very probable from the extensive destruction of red cells. The COAGULABILITY OF THE BLOOD is normal or increased. Hayera ' noted that the rapidly formed clot softens very soon, which Chvostek finds is observed in other conditions. The resistance of the red cells, as determined by Hamburger's method, was found by v. Hoff to have fallen to 38 percent NaCl, and Sabrazes, using Malas- sez' method, also found it reduced. Following the destruction of red cells there is usually a reduction in the alkalinity op the blood,. Robert believes this to be due to setting free of phosphoric and glycerin-phosphoric acids during the separation of oxyhemoglobin from the red cells. The regeneration of the blood is, in uncomplicated cases, very prompt, Bristowe and Copeman finding an increase of 500,000 cells after 5 days, and 600,000 after 6 days. Most confirmed cases, how- ever, suffer from a moderate or marked grade of chronic anemia, sometimes associated with enlargement of the spleen. 320 HEMOBBHAOIC DISEASES AND EEMOBBHAGIO DIATHESIS. Bibliography. Hemorrhagic Diseases. Ajello. Kif. Med., 1894, p. 103. Alhertoni. II Policlinico, 1895, II., p. 177. Arch. Ital. de Biol., T. 28, p. 369. Cent. f. inn. Med., 1895, p. 876. Auche. Jour, de Med. Bordeaux, 1896, 26, p. 209. Babes. Deut. med. Woch., 1893, p. 1035. Becquerel. Gaz. med. de Paris, 1847, p. 141. Bensaude. Bull. Soc. d. Hop., 1897, p. 36. Bignami, Dionisi. XI. Internal. Cong., II., p. 235. Billings. Johns Hopkins Bull., 1894, p. 65. Boas. Deut. Archiv f. klin. Med., 1883, Bd. 32, p. 355. Brandenburg. Berl. klin. Woch., 1895, p. 583. Bristowe, Copeman. Lancet, 1889, II., 256, 307. Buhl. Cited by Litten. Carriere, Gilbert. Compt. Rend. Soc. Biol., 1897, p. 328. Cantanl. Path. u. Ther. d. Stoffwechselkrank., IV. Rhachitis, Scorbut., Leipsic, 1884. Chalvet. Cited by Litten. Chvostek. Ueber d. Wesen. d. parox. Hemoglobinurie, Wien, 1894. Cima. Pediatrio (Naples), 1897, p. 281. Claude. Rev. des Mai. del'Enfance, 1896. Claisse. Archiv de Med. Exper., 1891, p. 379. Cohen. Zeit. f. klin. Med. Suppl., Bd. 17, 1890, p. 182. Dansac. Med. Mod., 1892. Dapper. Diss. Bonn, 1887. Duchek. Med. Jahrb. Wien., 1861, p. 39. Ehrlich. i Charite-Annalen, Bd. 13, p. 300. ^Verh. d. Verein, d. inn. Med., 1881, Mar. 21. K. Ehrlich, Lindenthal. Zeit. f. klin. Med., 1896, Bd. 30, p. 427. Falck. Archiv f. d. ges. Phvsiol., 1889, Bd. 45. Falk. Therap. Monatsch., 1890, p. 97. Fischl, Adler. Zeit. f. Heilk., Bd. 24. Fleischer. Berl. klin. Woch., 1881, No. 47. Frank. Berl. klin. Woch., 1898, p. 1095. Frazer. Edin. Med. Jour., 1897, IL, 313. Garrod. Monthly Jour., Jan., 1848. Gepperi. Zeit. f. klin. Med., Bd. 15, pp. 208, 307. Gotze. Berl. klin. Woch., 1884, p. 716. Grandidier. Die Hemophilic, Leipsic, 1877. Guarnieri. Bull. d. Soc. Lancisiana, May 24, 1888. Hamilton, Yates. Montreal Med. Jour., 1897, p. 117. Hanoi, Luzet. Archiv de Med. Exper., 1890, p. 772. Hayem. » Compt. Rend. Acad. Sci., 1896, p. 894. 2 Gaz. med. de Paris, T. 26, p. 126. 3 Gaz. hebdom., 1889, No. 11. Henoch. Vorles. u. Kinderkrank., II. Auflage. Hirsch. Handb. of Geog. and Hist. Path., Syndenham Soc. Trans., 1885, Vol. II. V. 1 Hoff. Berl. klin. Woch., 1897, p. 745. Hoppe-Seyler. Zeit. f. physiol. Chem., 1881, Bd. 5, H. 1. Immerman. Ziemssen's Handb., 1879, Bd. 13, 2. Kohert. Lehrbuch d. Intoxicationen, Stuttgart, 1893, p. 464. Koch. Zeit. f. Hygiene, Bd. 30. Kohler, Obermayer. Zeit. f. klin. Med., Bd. 13, p. 163. Kolb. Arb. a. d. Kais. Gesundheitsam., VII., p. 60. Kronig. Berl. klin. Woch., 1895, p. 998. Layral. La Loire Med., 1897, p. 255. Lebreton. Cent. f. Path., 1895, p. 834. Legendre. Soc. med. des Hop., 1893. Legg. St. Barthol. Hosp. Rep., 1874, X., p. 71. Lepine. La Semaine Med., 1886, p. 473. Letzerich. Untersuch. u. Kenntniss d. Purp. hem., 1890. Levi. Cited by Callarl, La clin. Mod., 1898, p. 226. Lewis. N. Y. Med. Record, 1898, 53, p. 769. Litten. Hemophilia, Nothnagel Sp. Path., Th. III., Bd. 8. I BIBLIOGRAPHY. 321 Lossen. Deut. Zeit. f. Chir., 1876, Bd. 7. McKenzie. Lancet, 1879, II., p. 725. McLeod. Lancet, 1897, I., 1410. Manteuffel. Deut. med. Woch., 1883, p. 665. Marchand. Archiv f. exper. Path., 1887, Bd. 22, 23. V. Mering. Das chlorsaure Kali, Berlin, 1885. Mesnet. Archiv gen. d. Med., 1881, May. Miller. The Microors:anisms of the Mouth, Phil., 1890. Muller. Deut. med. Woch., 1887, p. 27. Murri. i Rev. clin. Bologna, 1881, p. 215. « Ibid., 1885, No. 4. Neumann. Archiv f. Kinderheilk. , 1891, p. 54. Oertel. Therap. d. Kreislaufstor, Leipsic, 1854. Opitz. Prager Vierteljahrschrift, 1861, p. 108. Pavy. Lancet, 1866, II. , p. 33. Petrone. Gaz. d. osped. Milano, 1883, IV., 561. Riv. clin., 1887. ' Plehn. Archiv f. Schiffs u. Tropen Hygiene, 1899. Ponfick. Virchow's Archiv, 1888, p. 445. Popper. Oesterr. Zeit. f. prak. Heilk., 1868, p. 657. Recklinghausen. Handb. d. Allg. Path., 1883. Rolfe. Lancet, 1877, II., 81, 955. Rodet. Cited by Grawitz. Rosenbach. Berl. klin. Woch., 1880, p. 132 ; 1884, p. 751. Rosenell. Cited by Litten. Sabrazes. Compt. Rend. Soc. Biol., 1897, p. 923. Schmidt. Vide Manteuffel. Seeland. St. Petersburg med. Woch., 1882, p. 13. Silvestrini. II Policlinico, 1897, p. 13. Spietschka. Archiv f. Derm. u. Syph., 1881, p. 265. Stahel. Die Hemophilia im Wald. , Zurich, 1880. Stengel. 20th Cent. Practice, Vol. 7. Stewart. Cited by Legg, St. Barthol. Hosp. Rep., Vol. X. Sticker. Miinch. med. Woch., 1895, p. 644. Strobing. Deut. med. Woch., 1882, No. 1. lallquist. Exper. Blut-gift Anemie, Berlin, 1900. Thoma. Textbook of Pathology. Tomasselli. Congresso di med. interne, 1888. La intossicazioni chimica, etc., Catania, 1897. Uskow. Cent. f. med. Wissen., 1878, p. 499. Wertheim. Wien. med. Presse, 1868, p. 309. Widal, Therese. Cent. f. Path., 1895, p. 835. Wieruschky. Wratsch. 1889, cited by Litten. White. Med. Times and Gaz., 1883, 280. Wyss. Deut. med. Woch., 1894, p. 296. Zalesky. Arch. f. exper. Path., 1897, Bd. 23, p. 77. 21 CHAPTER XVIIL MISCELLANEOUS CONSTITUTIONAL DISEASES. DIABETES MELLITUS. The results of older observers led to conflicting opinions regarding the condition of the blood in diabetes, since it appeared that in cases of about equal severity the Hb, red cells, and albumens, might in one patient be increased and in another diminished, or might vary consider- ably at different times in the same patient, not infrequently rising above normal. The studies of Henocque were among the first showing that while individual differences may be considerable, yet in a series of cases the Hb does not vary greatly from the normal. Leichtenstern, finding an excess of Hb in an advanced case and a diminution in an early case, was led to refer the anomaly to the effects of diuresis in concentrating the blood. Although the extent of the diuresis depends largely upon the amount of sugar secreted, a very close relation between the Hb of the blood and the sugar of the urine has not apparently been estab- lished. Yet the above logical explanation has been generally accepted, and the marked changes observed in the blood of many diabetics may be referred to the uncertain balance between the amount of water ab- sorbed and that excreted in the urine. In the late stages of the disease other factors enter to further dis- turb the condition of the blood. General failure of nutrition belongs to the advanced disease and the blood must suffer proportionately. Yet even in extreme cases the anaemia is commonly masked by the concentration of the blood through diuresis, and it becomes necessary to class the blood of advanced diabetes, with excess of Hb and cells, as among the examples of oligemia or oligoplasmia. From these considerations it is evident that the estimation of Hb and the counting of red cells is of little use in diabetes unless very carefully controlled by reference to the diet and urinary excretion. Thus, the writer found in an emaciated anemic case with advanced phthisis, 78 percent of Hb and 4,100,000 red cells. In most re- ported cases in which the daily urine measured over three liters the red cells approached the normal figure. In Grawitz's case of coma the red cells rose from 4.9 millions, three weeks previously, to 6.4 millions 5 hours after the onset of the coma, and the dry residue of the blood from 21.4 percent to 24.75 percent. Possibly the result was in part referable to cyanosis. Habershon re- ports 3 cases of coma with excess of red cells (maximum 6.64 millions). DIABETES MELLITUS. 323 They all suffered from dyspnea, but in one the cells were lower than they were a week before death. In a cyanotic comatose case examined by the writer there were 6.8 million red cells. The general character of the blood in diabetes is well illustrated in the cases reported by James, as follows : No. Red Cella. Hb. Specific Gravity. Roy's Method. 1 6.73 66 1.056 2 6.10 61 1.059 3 4.80 58 4 5.25 60 5 5.60 65 6 3.55 52 1.054 7 52 1.060 8 5.30 75 1.056 9 6.28 108 1.055 10 5.38 96 1.055 11 5.64 112 1.056 12 6.20 112 1.057 13 4.46 55 1.054 14 6.00 96 James concluded from these results that the excess of red cells in diabetes can not result from concentration of the blood, otherwise the specific gravity would have been higher. Yet, considering the per- centage of the Hb, the gravity in the above cases is distinctly above the usual figure, indicating relative anhydremia with marked reduction of Hb, but with slight loss of red cells. The leucocytes, in the average hospital case if uncomplicated, are normal, subnormal, or moderately increased. In 7 of 20 untreated cases Habershon found less than 10,000 leucocytes, in the others a leu- cocytosis reaching a maximum of 19,800, and being rather uniformly present in the severer cases. With the advent of severe symptoms he found the leucocytes to increase, and in three cases of coma there was pronounced leucocytosis (maximum 28,500). Cabot found no leucocy- tosis in 13 cases, and only 4,200 in a comatose patient. The writer observed leucocytosis in four patients, one comatose, but autopsies re- vealed phthisis, suppuration, or pneumonia. Fat, occurring in traces in normal blood, is usually increased in dia- betes and may become extremely abundant. Usually microscopical ex- amination is required to demonstrate the minute extra-cellular glob- ules in specimens stained by osmic acid, but often the fat rises to the surface in demonstrable quantity if the specimen is centrifuged in the hematokrit, and occasionally the blood is milky from the abundance of fat. Its origin in the blood is not understood, but v. Noorden says that it is not connected with a previous ingestion of fat, as is the case with some other forms of lipemia. Chemistry. — The chief chemical alteration demonstrated in the blood is a pronounced increase in sugar. While normal blood contains .05— .15 percent of sugar, Pavy and Seegen have found as much as .6 324 MISCELLANEOUS CONSTITUTIONAL DISEASES. percent in severe diabetes, and Frerichs found the sugar of the blood to vary between .38-.44 percent, while that of the urine amounted to 5.5_8.4 percent, Naunyn obtained .7 percent in the blood of a fatal case in which the urine contained 4 percent. In Klemperer's " renal diabetes," the sugar of the blood remains normal.^ Henriques and Kolish have lately claimed that an excess of pre- formed sugar exists in the blood only in alimentary glycosuria, while in diabetes the preformed sugar of the blood is but slightly above normal. They find, however, that in diabetes the blood contains a marked excess of jecorin in combination with albumens which, during excre- tion, is split up into sugar and lecithin. Lepine and Barral believe that diabetes results from the failure of a glycolytic ferment elaborated by the pancreas and normally present in the blood, but this theory has not been demonstrated. (See Glycolytic Ferment.) Glycogen in the blood of diabetes has been demonstrated in marked cases by Gabritschewsky, who found it both in plasma and leucocytes, but the significance of the brownish extra-cellular granules has been brought into question. Livierato could find but little glyco- gen in the plasma and none in the leucocytes. Futterer claims to have demonstrated in the brain and medulla many thrombi composed of glycogen. Of 5 cases of diabetes the writer found in each a few leu- cocytes with many glycogen granules, while in the plasma were brown- ish staining granules of w^iich the significance seemed uncertain. The glycogen-holding leucocytes were less abundant than in cases of pneumonia. Alkalinity. — The alkalinity of the blood in normal subjects varies between 300 and 400 mg. of NaOH per 100 grms. of blood, and is always diminished in patients who are much reduced in strength. In diabetes when the patients are suffering little general disturbance, the alkalinity of the blood has been found to be very slightly reduced. (Minkowsky, Kraus, Lepine.) When diabetes reaches the stage of excretion of oxybutyric acid in the urine, the alkalinity of the blood has been found to be lower than in any other known condition, falling as low as 40 mg. of NaOH per 100 of blood. (Van Noorden.) In diabetic coma, Minkowsky found as low as 3.3 vol. of CO2 per hun- dred of blood. This relative acidemia has been referred to the presence in the blood of various acid products of proteid metabolism, /9-oxybutyric, and other fatty acids. These acid products are supposed to act either as direct poisons, or indirectly, by lowering the alkalinity of the blood. The basis of this theory is found partly in the demonstration of diminished alkalinity in the blood, but largely in the abundance of these acids or their salts in the urine. OBESITY. The majority of healthy obese subjects show an excess of Hb, as is clearly indicated by the extensive observations of Kisch, who found a ADDISON'S DISEASE. 325 notable excess, and maximum of 120 percent, of Hb, in 79 of 100 fat subjects, while in 21 it was diminished, reaching 55 and 60 percent in some instances. The existence of distinct anemia in some apparently healthy fat subjects had long been accepted, but the opinion was ren- dered much more certain by Leichtenstern's observation of four such cases, showing considerable reduction of Hb. Oertel also found an excess of 5-8 percent of Hb, and Grawitz reported an increase of the dry residue of the blood in very fat subjects. It appears, therefore, that there is a distinct tendency toward true plethora associated with obesity, while the anemia occasionally observed, as in the majority of Kisch's anemic cases, is referable to other causes. Grawitz, in two cases, observed a considerable reduction in dry resi- due both of the whole blood and of the serum, as a result of a course of treatment which reduced the body weight 10-12 lbs. That the plethora of obesity is sometimes more apparent than real is suggested by the constant perspiration, frequent dyspnea, and ten- dency toward venous congestion, from which these subjects suffer. ADDISON'S DISEASE. Established cases of this disease are usually attended with a severe grade of anemia. In four cases reported by Tschirkoff the red cells be- fore treatment numbered from 2,733,000 to 3,280,000. In three of Cabot's cases, and in three examined by the writer, the same grade of oligocythemia was observed, while Neumann reported more severe anemia with 1.12 million red cells in one case. Morphological changes in the red cells are not marked, but microcytes are sometimes quite abundant. Leucocytes are commonly diminished, sometimes slightly increased, and are subject to the usual variations of secondary anemia. Eosino- phile cells were present in high normal proportions in Cabot's and in the writer's cases. Opposed to the typical cases are others reported in which the red cells were in excess of 5 millions, reaching, in Neumann's observation, the remarkable figures of 7.7 millions, while Tschirkoff found that the Hb in early cases might exceed the normal. It appears likely that disturbances in the circulation may be responsible for some of these anomalous results, but the subject requires further observation. The deposits of pigment in Addison's disease have been referred by Riehl and Afanissiew to the results of minute thrombi and hemorrhages in the superficial vessels, and Afanissiew observed the disintegration of red cells in local pigmented areas, v. Kahlden, however, showed these lesions to be inconstant and probably of secondary significance. An interesting contribution to the pathology of Addison's disease, bearing on this point, has been furnished by Tschirkoff, who examined the blood of two cases with special reference to the pigments, by means of Glan's spec- trophotometric method, by which he was able to estimate the relative amounts of oxy- and reduced Hb. By this method he found in advanced 326 MISCELLANEOUS CONSTITUTIONAL DISEASES. cases a much greater proportion of reduced Hb than is present in normal blood, and at times more reduced Hb than oxy-Hb. During improvement he found an increase in the proportion of oxy-Hb at the expense of the reduced Hb but without coi-responding increase in the total Hb of the blood. He found evidence pointing to the presence of methemoglobin in the blood, but could not observe any quantitative relation between these changes in the Hb of the blood and the pigmentation of the skin. He concluded that Addi- son's disease is associated with a qualitative rather than quantitative change in the hemoglobin of the blood, with a corresponding failure to furnish oxy- gen to the tissues. His results have not yet been verified. GOUT. From the comparatively few recorded observations it is evident that acute gout has little eiFect upon the red cells and Hb of the blood, in which respect it differs from acute rheumatism. In chronic gout, also, Duckworth found cases with inappreciable anemia and concluded that when the disease is associated with anemia, the condition of the blood is referable principally to complications such as hemorrhage, nephritis, lead poisoning, bad hygiene, etc. The writer found uniform but mod- erate anemia in a series of chronic cases examined at Roosevelt Hos- pital, the patients all coming from the poorer classes. The leucocytes are probably increased in acute attacks, but there are apparently no direct observations on this point. In the chronic cases leucocytosis of moderate grade may be observed, but it is difficult to determine its relation to the gouty process, as many of these patients suffer from other complaints. Neusser ' mentions the usual presence of mixed leucocytosis in a series of 100 cases of the uric acid diathesis. Chemistry. — The chemistry of the blood in gout has for many years been the subject of extensive investigation and while the presence of abnormal principles has been demonstrated, an important relation of these principles to the pathogenesis of the disease has not been estab- lished. Garrod first demonstrated an excess of uric acid in the shed blood, finding .025—175 grains of uric acid for 1,000 grains of serum, in five cases during and shortly after the attack, and this uricacidemia has been noted by Salomon,^ v. Jaksch, and Klemperer.' Recently Magnus- Levy ^ failed to find, in several cases, any constant variations in the amount of uric acid which might be regarded as an increase. Garrod perfected a ready clinical method for the demonstration of uric acid in the blood or serum of gouty patients, commonly known as the '* thread test." One to two drams of serum from fresh blood drawn from a vein are placed in a broad, flat dish, about 3 inches in diameter and + inch deep, and to it are added 6 minims 33-percent acetic acid to each dram of serum. A couple of old but clean linen threads are submerged in the fluid and the ves- sel is allowed to partially evaporate at room temperature for 36-60 hours, when the threads will be found to have gathered minute crystals of uric acid, if any considerable trace is present in the blood. They may be washed in water and freed from any adherent phosphate and identified under the mi- croscope. Garrod found that crystals begin to deposit on the thread when OSTEOMALACIA. 327 the serum contains .025 gr. of uric acid per 1,000, and become very numerous with. 08 gr. per 1,000. The test must be carefully performed and may fail, especially from too rapid or prolonged evaporation or from heating above 75° F. Serum obtained from a blister may be used, but not when obtained from the neighborhood of an inflamed joint. Excess of uric acid in the blood is not, however, pathognomonic of gout, being absent in some undoubted cases (Duckworth), and being abundantly present in many other conditions, such as pneumonia (Sal- omon),^ and other conditions with leucocytosis, emphysema, nephritis, severe anemia (v. Jaksch ^), leukemia (Magnus-Levy),^ and after a diet rich in nucleo-proteids (Weintraud). Alkalinity. — The assumption by Garrod that an excess of uric acid in the blood is associated with diminished alkalinity has not been sup- ported by late observations. Klemperer,^ estimating the alkalinity by the content of the blood in CO2, found no distinct variations from the normal in three cases during the attack. Luff, using Wright's method, also failed to find any loss of alkalinity during the course of an acute attack, and Magnus-Levy, using Lowy's method, failed to find any con- stant variations in a series of 16 cases examined at various periods. From the recent studies of the chemistry of the blood in gout, it is evident that Garrod's theory of the nature of the disease must be abandoned. OSTEOMALACIA. The usual condition of the blood in osteomalacia appears to be that of moderate chlorotic anemia. Such cases are reported by Eisenhart, Tschistowitch, and others, while Seligman, in an advanced case, found 41 percent of Hb. The red cells are usually normal or very slightly reduced. Shortly after abortion Tschistowitch's case showed 3.1 million cells, but the numbers increased rapidly, later to oscillate between 3.4 and 6.8 mil- lions. One of Rieder's cases had 3.8 million cells, another 4.9 millions. The leucocytes have varied from subnormal to moderately increased numbers. The lymphocytes are usually very numerous, an excessive proportion (maximum 56 percent) having been found by Tschistowitch, while Ritchie observed moderate lymphocytosis. Myelocytes have been found by Neusser and by Tschistowitch. Neusser ^ proposed to divide cases of osteomalacia into two groups, one showing myelocytes in the blood, the other showing high propor- tions of eosinophile cells, of which he had seen examples. High normal proportions of eosins have been found by several observers, but not by others, viz., Fehling, Sternberg, and Chrobak, who were prob- ably looking for a distinctly abnormal number. In Tschistowitch's case their numbers varied greatly during the year over which his ob- servations extended. Chemistry. — The alkalinity usually varies between normal limits (Limbeck) but has been found both increased and diminished. Lactic acid has never been fully demonstrated in the blood. Peters claimed 328 MISCELLANEOUS CONSTITUTIONAL DISEASES. to have found in the blood a nitrifying micro-organism, cultures of which produced in dogs a condition resembling osteomalacia. This claim has not been established. RACHITIS. The state of the blood in rachitis varies with the extent and sever- ity of the primary disease and is markedly affected by complications. Nearly all series of cases reported contain some examples of pro- nounced rachitis in which the red cells were nearly normal and the Hb very moderately reduced, and some observers, as Felsenthal and Morse, did not meet with severe anemia in any case. Nor does it ap- pear that the anemia bears a very close relation to the extent of the disease, although as a general rule the advanced cases with enlarge- ment of the spleen show greater anemia than do those without involve- ment of the viscera. (Monti, Berggrun.) It is agreed on all sides that the disease is not associated with any peculiar type of anemia, al- though the hyperemia of the bone marrow might be expected to yield an unusually large number of nucleated red cells. Yet these cells are apparently not more abundant in rickets than in congenital syphilis (Monti), while Gundobin found the same grade and type of anemia in rachitic as in other poorly developed infants. Rachitis figures about equally with syphilis, etc., in the etiology of v. Jaksch's anemia. The existence of rachitis in patients with normal blood raises a question as to the cause of the anemia usually found in the disease, and the conclu- sion seems justified that the anemia of rickets is largely referable to the gen- eral malnutrition of the patient and to a variety of complications, especially gastroenteritis. Moreover the pathological nature of the disease renders it extremely improbable that the lesion in itself has very marked effect upon the blood. So far as our present knowledge goes, the anemia of rickets is practically that of marasmus and gastroenteritis. Simple chlorotic anemia is the usual condition of the blood of rachitic children. No other forms were encountered among the cases of Felsenthal and Morse, most of which were described as of moderate grade ; and a few examples of the same type have been described by others. Although the red cells may be above 5 millions, the Hb is invariably reduced, the color-index frequently falling as low as in primary chlorosis. Morse found an average Hb-index of .7, and Felsenthal obtained only 50 percent of Hb in one case with over 5 million cells. The usual morphological changes in the red cells are to be observed, and a few nucleated red cells are nearly always to be found in the young subjects. The leucocytes in cases with mild anemia are usually found at the higher normal limits or distinctly increased. Rickets does not appear among Monti's cases of simple anemia with- out leucocytosis. Grave secondary anemia is observed in many cases of rachitis in which there are almost always some serious complications. Although the progress of rachitic anemia is sometimes moderately rapid, the cause RACHITIS. 329 of the destruction of red cells is not always apparent, v. Jaksch saw the red cells fall from 1.6 millions to 750,000 within three months, while Luzet saw a reduction of 500,000 cells in three weeks. Usually the loss of cells occurs more slowly, and in the grave forms of anemia the impoverishment of the blood is usually attended with enlargement of the spleen, and often of the liver. The changes in the red cells are then similar to those of other forms of severe secondary anemia of chronic course, but poikilocytes and nu- cleated red cells are unusually numerous. The Hb-index varies. In a few instances, especially those of more rapid course, the Hb is defi- cient, but usually the Hb-iudex rises and may become abnormally high. Thus, Monti and Berggrun found 50-55 percent Hb with 2.35 million cells. In all cases of this type there is leucocytosis. Progressive pernicious anemia appears to be a relatively infrequent complication of rickets. In 11 cases of pernicious anemia in young children or infants, collected by Monti and Berggrun, rickets existed or was mentioned in only two, and one of these was complicated by congenital syphilis. The writer has had a similar experience, never having found the well-marked characters of progressive pernicious anemia in a rickety child. Leucocytes. — It has already been shown that in many mild cases of rickets the leucocytes do not exceed the normal limits for young in- fants. The slight excess over 10,000 seen in most of the mild cases reported is not, as Cabot points out, abnormal for infants. Felsenthal found 30,000 white, with 4 million red cells, in one mild case in which no complication was noted. In the severer cases leucocytosis is nearly constant, but does not usually exceed 30,000. Yet Luzet believes that there are all grades of the leucocytosis of rickets up to that of leukemia, where it is certainly difficult at times to distinguish the leucocytosis of a rachitic anemia from that of v. Jaksch's anemia. The lymphocytes are usually quite numer- ous, but not often excessive for the age. Morse found an average of 43 percent of lymphocytes in twenty rachitic children under two years of age. Distinct lymphocytosis was observed by Rieder, Weiss, Monti and Berggrun, and Morse, and, while probably not uncommon in young subjects, is without special significance. Eosinophile cells are often relatively numerous. Morse found an average of 3 percent. They may be distinctly increased, as Hock and Schlesinger found 20 percent in one subject and Weiss, 16 percent in another, or they may be scanty. (Rieder, Cabot.) The conditions leading to excess of leucocytes in rickets are not well understood. Limbeck refers the excess principally to the co-existent gastro-enteritis which is often present. Whilst most cases with leu- cocytosis show hyperplastic splenitis, these two conditions are most probably separate effects of a common cause. Not all cases with en- larged spleen show leucocytosis, which is commonly absent in pro- nounced cases of " splenic anemia " in infants. Luzet, however, con- nects the excess of leucocytes with the hyperplasia of the spleen. 330 MISCELLANEOUS CONSTITUTIONAL DISEASES. It seems not improbable that the hyperemia of the marrow may tend to discharge an unusual number of white cells into the circulation, a possibility which is favored also by the large number of nucleated red cells commonly seen, by the frequent presence of a few leucocytes with mitotic nuclei' (Hock, Schlesinger), and by the nearly constant occur- rence of myelocytes. Chemistry. — The specific gravity of the blood was reduced in pro- portion to the loss of Hb in the cases of Hock and Schlesinger. Even in patients who appeared otherwise in good health the specific gravity of the blood was distinctly reduced in 12 cases examined by Felsenthal and Bernhard. MYXEDEMA. The anemia of myxedema, in the majority of pronounced and un- treated cases, is of the secondary chlorotic type, with slight leucocy- tosis, and of moderate grade. Bramwell observed well marked anemia in 26 of 33 cases. Of 23 cases collected by Murray the blood was normal in 7, while in the cases of Cabot and in those of Kraepelin there was no anemia. The most severe anemia recorded was that of Le Breton's case, in a child of three years, with 1.7 million cells, 65 per- cent of Hb, and 4,500 leucocytes. Many other cases have shown a reduction of the red cells to about 3 millions (Bramwell, Putnam, and 2 cases of the writer's). Kraepelin noted by careful measurements a distinct increase in the diameters of the red cells in 4 cases, and claimed this feature to be one of the typical symptoms of myxedema. The chemical examinations of the blood of these patients supported this conclusion, by showing excess of solids. Le Breton also noted a sim- ilar condition in the blood of his case but the patient was very anemic (1.7 million cells). Cabot failed to find any such changes in the red cells of three cases. When present it is probably referable to the increased venosity of the blood which, according to Horsley, is commonly pres- ent in myxedema. The Hb is usually moderately reduced, but less than in most other forms of secondary anemia. In Le Breton's case the Hb-index was abnormally high (L8), but since large nucleated red cells were present the condition of the blood seems to accord with that of pernicious anemia. In two of Kraepelin's cases, while the Hb by Fleischl's method was not excessive (5.4, 5.6 million cells, 93-98 percent Hb), by the spectrophotometric method an abnormal quantity of blood pigment was demonstrated. The specific gravity was also abnormally high. In several other cases, even when the red cells were considerably deficient, the Hb-index was normal or increased. (Bramwell.) Putnam re- ports a few myelocytes in one case in which the blood was otherwise nearly normal. Thyroid treatment has usually been followed by marked improvement in the blood. In 6 weeks Kraepelin's case gained 700,000 red cells, and nucleated red cells and leucocytes disappeared. Putnam observed polythemia (5.7 millions) after 6 months' treatment. Marked anemia BIBLIOGRAPHY. 331 followed the administration of over-doses of thyroid extract in one of Bramwell's cases, but on smaller doses the patient did well. The course of the blood changes in this case was as follows : Red cells. Hb. October 28 (before treatment) 3.8 millions. 65 percent. November 28 (after acute thy roidism) 2.6 " 54 " December 21 (smaller doses) 3.8 " 68 " January 13 4.3 " 70 " Leucocytes are usually not increased, but in 6 of Murray's 23 cases there was leucocytosis. In uncomplicated cases without severe anemia it appears to be nearly always absent. The proportions of the various leucocytes are not altered. Cabot found 5 and 4.4 percent of eosins in two cases, but Schmidt found only 1.5 percent of these cells. Putnam reports a few myelocytes in one case in which the blood was otherwise nearly normal. Chemistry. — Very numerous and elaborate studies of the chemistry of the blood in myxedema, and especially in thyroidism, have failed as yet to demonstrate the exact nature of this peculiar toxemia. Halli- burton in 1885 demonstrated mucin in the blood of thyroidectomized dogs, an observation which has been fully confirmed, and Levine has recently produced peculiar chronic toxemia by injection of mucin in animals after partial thyroidectomy, although similar injections in healthy animals were innocuous. Yet in the human subject an excess of mucin in the blood has not been demonstrated. The chemical analysis of Kraepelin's cases, by Schneider, showed the exaggerated effects of chronic venosity of the blood, a condition first noted clinically in myxedema by Horsley. Here there was considerable increase in specific gravity of the blood (1.0625-1.0636) and of the serum (1.0317-1.0329), while the dry residue of whole blood, serum, and red cells was also increased. Bibliography. Constitutional Diseases. Afanissiew. Cited by Tschirkofl' (Russian). Arthus. Archiv de Physiol., T. 24, p. 337. Bramwell. Anemia. Chrobak. See Fehling. Duckworth. Treatise on Gout, London, 1890. Eisenhart. Deut. Archiv Klin. Med., Bd. 49, p. 181. Fehling. Zeit. f. Geburtsh., Bd. 30, p. 471. Felsenthal. Archiv f. Kinderheilk., Bd. 17, p. 333. Frerichs. Diabetes, Berlin, 1884. Futterer. Verh. d. ph. med. Gesell., Wurzburg, 1888, p. 112. Gabriischewsky. Archiv f. exper. Path., Bd. 28, p. 272. Garrod. Gout and Rheumatic Gout, London, 1876, p. 80. Gundobin. Jahrb. f. Kinderheilk., 1893, Bd. 35, p. 187. Hahershon. St. Barthol. Hosp. Rep., 1890, p. 152. Halliburton. Cited by Horsley. Henriques. Zeit. f. Physiol. Chemie, Bd. 26. Hock, Schlesinger. Beitrage zur Kinderheilk., Wien, 1892. Horsley. Brit. Med. Jour., 1885, I., p. 419. V. Jaksch. iZeit. f. Heilk., 1890, p. 415. James. Edin. Med. Jour., 1896, Vol. 62, p. 93. 332 MISCELLANEOUS CONSTITUTIONAL DISEASES. V. Kahlden. Vircliow's Archiv, Bd. 114, p. 112. Kisch. Zeit. f. klin. Med., Bd. 12. Klemperer. 'Deut. med. Woch., 1895, No. 40. ^Untersuch. u. Gicht., Berlin, 1896. "Deut. med. Woch., 1896, p. 46. Kolish, Stejskal. Wien. klin. Woch., 1897, p. 1101. Kraepelin. Deut. Archiv klin. Med., Bd. 49, p. 587. Kraus. Archiv f. Heilk., 1889, p. 106. Krypiakiewiez. Wien. med. Woch., 1892, No. 25. Le Breton. Bull. Soc. des Hop., 1895, p. 22. Leichtenstern. Untersuch. u. d. Hmglb'n, 1878. Lepine. Kevue de Med., 1887, p. 224. Lyon Med., T. 62, p. 619 ; T. 63, p. 83. Lepine, Barral. Lyon Med., T. 62, p. 619 ; T. 63, p. 83. Levine. Proc. N. Y. Path. Soc, 1900. Livierato. Deut. Archiv klin. Med., Bd. 53, p. 303. Luff. The Pathology of Gout, London, 1898. Luzet. Etude sur les Anemies de I'Enfance, Paris, 1891. Magnus-Levy. 'Zeit. f. klin. Med., Bd. 36, p. 366. ^Virchow's Archiv, Bd. 152, p. 107. Minkowsky. Archiv f. exper. Path., 1884, Bd. 18, p. 35. Berl. klin. Woch., 1892, No. 5. Monti, Berggrun. Chronische Anemie d. Kindesalter, 1893. Morse. Boston Med. Surg. Jour., Vol. 136, p. 369. Murray. 20th Cent. Practice, IV., p. 710. Naunyn. Nothnagel Spec. Path., Bd. 7, Th. 6, p. 150. Neumann. Deut. med. Woch., 1894, p. 105. Neusser. 'Wien. klin. Woch., 1892, p. 41. nVien. klin. Woch., 1894, p. 727. V. Noorden. 20th Cent. Practice, Vol. II., p. 99. Oertel. Allgem. Ther. d. Krieslauf st. , Leipsic, 1891, p. 36. Pavy. Lancet, 1878, I., p. 557. Peters. Rif. Med., 1892, No. 78, pp. 119, 163. Putnam. Amer. Jour. Med. Sci., 1893, Vol. 106, p. 125. Biehl. Zeit. f. klin. Med., Bd. X. Ritchie. Edin. Med. Jour., 1896, Vol. 42, p. 208. Salomon. • Charit^-Annalen, 1880, V., p. 137. "Zeit. f. physiol. chemie, 1878, p. 65. Seegen. Diabetes Mellitus, Berlin, 1893. Seligman. Cent. f. Gyn., 1893, pp. 374, 649. Sternberg. Zeit. f. klin. Med., Bd. 22, p. 265. Tschirkoff. Zeit. f. klin. Med., Bd. 19, Sup., p. 87. Tschistowitch. Berl. klin. Woch., 1893, p. 919. Weintraud. Deut. med. Woch., 1895, V. B., p. 185. Weiss. Jahrb. f. Kinderheilk. , Bd. 35, p. 146. ! I CHAPTER XIX. NERVOUS AND MENTAL DISEASES. MANIA, GENERAL PARESIS, MELANCHOLIA, DEMENTIA, EPILEPSY. The blood of inmates of insane asylums has been the subject of much careful study. In 1873 Sutherland examined the blood of 149 insane patients, finding in 4 of them grave secondary anemia with leucocytosis. Ex- cess of leucocytes occurred in many other milder or earlier cases. McPhail contributed the first scheme of the blood changes in gen- eral paresis finding from a study of fifteen cases that : 1. The Hb was moderately low on admission (62.70 percent), im- proved for a time, owing to the better hygienic conditions prevailing in the institutions, but fell (52.66 percent) in the terminal stages of the disease. (2) The red cells diminished steadily with the progress of the disease, reaching in one case 3.4 millions. (3) Leucocytosis (minimum 12,700) was noted in all cases, progressed with the disease and reached a well-marked grade in the terminal stages, 27,700 to 36,600. Lewis, Steele, Seppili, and others obtained results very similar to the above and added many new details. Smyth examined the blood in cases of melancholia, epilepsy, general paresis, and secondary de- mentia. He found the red cells as a rule below the normal, while the Hb was reduced in much greater degree. The average results ob- tained were as follows : Cases. Hb?(. Red cells. S. G. During ConTuIsions. Melancholia 10 50 40 12 69.7 62.8 68.7 53.7 4. 684 mil. 4.520 " 4.700 " 4.070 " 1.0572 1.0568 1.0605 1.0612 EuileDsv 1.0596 General paresis Secondary dementia The more marked anemia, with high specific gravity, of secondary dementia, and the increase of specific gravity after epileptic convulsions, which was not invariable, are specially worthy of note. Winckler also found a distinct loss of Hb, especially in the depressed forms of insanity. Paroxysms of mania or of melancholia caused a loss of both cells and Hb, which was restored after the recovery from the attack. Terminal dementia was associated with a progressive anemia. Capps examined 12 cases of general paresis in which the red cells never fell below four while never reaching five millions (average 334 NERVOUS AND MENTAL DISEASES. 4.789 millions). The Hb varied between 73.92 percent (average 85 per- cent), and the specific gravity between 1.058-1.066. Convulsions ap- peared to concentrate the blood. JelliiFe, working at the Binghampton State Hospital, altitude 400 meters, found moderate polycythemia in 14 of 17 cases, but the Hb was usually reduced, in one case to 52 per- cent. The specific gravity ranged from 1.047 to 1.060. Langdon and Bamford, at the Hudson River State Hospital, obtained in a reasonably short time 20 very anemic cases of melancholia, with red cells between 2-3 millions, Hb between 60-85 percent. Treatment by bone mar- row rapidly improved the condition of the blood. Severe anemia, red cells 2.5 millions, Hb 25 percent, is also reported by Howard in a case of terminal dementia, while in Steele's 35 cases of melancholia the aver- age of red cells was 3 millions, of Hb, 75 percent. The Leucocytes. — In early cases there is usually no leucocytosis, and the uniform tendency to leucocytosis observed by McPhail has not been found by all subsequent observers. Smyth reported an excess in many cases but without relation to the symptoms of the disease. Most of Capps' patients had leucocytosis, as did also the very anemic subjects examined by Langdon and Bamford, while in the tables of Jelliffe and Somers, leucocytosis was absent in the majority of cases. Leucocytosis is more frequent when there is anemia, in the last stages of the disease, and it has been found rather more frequent in the more acute forms of insanity and in terminal dementia. Somers found an average of 8,315 leucocytes in 19 cases of mania, 7,947 in 19 cases of melancholia, 10,473 in 19 cases of dementia, and 8,800 in 5 cases of general paresis. In 16 cases of epilepsy Kohlmann obtained dis- tinct leucocytosis in only one (14,000). Pease followed the leucocy- tosis long enough to show that it is often transitory. The proportions of the various forms of leucocytes show no constant or significant variations. When the total numbers are increased the polynuclear forms are usually most affected. Lymphocytosis is fre- quently observed, especially in epilepsy. The eosins show irregular variations, being absent in some cases, reaching high normal figures in others, while Roncorini found a marked excess, up to 25 percent, in maniacal cases. Krypiakiewicz also found an increase of eosins in the more acute forms of insanity, and comparatively low proportions in the chronic forms. Capps could not find that the restlessness of the pa- tient was always connected wnth any increase in eosins. To recapitulate the results of the above studies, it has been shown that the common forms of insanity may develop in an anemic subject or in one whose blood is normal. The disease may run its course without anemia, but usually the state of the blood accords with the general state of nutrition. In some cases, especially of melancholia, the blood suffers more than the general nutrition. Leucocytosis is frequently observed, especially in anemic subjects, in late stages, and after convulsions. Convulsions may concentrate the blood and induce leucocytosis. Acute mania is not associated with any specific changes in the morphology of the blood. CHOREA. 335 BERI-BERI. This form of infectious neuritis is often associated with severe or even pernicious anemia (Spencer), but in the cases which reach America the anemia is usually moderate and of the chlorotic type. In three recently imported cases of the disease, each with moderate fever (100- 101.5° F.), the red cells ranged between 3 and 3^ millions, and were very deficient in Hb, while showing no other changes, except for the presence of microcytes. Leucocytosis was absent and the eosins were not increased. Cabot records an afebrile case with 3.9 million red cells, 48 percent Hb, and 7,800 leucocytes. Daubler found normal numbers of red and white cells, but increase of fat, in the blood of three cases. Numerous attempts to demonstrate the infectious agent in beri-beri are thus far negative. Various bacteria have been demonstrated in the blood post-mortem. The latest contribution is that of Fajardo who believes that he has seen in the red cells of 59 cases a protozoan resembling the parasite of Texas fever but smaller and more difficult to stain. CHOREA. The coincidence of St. Vitus dance and chlorosis was such a fre- quent clinical picture that the older writers regarded the anemia as one of the essential causes of such nervous manifestations. Although this view has been proven untenable it remains true that anemia and chorea have many common antecedents, that the blood in chorea usually shows a slight impoverishment, and that anemia is a very fre- quent predisposing cause of chorea. Burr found moderate chlorotic anemia in all of 36 cases, but any severe grade was always referable to complications. Leroux found 5 million cells in 2 cases, between 3.4 and 4.8 millions in 5, and 2.2 mil- lions red, 89,000 white in an infant in which the disease was compli- cated by furunculosis. Zappert's four cases showed moderate reduction in red cells, 3.9-4.5 millions. Litten has recorded two fatal cases arising in the course of pernicious anemia. Cabot refers to 12 cases showing no abnormality except eosinophilia. The leucocytes in uncomplicated cases, have been found normal in numbers but often with excess of eosins. Zappert counted 630-1360 (8-19 percent) of eosins in 4 cases. The bacteriological examination of the blood during life has thus far been negative, although Leredde obtained the Staphylococcus albus from the blood of one case complicated by endocarditis. The blood of the cadaver has been found to contain various micro-organisms, Triboulet having 3 positive results in 15 cases. Functional nervous diseases, such as hysteria, neurasthenia, hypochondriasis, and tetany, do not give rise to anemia, but numer- ous associated conditions may lead to impoverishment of the blood. Cabot and Reinert report mild grades of chlorotic anemia in hysteria and neurasthenia. In neuritis Cabot reports chlorotic anemia with 336 NERVOUS AND MENTAL DISEASES. marked leucocytosis (16,000-28,000) in a febrile case in a young sub- ject, while in 4 of 6 cases of afebrile alcoholic neuritis the leucocytes were slightly increased. GRAVES' DISEASE. Anemia does not appear to be essentially connected with exophthal- mic goitre, as some cases show nearly normal cells and Hb. (Oppen- heimer.) Bramwell reports anemia, grade not stated, in only 27 of 64 cases. Yet many patients suffer from a form of anemia closely re- sembling chlorosis, which is sometimes quite severe. Thus Zappert records two cases with 2.8 and 2.7 million cells, and 32-30 percent of Hb. The writer has examined the blood in several such cases, in one of which the majority of red cells were distinctly undersized. The term " thyroid chlorosis " is applied by Capitan to cases of chlo- rosis associated with enlargement of the thyroid. Referring to Hayem's report of 29 goitres among 35 chlorotic patients, Capitan finds reason to believe that one type of chlorosis is referable to thyroid intoxication. Some years ago the writer was struck by the peculiar clinical type of chlorotic anemia observed in two elderly women suffering from Graves' disease, but has not since encountered such cases. The leucocytes, in the absence of complications, are normal or di- minished. Neusser and Cabot have noted relative lymphocytosis. Bibliography. Nervous and Mental Diseases. Burr. Cited by Cabot. Capitan. Compt. Rend. Soc. Biol., 1897, p. 1073. Capps. Amer. Jour. Med. Sci., Vol. Ill, p. 650. Daubler. Archiv f. Tropenhvg., 1897, p. 372. Fajardo. Cent. f. Bact., Bd."24, p. 558. Howard. Bull. N. Y. State Hosp., I., 1896, p. 139. Jellife. Bull. N. Y. State Hosp., 1897, p. 397. Kuhlmann. Bull. N. Y. State Hosp., 1897, p. 77. Langdon, Bamford. Bull. N. Y. State Hosp., 1896, p. 239. Leredde. Rev. Mens. Mai. de I'Enfanee, 1891. Leroux. Mai. de I'Enfanee (Grancher), IV., p. 819. Lewis. Text-book of Mental Diseases, 1889, p. 287. Litten. Charit^-Annalen, 1884, p. 265. McPhail. Jour, of Mental Sci., Vol. 30, pp. 378, 488. Oppenheimer. Deut. med. Woeh., 1889, p. 882. Pease. Bull. N. Y. State Hosp., 1896, p. 133. Roncorini. Archiv di Psychiat., 1894, p. 293. Seppili. Ref. in Jour. Mental Sci., Vol. 34, p. 143. Smyth. Jour, of Insanity, 1890, p. 505. Somers. Bull. N. Y. State Hosp. , 1896, p. 75. Spencer. Lancet, 1897, I., p. 32. Steele. Amer. Jour, of Insanity, Vol. 49, p. 604. Sutherland. Jour, of Mental Sci., Vol. 31, p. 147. • Triboulet. These de Paris, 1893, cited by Leroux. Winckler. Diss. Bonn, 1891. PART V. GENERAL DISEASES OF VISCERA. CHAPTER XX. THE HEMOPOIETIC SYSTEM. THE LIVER. Relation of Its Functions to the Blood. — Although the chief func- tions of the liver in the secretion of bile and the elaboration of ab- sorbed food products are directly concerned with the maintenance of the blood albumens, specific effects of inhibition of these functions have not yet been traced in the blood. These effects are doubtless seen in the anemia of advanced cirrhosis but are even then not to be dis- tinguished from similar changes in other secondary anemias. The function of transforming into bilirubin and removing from the system the pigment of disintegrated red cells is being constantly ex- erted in health and disease. When from any cause the excretion of this product is obstructed, as in acute yellow atrophy, or the demands upon the function are excessive, as in pernicious malaria, either hemo- globin, or its derivative bilirubin, accumulate in the blood giving hem- oglobinemia or jaundice. In the former case the content of the liver in hematoidin and hemisiderin is an accurate index of the grade of blood destruction. There is as yet no proof that the liver plays more than a passive part in the destruction of blood. The embryonal function of the liver as a depot of forming red cells is not infrequently retained till after birth, and there is reason to be- lieve that the renewal of this function may sometimes give peculiar character to the anemia of infants described by v. Jaksch. Effects of Bile upon the Blood. — A globulicidal action of bile ■was first demonstrated experimentally by Hunefeld (1840), while the fact that the biliary acids are the active agents in dissolving red cells in jaundice was show^n by v. Dusch (1854), and fully established by many later studies, v. Dusch not only demonstrated the power of bile to dissolve red cells, but claimed that it may also dissolve leuco- cytes and even liver cells, a property noted also by Rywosch. The relative globulicidal power of the various salts of biliary acids was ully tested by Rywosch, who found that sodium chenocholate and taurocholate dissolve red cells when in a concentration of 1-700, or 1-600, that Na-cholate is only moderately active at 1-200, while Na- 22 338 THE HEMOPOIETIC SYSTEM. glycocholate destroys red cells only when in a solution of 1-50. In the process of solution the red cells first lose their central depression, become variously deformed and finally dissolved, leaving no detritus. These changes have been observed only in the test-tube, but in the blood of most cases of jaundice the process appears to differ from that of other forms of hematocytolysis. The resistance of the red cells was found by Limbeck, in all cases of jaundice, to be distinctly increased, the cells not dissolving in solutions of more than .32 percent NaCl (normal .46 percent). This change he referred to the destruction of less resistant cells and a chemical union of Hb and stroma brought about by the action of bile salts. An increase in specific gravity of the blood was observed by Grawitz, following the injection of bile in animals, and referred by him to transudations excited by some specific action on the part of the bile. Limbeck also found a relative decrease in volume of the serum in obstructive jaundice, as well as a lower percentage of chlorides. At the same time the volume of the red cells was increased. Coagulation is hastened, according to Rywosch, by the presence of one part of Na-chenocholate or taurocholate in 500 of blood, but entirely inhibited when the proportion reaches 1-250. The presence of bile pigments has no deleterious action on the blood, although Rywosch found that injections of pure bilirubin, while not affecting the blood, produce mild general toxic symptoms. Diseases of the Liver. Jaundice. — In the human subject clinical observations have shown that small traces of bile are demonstrable by the greenish-yellow color of the serum of very mild cases in which, however, there are usually no other changes in the blood. In severe cases, when con- siderable quantities of bile are present in the circulation, many of the effects observed in the experimental study of jaundice become appar- ent. Any tendency toward concentration which the presence of bile may exert is seldom seen clinically, as the majority of severe cases of jaundice exhibit some reduction of red cells. This anemia is usually referable to preexisting conditions, but in very severe cases actual destruction of red cells may be traced in the dried specimen. A tendency to very rapid crenation, and total absence of rouleaux, have been observed by Grawitz. In the blood of icteric infants Hofmeier also noted deficiency in the formation of rouleaux, which, however, are normally less numerous than in the blood of adults. Evidences of solution of red cells were found by Silberman' in the blood of many jaundiced infants. These consisted in the presence in the blood of fragments of red cells, and, in the liver and spleen, of many phagocytes inclosing red cells and blood pigment. Many other degenerative changes in the red cells were noted but they were not necessarily connected with the jaundice. In severe cases examined by the writer, besides the ordinary signs of anemia, there were in the CANCER OF LIVER. 339 blood many very pale and some fragmented red cells, and in the viscera deposits of pigment equal to those of pernicious malaria. After death reddish acicular crystals (bilirubin ?) were sometimes found in the blood. A POLYCYTHEMIA referable to the concentratino- effect of bile in the blood has not often been reported. Yet Becquerel and Rodier reported the highest figures in their experience as occurring in a case of jaun- dice, and later observers have abundantly shown that well-marked jaundice markedly increases the specific gravity of the blood, in pro- portion to the intensity of the jaundice. Grawitz saw the gravity rise from 1.050 to 1.061 in a severe case, but does not report the number of red cells, while Limbeck and v. Noorden ^ found the dry residue of the blood to be 22-25 percent in several cases. The gravity of the serum was unaffected by the presence of bile in Hammarschlag's 1 2 cases, but in Limbeck's 2 cases the serum was poor in chlorides. The results of Limbeck's chemical analyses indicate that the volume of red cells in jaundice is much increased. The ALKALINITY of the blood is in mild cases unchanged (Limbeck), but in acute yellow atrophy, de Renzi, using Reale's method, found a neutral reaction, while an acid reaction was present in severe catarrhal jaundice and in suppurative hepatitis with jaundice. The leucocytes, in simple catarrhal jaundice, have usually been found imaifected, but in Grawitz' experience an increase has been common and occasionally the leucocytosis has been very marked (38,000-40,- 000). It is probable that different observers refer to different types of cases. Cancer of Liver. — A large proportion of the carcinomata of the liver being primary in the stomach, the anemia of this condition is fre- quently grafted upon that of gastric cancer. The reports of Wlajew include one case of secondary pernicious anemia with only 850,000 red cells, and none showing over 3 millions, while the leucocytes ran be- tween 20,000 and 40,000, indicating that this anemia is usually severe. The writei''s experience at autopsies in these cases accords in general with Wlajew's results, as marked emaciation and distinct oligemia are usually prominent. The sudden onset of severe jaundice in many of the cases suggests the probability that the blood may at times suffer concentration. Cabot's series of 17 cases showed well-marked anemia, cells under 3 millions in only two, while in 10 there were over 4 millions. The LEUCOCYTES are usually increased in cancer of tlie liver but the increase is often intermittent. Of 19 cases Cabot found only 6 with white cells under 10,000, and in none of these six was the ex- amination repeated. Alexander found leucocytosis in two cases, but only after repeated examinations. The grade of leucocytosis is usually moderate. The fact that when attacking the liver most carcinomata are already generalizing, the tendency toward rapid growth in the organ, and the early formation of further metastases, are some of the factors which account for the frequency of leucocytosis in hepatic cancer. 340 THE HEMOPOIETIC SYSTEM. Hypertrophic Cirrhosis with Jaundice. — More severe alterations of the blood have been observed by Hayem in a case of hypertrophic cirrhosis in which the red cells fell below 2 millions, with increased Hb index and leucocytosis. Severe anemia was also present in two cases verified at autopsy and reported by Cabot, but in one, only 2,400 leu- cocytes were found. Hanot and Meunier have insisted upon the presence of leucocytosis in this form of cirrhosis of liver, finding therein evidence of the infec- tious nature of the disease. They found from 9,000-21,800 leucocytes in 5 cases. Lukachewitch also found 20,000 white cells in another case, and Auche reports in three cases intermittent leucocytosis, 12,400 to 18,600 cells. Bacteria, usually Staphi/lococciis pyogenes aureus, have been found in the blood of " infectious jaundice," sometimes with hy- pertrophic cirrhosis, by Netter. Cholelithiasis. — Considerable interest attaches to the bacteriological examination of the blood in cholelithiasis from the recent discovery of bacteria in the blood, in several cases, usually in those marked by intermittent fever. Staphylococcus j^yogenes aureus has been isolated in two cases by Netter, in three cases, once with the colon bacillus, by Sittmann, and in the pus of metastatic abscesses by Gilbert and Girode, Brieger, and many others. The pneumococcus of Franhel and Strepto- coccus pyogenes were obtained from the blood in one case each, both complicated by hepatic abscess, by Canon and by Zancarol, while a small bacillus, not identified, was obtained from a complicating hepatic abscess and from the ulcerated heart valves, by Netter and Martha. The anemia in these cases varies with the previous condition of the patient and with a great variety of associated lesions. The behavior of the leucocytes is of interest, and may be of value in differential diagnosis. Biliary colic may supervene in severe attacks without producing any leucocytosis. The onset of jaundice is some- times accompanied by moderate leucocytosis, at other times fails to cause any increase of white cells. The difference probably depends upon the grade of inflammation excited by the gall stones, etc. Here as elsewhere one may expect no leucocytosis with simple catarrhal in- flammations but with active suppurative lesions leucocytosis may reach a pronounced degree (up to 50,000, Cabot). As some of these lesions run into abscess of the liver, or pyemia, even this high figure is prob- ably not the limit of the leucocytosis. Abscess of Liver. — The lesion in abscess of the liver may follow one of several types. 1. There may be a large number of small or miliary abscesses throughout the liver, filled with creamy or muco-pus. These abscesses arise from extension of suppurative cholangitis. The leucocytosis is probably a continuation of that observed in cholangitis, but there ap- pear to be no direct observations on the blood of this condition. 2. There may be one or two large cavities filled with creamy pus or stringy muco-pus, and well shut off from the liver tissue by mucous or pyogenic membrane. The writer has seen such an abscess run its I I CIRRHOSIS OF LIVER. 341 course with little or no increase of white cells in the blood, and the leucocytosis is usually slight. In one adult case, the day before J liter of muco-pus was evacuated at operation, with temperature 101° F., the blood showed 11,000 leucocytes, 52 percent of which were mononuclear, 48 percent polynuclear. A previous examination had also failed to show distinct leucocytosis. 3. Large necrotic abscesses with actively- suppurating walls are practically always attended with marked leucocytosis, up to 50,000, but this increase may be distinctly intermittent. Acute Yellow Atrophy. — Cases recorded by Grawitz and Cabot showed 5.15 million, and 5.52 million red cells, and 12,000-16,000 leucocytes. The exact limits of this disease are as yet imperfectly de- termined, but in three cases seen by the writer there was in each a moderate leucocytosis, 15,000-21,000. Two of these occurred in par- turient women at Sloane Maternity Hospital and showed the usual lesions in the liver. In another, at Roosevelt Hospital, the disease lasted three weeks, and the inner four-fifths of the hepatic lobules were necrotic. The red cells were not counted, and the absence of rouleaux and the early crenation mentioned by Grawitz were not noted. In two cases bacteriological examinations of the blood during life, by Sitt- mann, were negative, but Vincent obtained the colon bacillus from the blood of one case during life. Echinococcus Cyst of Liver. — Hayem classes the echinococcus cyst of the liver as one of the conditions leading to leucocytosis and excess of fibrin in the blood. Neusser refers to an increase of eosins in the same condition. Auche in one case ruled out the diagnosis of hypertrophic cirrhosis from the absence of an excess of leucocytes, the patient showing an echinococcus cyst at autopsy, while Wlajew found no change in the blood in a case of hydatid cyst of the liver. Cirrhosis of Liver. — In the early stages of the disease, when digestion is not much impaired and there are neither jaundice nor hem- orrhages, the blood is but little altered. Almost invariably the Hb is deficient. During the progress of the disease the blood steadily de- teriorates, giving some of the most typical examples of secondary anemia. Abundant cause of the anemia is found in all cases in the disturbance of the functions of the liver and stomach, and in other cases by the complications of the particular type which the disease follows. In slowly progressing cases without special complications, the blood commonly shows between 3-4 million cells, 55—75 percent of Hb, and little or no leucocytosis. The ordinary progress of the anemia is perhaps most affected by PROFUSE HEMORRHAGES, after which the blood shows the usual reduc- tion of red cells and Hb, with nucleated red cells and leucocytosis. Single hemorrhages may occur in subjects which are only slightly anemic, and the blood is then rapidly restored. In the later stages of the disease, single or repeated hemorrhages if not at once fatal, often ead to very grave secondary pernicious anemia. Ascites, although draining the blood of much albumen, is com- 342 THE HEMOPOIETIC SYSTEM. monly found with lesser grades of chlorotic anemia. It is not un- common to find, in well advanced cases requiring paracentesis, 4| to 5 million red cells and Hb above 65 percent. Grawitz believes that the ascitic transudate leads to some concentration of the blood and the masking of anemia. Such a condition is strongly suggested at times by the disproportion between the anemic appearance of the patient and the slight changes demonstrable in the blood. The effects of tap- ping vary. Limbeck observed a case of Laennec's cirrhosis, with ex- treme ascites, in which the removal of 18 liters of fluid raised the red cells after 24 hours from 3.28 millions to 5.16 millions. Three days later they fell to 3.54 millions. Osterspey noted the same result after several tappings of the same patient. On the other hand Grawitz saw the red cells fall from 4.7 to 4.3 millions shortly after tapping, but the first examination came three days before the operation. The eifects of tapping may very well vary with the immediate effects upon the circulation, with the state of the kidneys, and with the reappearance of the ascites. Jaundice, if mild, has no demonstrable influence on the blood of cirrhosis ; nor do the ordinary chronic forms of moderately severe jaundice appear to be accompanied by any special alteration in the blood except a tendency toward leucocytosis. In most of the jaundiced cases reported by Hayem, Cabot, and others, there was a moderate leucocytosis, reaching 20,000 or more. Severe jaundice usually occurs in advanced and anemic cases, or in the hyper- trophic forms. The ordinary cirrhotic process in the liver is unaccom- panied by leucocytosis, which however may frequently arise from many complications or accidental causes. Thus Hayem, Osterspey, Cabot, and others found no leucocytosis except from complications, but Gra- witz, Rosenstein, and Wlajew refer to cases with moderate leucocytosis, without giving details. DISEASES OF GASTROINTESTINAL TRACT. Esophagus. — Stenosis of the esophagus, when extreme, by diminish- ing the quantity of fluids ingested, has been shown to lead to marked concentration of the blood. V. Noorden ^ refers to two cases of can- cer of esophagus in which the dry residue of the blood reached 26.5 and 27.3 percent (normal 21-22 percent). When to this condition of con- centration are added the cachexia which belongs to carcinoma and to starvation, a marked degree of oligemia may result, approaching that of pernicious anemia. A distinctly reduced quantity of blood was noted at autopsy in one of two cases of epithelioma of esophagus, observed by the writer at Roosevelt Hospital, while in the other, in which death resulted from pulmonary extension and pneumonia, the stenosis was not extreme. These same factors may perhaps be re- sponsible for some relative increase in the numbers of red cells which has been observed in a few cases. The red cells in reported cases have been above normal, or very slightly reduced, as in other forms DISEASES OF THE STOMACH. 343 of cancer, or greatly reduced, as in the severe anemia of cancerous cachexia. Comparing his own results in a single case with those of Escherich, Pee, and Osterspey, Rieder concluded that the cachexia of epithelioma of the esophagus differs from that of carcinoma in other regions in the absence of leucocytosis. Reinbach also reported two cases with relative lymphocytosis. Cabot and Hoffman, however, report several cases with moderate leucocytosis. In the writer's cases there was ante-mortem increase of white cells. There has been no attempt to refer these varying results to the special causes of leucocytosis to which they are probably due. Diseases of Stomach. General Considerations. — Since the blood depends for its supply of albumens chiefly upon the functional activity of the stomach, there is an intimate relation between the condition of the blood and the ac- tivity of digestion and absorption in this organ. 1. Transitory functional disturbances of the stomach usually fail to notably alter the blood, but persistent vomiting with hyperacidity has resulted in distinct concentration of the blood and reduction of chlorides both in blood and urine. 2. While total withdrawal of food for a short period causes com- paratively little change in the blood, a long-continued failure of diges- tion or absorption is one of the most efficient causes of anemia. It has already been shown that organic lesions of the stomach are promi- nent among the pathological findings in cases of pernicious anemia, including chronic gastritis, atrophy of gastric mucosa, stenosis of the pylorus, and cancer. All the lesser grades and types of anemia are ob- served, with even greater frequency, to be associated with gastric lesions, while the anemia of marasmic infants is doubtless largely owing to fail- ure of digestion and absorption, without organic lesion in the stomach. Meynert's theory of chlorosis further illustrates the well-founded be- lief in the importance of normal digestion in maintaining the condi- tion of the blood. 3. The stomach is a frequent site of loss of blood by hemorrhage. If, as is indicated by the studies of Quincke, and of Dettwyler and Sil- berman,^ anemia or hemoglobinemia are strong predisposing causes of ulceration of the stomach and duodenum, there is in the interaction of these two conditions a " circulus vitiosus " which readily explains their frequent association in the same subject, and the very severe grades of anemia which thereby result. Smaller and more frequent bleedings also greatly aggravate the anemia which accompanies ulcerating carcinoma. 4. Much has been written about the effects of gastro-intestinal toxemia, but the chief source of autotoxic agents is without doubt the intestine. Special Diseases. Acute Gastritis, Dyspepsia, Hyperacidity. — These conditions, if not prolonged, have no tendency to produce anemia. If vomiting is 344 THE HEMOPOIETIC SYSTEM. excessive the cells and Hb may be increased. The leucocytes are sometimes increased, sometimes normal or diminished. Their numbers may indicate the intensity, but not the etiology of the inflammation. Chronic Gastritis. — Chronic gastritis may be tolerated for long periods without leading to any marked anemia, as the writer has often observed in private cases addicted to alcohol. When the patients be- come disabled, anemia, of the secondary chlorotic type, is usually found to be associated with loss of flesh. The leucocytes are low, and lym- phocytes usually prevail. (Blindeman, Cabot, Hofiman.) Digestion leucocytosis may be diminished and is sometimes absent, as in cancer. Associated with chronic gastritis and fibrous stricture of the pylorus, the Avriter has twice observed slowly progressive anemia lasting four and five years. The changes in the blood were first chlorotic, later intermediate, and finally those of pernicious anemia, without many megaloblasts in the blood, but with typical changes in the marrow. Much more acute and typical cases of pernicious anemia have been re- ferred to chronic gastritis. (Pepper, Stengel.) Dilatation of Stomach. — In cases of gastrectasis from chronic gas- tritis or stenosis of pylorus the blood does not differ from that of simple uncomplicated chronic gastritis. Although the patients become cachectic and lose flesh, the blood usually fails to show corresponding anemia. In the cases reported by Reinert, Cabot, Blindeman, and others, the red cells and Hb were normal or increased or but slightly reduced. It is generally agreed that oligemia exists in these cases, but its origin is not clear, and it is by no means evident that the con- centration of the blood is affected, as Kussmaul suggests, by excessive vomiting. Defective absorption is j)robal)ly a more important factor. The Blood in Gastro-intestinal Diseases of Infants. — Acute gas- tro-enteritis occurring in healthy infants usually tends to concentrate the blood, but these effects are demonstrable only in severe cases. Fel- senthal and Bernhard, in 20 cases, found uniform polycythemia reach- ing 7.5 million cells, and usually moderate leucocytosis. Hock and Schlesinger found a moderate increase in specific gravity in very se- vere but not in ordinary cases. On the other hand in an anemic rachitic infant with diarrhea the Hb fell in one week from 80 to 60 percent, indicating that a preexisting anemia may be aggravated by such attacks. The leucocytes are usually increased, Fischer finding a well-marked lymphocytosis with 60 percent of lymphocytes, while Hock and Schlesinger in the above-mentioned case found 38,000 leucocytes the majority of which were large lymphocytes. In hoo cases of cholera infantum, reported by Weiss, the lymphocytosis approached the grade of lymphatic leukemia. In chronic gastro-enteritis there may be progressive anemia and emaciation, as in Hayem's case in which after 2 months the red cells numbered only 685,000, leucocytes 18,910. On the other hand, Monti and Berggrun (p. 17) report a case lasting 6 weeks in which the blood of the emaciated subject showed 4.1 million red cells and 65 ULCER OF STOMACH. 345 percent of Hb. In most of Felsentbal's cases the Hb rose above 65 percent, only when there was marked polycythemia, while moderate leucocytosis was the rule. The grade of anemia observed in marasmio wfants with mild gastro- intestinal disturbance is very variable, and its different phases illus- trate all the stages of secondary anemia. It is usually characterized by relative or absolute lymphocytosis. Ulcer of Stomach. — The state of the blood in ulcer of the stomach is extremely variable and it is not always evident what conclusions to draw from the results of its examination. From the very numerous cases reported in greater or less detail by Leichtenstern, Laache, Haeberlin, Oppenheimer, F. Muller, Schneider, Reinert, Osterspey, Cabot, Rochemont, and others, as well as from the writer's own obser- vation, these changes seem to fall in some rather distinct classes. 1. Some cases appear to have suffered no permanent change in the composition of the blood, but from their anemic appearance it is prob- able that the total volume of blood has been reduced, without marked alteration in its quality. Most of these patients have never sustained a large hemorrhage, or a former loss of blood has been replaced through the active regeneration which is often seen in this condition. Muller and Oppenheimer have re- ferred especially to such cases, while Grawitz demonstrated in one patient that the blood was decidedly hydremic, the dry residue of the serum being only 8.56 percent, while that of the whole blood was 19.75 percent, and the red cells numbered 4.34 millions. This patient was recovering from more severe anemia and still gave anemic heart murmurs. A single chemical analysis, while most important, cannot with certainty be applied to all these patients, however, and several of Oppenheimer' s reports indicate that the blood may be entirely normal in some cases of gastric ulcer. It does not seem probable that vomiting or other causes of concentration of the blood are important factors in the majority of these cases. 2. The usual state of the blood is one of marked secondary chlorotic anemia, with little or no leucocytosis, the presence of which depends largely upon the manner of feeding. While the regeneration of the blood after hemorrhage is often very rapid, it is seldom complete while the patient remains in the hospital, and there seems to be some influ- ence, possibly found in a diminished digestive power in the stomach, which causes the anemia to persist. The writer followed one male patient for 5 years after a dangerous hem- orrhage, but although all local and other general symptoms had disap- peared, the blood never showed more than 75 percent of Hb. 3. The bleedings of ulcer of the stomach furnish some of the most remarkable examples of post-hemorrhagic anemia, the characters of which have been described. Surviving patients seldom show less than one million cells, but from this point may improve rapidly, with low Hb-index, normoblasts, and leucocytosis. ' Cases with cells under 3 millions have usually suffered from hemorrhages, which if frequently repeated may lead to grave anemia with rising Hb-index and increas- ing diameters of red cells. 346 THE HEMOPOIETIC SYSTEM. Leucocytes. — Leucocytosis of moderate grade follows hemorrhage, or local inflammation. In one case with marked leucocytosis the writer found two perforated ulcers, one opening into a large cavity in the liver, the other connected with an abscess of the liver. In quiescent periods, with rectal feeding, there is usually hypoleucocytosis, while the change to feeding by the stomach may excite considerable diges- tion leucocytosis, as in Cabot's case, in which the first meal raised the white cells from 4,000 to 15,000. These features may be of diagnos- tic value when cancer is suspected, but should be interpreted with cau- tion. Digestion leucocytosis has been absent in 10 cases of ulcer re- ported by Hoffman, Hassmann, and Schneyer. In three of these there was stenosis of pylorus. Cancer of Stomach. — The usual course of the anemia of malignant neoplasms is varied by the peculiar conditions connected with a tumor of this organ, and by the extent and character of the new growth. In the reports of some 200 cases to be found in the literature there has been no systematic attempt to classify the types of anemia with reference to the character of the tumor, and while it appears by no means certain that such a classification would be of clinical value, there are, nevertheless, certain features of the growth and complica- tions of these tumors which chiefly determine the state of the blood. 1. 8ome cancers of the stomach exist for considerable periods without leading to marked impoverishment of the blood, and there is no necessary <5ause of anemia in the early stages of most of those which do not bleed, ulcerate, or contract the pylorus. When chronic gastritis or ulcer precede cancer, the preexisting anemia suffers at the time no appreciable change. Accordingly, it is a common experience to find in the blood at this time no distinct indication of the presence of a ma- lignant tumor in the body. At this time the general symptoms are more reliable diagnostic signs than is the state of the blood. Moreover, if vomiting is a prominent early symptom, an initial anemia may be masked by temporary concentration of the blood. Thus Cabot re- ports over 4 million red cells in 34, and over 5 millions in 19, out of '72 cases. The first sign of affection of the blood is seen in the falling Hb which has been found considerably reduced in many of the cases with normal or nearly normal red cells. 2. In the majority of case^ of gastric cancer the red cells are reduced in number, the Hb is deficient, and the blood shoios the changes of second- ary chlorotic anemia. The reduction of cells is sometimes remarkably slight, but the Hb usually suffers first and most severely. When dis- tinct cachexia is present, both red cells and Hb are nearly always re- duced, unless there is active vomiting. Poikilocytosis appears in the more chronic and severe cases. Nucleated red cells are usually seen in small numbers. Sometimes they are present when the anemia is very slight. Jez found them so common in cancer and so rare in ulcer as to furnish diagnostic evidence between these conditions. The Hb-index is commonly subnormal but not so low as in chloro- CANCER OF STOMACH. 347 sis. Osier and McCrae, however, found an index of .63 in 52 cases, which is nearly as low as that of chlorosis. The Hb is very much less likely to increase when once low, while it does increase, frequently and rapidly, in simple ulcer. (Blindeman.) The conditions which tend to develop severe anemia in gastric can- cer are chiefly hemorrhage, ulceration, and metastasis. Diffuse growths and annular pyloric tumors have less effect upon the blood. 3. A moderate 2yroportion of cancers of the stomach pursue a course which clearly resembles that of jiernicious anemia. Without very marked gastric symptoms and with no demonstrable tumor, the blood becomes excessively anemic, the red cells falling to a low figure, usually about 1^-1 millions, megalocytes are abundant, and nucleated red cells, some of which may be of large size, make their appearance. The type of carcinoma €ound in tliese cases varies. Sometimes there is a small tumor of the stomach wall with moderate neighboring metastasis, suggesting the development of cancer in an old round ulcer. Sailer and Sadler report such cases, and the writer has seen two others at autopsy. Usually the tumor is of considerable size and ulcerating, as in another case reported by the writer. It is not probable, however, that the grave form of anemia is necessarily associated with any particular forms of cancer, as both the above types may be seen without extreme anemia. In the diagnosis of this type of gastric cancer from pernicious anemia, one is usually aided in the examination of the blood by the following features : (a) The majority of megalocytes do not show an excess of Hb, which is distinctly more deficient than in pernicious anemia. (6) Megaloblasts are rare, and nucleated red cells, if present, are mostly of normal size or but slightly larger than normal. (c) There is usually polynuclear leucocytosis in advanced carcinoma of stomach. The leucocytosis may fail and a few megaloblasts may be present, but the writer has been unable to find a case of grave anemia in cancer of the stomach in which the general deficiency of Hb was not marked. Henry calls attention to the fact that in cancer the cachexia exceeds the oligocythemia, while in pernicious anemia the opposite relation holds, but the anemic type of gastric cancer has in several reported cases reduced the red cells below IJ millions, the limit noted in his experience. In his statement that at death pernicious anemia leaves fewer red cells (usually below 1 million) than gastric cancer (usually above 1 million) most observers will concur. Leucocytosis in G-astric Cancer. — Polynuclear leucocytosis is present in the majority of cases but not in all. In the early stages of the dis- ease, when diagnoses are difficult, leucocytosis is frequently absent, and if we exclude the slight increases, the majority of early cases fail to show leucocytosis. This fact probably depends on the usual absence of any cause of leucocytosis, which the tumor itself does not furnish. Without much regard to the position of the tumor or the rapidity of growth, leucocytosis makes its appearance as a result of ulceration, 348 THE HEMOPOIETIC SYSTEM. exudative inflammation about the tumor, hemorrhage, metastasis, and anemia. In the writer's experience these causes are most active in the order named. A single examination of the blood is, however, quite insufficient to exclude the occurrence of the leucocytosis which is now under consideration. The leucocytosis of secondary anemia is often intermittent, according to the process which it follows, and its detection may require repeated examinations. The grade of leucocytosis is usually not high, but since in low states of nutrition the leucocytes are usually low, the presence of 10,000 leucocytes has here more significance than in states of health. The majority of advanced cases show between 10,000 and 20,000 white cells. Extreme leucocytosis has usually been found to result from complicating inflammation, extensive ulceration, or numerous meta- stases. It is said that some cases may pursue their course entirely without leucocytosis. When the total numbers are distinctly increased the proportion of polynuclear leucocytes is usually high. This fact is specially apparent in the cases reported by Osier and McCrae. Eosinophile cells, how- ever, are nearly always present, and sometimes their numbers and pro- portions are considerably increased (5-6 percent). It is of interest to note that Labbe has found many eosinophile cells in the tissue in and about an excised cancer of the stomach, as well as in the adjoining lymph nodes. Digestion Leucocytosis in Cancer of Stomach. — R. Muller, in 1890, first called attention to absence of digestion leucocytosis in cancer of the stomach, and this fact has since been abundantly verified by the reports of Schneyer, Hoffman, Hartung, Capps, Sailer, Cabot, Hassman, and Jez Muller, Schneyer, Jez, and Hartung found no digestion leucocytosis in their 41 cases, while in simple ulcer it was always present, except in one case of hematemesis. Hartung found it present in four cases of cancer of other organs. Capps and Cabot report digestion leucocytosis of 3,270-3,850 cells, in 3 of 37 cases, while it was present in only 5 of 10 cases of chronic gastritis and in one case of fibrous stricture of pylorus. More recently Hoffman , finding digestion leucocytosis in 3 of 24 cases of cancer, but in only 2 of 9 cases of ulcer of stomach, and in only 2 of 11 cases of anacidity from other causes, denies any great diagnostic value in the test. Hassman and Schneyer also found no digestion leucocytosis in three cases of ulcer with pyloric stenosis, while Sailer found no digestion leucocytosis in one case of tuberculous ca- chexia, and in two cases Avith preexisting inflammatory leucocytosis. Osier and McCrae demonstrated an increase of 33 percent of leucocytes following digestion in 10 of 22 cases of gastric cancer. The above studies showing that digestion leucocytosis is absent when the stomach fails either to digest or absorb with considerable activity, conditions which usually but not necessarily accompany carcinoma but which may also be found in simple ulcer, chronic gastritis, stenosis of pylorus, gastroptosis, anacidity, tuberculosis, etc., it appears that this test can hold only a very subordinate position in the diagnosis of can- cer of the stomach. DISEASES OF THE INTESTINE. 349 Duodenal Ulcer closely resembles gastric ulcer in its effects upon the blood. Diseases of the Intestine. The blood is very closely dependent upon the absorptive and ex- cretory functions of the intestine both for its transient variations in con- centration and its more permanent supply of albumens. So marked and prompt is the response of the blood to these changing conditions, that no examination of the blood requiring accurate calculation should be made without full regard to the condition of the gastro-intestinal tract. The writer has repeatedly seen an attack of typhoid fever com- pletely obliterate the signs of a moderate chlorosis ; and has previously referred to the remarkable transformation of the blood observed when, at Montauk in 1898, typhoid fever developed in subjects of malarial cachexia. Of the two main intestinal functions the blood responds much more promptly to excessive excretion than to increased absorption. Effects of Absorption. — During the early stages of digestion of solid food the red cells tend to increase, owing to the discharge of a considerable bulk of digestive fluids. (Buntzen, Sorenson, Leichten- stern.) With largely fluid diet, on the other hand, the absorption of fluids may tend from the first to diminish the red cells, so that their lowest proportion is found about four hours after meals. At the same time coagulation is slower. (Vierordt, Leichtenstern.) The increased flow of lymph doubtless contributes to this result. With proper diet in healthy adults, these processes usually fail to cause important vari- ations in the red cells and Hb of the blood, and even when the quan- tity of fluid in the food is excessive or extremely deficient, the demonstra- ble effects are slight and very transient. Leichtenstern could detect no change in a patient who drank 21.5 liters of water during three days, and Schmaltz found only a slight reduction in gravity (1.059 to 1.057) forty-five minutes after the ingestion of 4 liters of physiological salt solution. The ingestion of very large quantities of fluid can dilute the blood only to a slight degree, beyond which the activity of the kidneys is sufficient to carry off all excess and keep the gravity of the blood within somewhat narrow limits. Depletion of the blood by limited ingestion of fluids and especially by watery exudates leads, on the other hand, to very prompt and consider- able concentration of the blood. Numerous clinical illustrations of this fact have already been considered as arising from disorders of the stomach, but the more extreme examples are connected with the use of purges and with diseases of the intestine, especially those attended with active diarrhea. Effect of Purges Upon the Blood. — It was held by Poisseuil in 1839 that when salines are administered in greater concentration than that of the blood serum, the laws of osmosis determine the passage of fluids from the blood into the intestine. It has since been shown that with many laxatives other factors lead to the same result, especially he increased intestinal secretion and the failure of absorption. 350 THE HEMOPOIETIC SYSTEM. Brouardel first demonstrated an increase of blood cells following purga- tion finding that with severe diarrhea the red cells might rise as much as 1 456 000 per cmm. The leucocytes were at the same time increased, maximum 5,880, but apparently for different reasons and not in proportion to the red cells. Later, Hay found an increase of nearly 2 million red cells within 1\ hours after the administration of a large dose of Glauber's salt. The increase began in a few minutes, and the normal condition was restored in about 4 hours. The effects were more marked the more concentrated the solution ; but after a certain concentration of blood had been reached, no further effects could be induced and no purgation followed the continued ad- ministration of the saline. Likewise Grawitz found the administration of 15 grains of Epsom salt in 50 cc. water to raise the gravity of the blood, be- ginning in 5 minutes and reaching a maximum increase, from 1.0591 to 1.0539 in 42 minutes, after which there was a steady decline. Common salt was even more active ; 15 grains in a little water on an empty stomach rais- ing the gravity of the blood from 1,050 to 1.060 within 20 minutes. The ef- fects of such salines he found were largely nullified when the stomach and intestines contained much food. Colitis (Dysentery). — Acute colitis of croupous or ulcerative type usually causes some concentration of the blood and is attended with considerable leucocytosis. After the acute febrile period, if the patient survives, anemia replaces the oligemia and the leucocytosis becomes intermittent. Yet the writer has seen excessively emaciated cases of acute amebic colitis die, without fever, and without marked reduction in the proportion of red cells. Chronic ulcerative colitis, when attended by much purulent exu- date and numerous hemorrhages may lead to extreme anemia of secondary chlorotic or pernicious type. Leucocytosis if present is intermittent. Catarrhal colitis has little effect upon the blood, tending by deple- tion to raise the proportion of red cells. There is usually no leuco- cytosis. Simple acute diarrhea causes a slight and transitory increase of cells and Hb. Cholera morbus and the severe gastro-enteritis of infants must have a pronounced effect in concentrating the blood and masking anemia, but no detailed studies in this field are apparently at hand. Cholera. — The most extreme grades of concentration of the blood are seen in those cases of Asiatic cholera w'hich die in the algid stage, and in which the system has been excessively depleted by prolonged diarrhea. The diminution in the quantity of blood may even be plainly visible to the naked eye, in the scanty hemorrhage which fol- lows section of an artery. (Dieffenbach.) The principal changes in the blood of cholera are fully set forth in the first systematic study of the subject by Schmidt, who in specimens obtained by venesection found the specific gravity of the whole blood raised from 1.050-1.059 to 1.065 or 1.073, of the serum from 1.029 to 1.047. Even the red cells were found to suffer a marked loss of water, their gravity rising from 1.088 to 1.102. Moreover, while the in- crease in density of the blood and serum was not uniform nor in all cases well marked, the changes in the red cells were pronounced and CHOLERA. 351 invariable. Schmidt's analyses were not made in the most extreme cases, however, and the relation between the gravities of the blood, the cells, and the serum, indicate that the relative quantity of serum had remained about normal. Other and older observers (Thomson) had previously noted a gravity of the serum as high as 1.057, indi- cating that the serum had begun to suffer disproportionately in bulk, and that a true oligoplasmia existed. Schmidt and Biernacki added to the knowledge of the subject the fact that all chemical constituents are increased in rather uniform pro- portion except the sodium chloride, which seems to transude in the stools in excessive quantity. Biernacki, finally, has shown that these excessive grades of depletion of the blood do not exist unless the tissues are themselves seriously drained of fluids and can no longer supply water to the blood. The coagulability was found by Hayem and Winter to be sometimes normal, but often, in the stage of reaction, much increased. They found a much diminished capacity of the blood to absorb oxygen. The alkalescence of the blood in cholera was regarded by the older observers as extremely deficient, and Cantani in 1884 claimed to have demonstrated a rapid decrease of alkalinity, a neutral reaction during the algid stage, and even an acid reaction before death. This change he referred to excess of COg, and he recommended treatment by alka- lies. Biernacki refers the diminished alkalinity to loss of sodium. Hayem and Winter examined twelve specimens of blood during the stage of collapse, finding four of them neutral or slightly alkaline, and eight slightly acid. The reaction has not been determined by more recent methods. The red cells are markedly increased in nearly all cases. In one instance they rose to 7.662 millions within 24 hours, the patient finally recovering. Usually the red cells rise to 6^-7J millions, and this ex- cess is usually, but not always, proportionate to the other evidences of concentration. One case showed a maximum of over 8 million red cells, while another gave only 3.193 millions during the stage of reac- tion. Beginning concentration ha^ been noted as early as the third hour. Biernacki attributes to the thickening of the blood and espe- cially to the depletion of tissues considerable importance in the patho- genesis of symptoms but found very severe symptoms in patients not showing much diarrhea. Morphological changes in the red cells have not been noted by most observers, but Nicati reports that in hemorrhagic and jaundiced cases there are many shadows of dissolving red cells, and often pigmented leucocytes, to be seen, especially in the stage of reaction. Leucocytosis is practically constant in all marked cases. Biernacki found 21,250 cells in one mild case suffering chiefly from diarrhea, but only 4,375 in another very similar case. Eight cases showing in the algid stage very high leucocytosis, 40,000-60,000, were fatal. Some cases with more moderate leucocytosis (20,000-30,000) were also fatal, but less rapidly. In the stage of reaction the leucocytes are 352 THE HEMOPOIETIC SYSTEM. usually lower, but some fatal cases show a persistent or increasing leu- cocytosis in this stage. The increase of leucocytes is relatively much greater than that of the red cells. It has been noted as early as the twelfth hour, and may persist as long as 6 days. The polynuclear cells are in great excess, and eosins are usually absent. Bacteriology of tlie Blood. — The comma bacillus has apparently not been found in the blood during life. After death it was recovered from the heart's blood in a small proportion of 48 cases examined by Lesage and Macaigne, and in 4 of 11 cases, examined 8-24 hours after death by Kekowsky. Wlaew failed to obtain positive results from the heart's blood in several cases, and one specimen of blood drawn during life was sterile and non-pathogenic to animals. Dia- troptoff also found 3 sterile specimens of heart's blood. Bibliography. Hemopoietic System. Alexander. These de Paris, 1887. Auche. Gaz. hebdom., 1897, p. 459. Becquerel, Rodier. Loc. cit., p. 103. Biernacki. Deut. med. Woch., 1895, p. 795. Blindeman. Wien. med. Blatt, 1895, p. 695. Brieger. Zeit. f. klin. Med., 1886. Brouardel. L' Union Med., 1876, T. 22, p. 405. Canon. Deut. med. Woch., 1893, p. 1038. Cantani. Cent. f. med. Wissen., 1884, p. 785. Capps. Boston Med. Surg. Jour., 1897, Nov. 4. Diatroptoff. Deut. med. Woch., 1894, p. 691. Dieffenhach. Cited by Lukjanow, Grundz. allg. Path. d. gefasssyst., Leipsic, 1894, p. 221. V. Duscli. Inaug. Diss. Leipsic, 1854. Escherich. Berl. klin. Woch., 1884, p. 145. Ewing. Trans. N. Y. Path. Soc, 1897-8, p. 136. Felsenthal, Bernhard. Archiv f. Kinderheilk., Bd. 17, p. 333. Fischl. Zeit. f. Heilk., Bd. 13, p. 277. Gilbert, Girode. Semaine Med., 1890, p. 481. Haherlin. Miinch. med. Woch., 1888, No. 22. Hanoi, Meunier. Compt. Rend. Soc. i?iol., 1895, p. 49. Hartung. Wien. klin. Woch., 1895, p. 697. Hassman. Wien. klin. Woch., 1896, p. 314. Hay. Jour, of Anat. and Physiol., 1882, p. 430. Hayem. Gaz. hebdom., 1889, p. 726. Hayem, Winter. Gaz. hebdom., 1885, pp. 118-138. Henry. Archiv f. Verdauungskrankh., 1898, p. 1. Hock, Schlesinqer. Beitriige zur Kinderheilk., 1892. Hoffman. Zeit. f. klin. Med., Bd. 33, p. 460. Hqffmeier. Die Gelbsucht d. Neugeborenen, Stuttgart, 1882. Hunefeld. Der Cham, tierisch. Organis., Leipsic, 1840. Cited by Eywosch. Jez. Wien. med. Woch., 1898, p. 633. Kussmaul. Berl. klin. Woch., 1872, No. 37. Laache. Die Anemic, 1883. Labbe. Bull. Soc. Anat. Paris, 1899, p. 85. Leichtenstem. Untersuch. u. d. Hb-gelialt. des Blut., 1878. Lesage, Macaigne. Annal. Institut Pasteur, 1893, "p. 17. Lukachewitch. Cited by Auche. Monti, Berggrun. Chron. Anemie d. Kindesalter, Leipsic, 1892. F. Muller. Verb. d. Vereins. inn. Med., 1888, p. 378. R. Muller. Prag. med. Woch., 1890, No. 17. Netter. Progres. Med. , 1886, p. 992. Netter, Martha. Archiv de Physiol., 1886, p. 7. BIBLIOGRAPHY. 353 Nicati. Marseilles Med., 1884, p. 684. V. Noorden. iPath. d. Stoffwechsels, p. 280. ^itid., p. 461. Oppenheimer. Deut. med. Woch., 1889, Nos. 42-44. Osier, McCrae. N. Y. Med. Jour., Vol. 71, p. 757. Osterspey. Inaug. Diss. Berlin, 1892. Pee. Diss. Berlin, 1890, p. 8. Poisseuil. Cited by Grawitz. Quincke, Dettwyler. Corresp. Schweiz. Aerzte, 1875, p. 101. Reinbach. Langenbeck's Archiv, Bd. 46, p. 486. Reinert. Zablung d. Blutkorp, 1891. de Renzi. Virchow's Archiv, Bd. 102, p. 218. Rochemont. Miinch. med. Woch., 1897, p. 1466. Rokowsky. Archiv des Sci. Biol., St. Petersb., 1892, p. 517. Rosenstein. Cong. f. inn. Med. , 1892, p. 74. Sadler. Fort. d. Med., 1892, Suppl. Sailer, Taylor. Internat. Med. Mag., 1897, July. Schmidt. Charact. d. epidem. Cholera, Leipsic, 1850. Schneider. Diss. Berlin, 1888. Schneyer. Zeit. f. klin. Med., Bd. 27, p. 476. Silberman. i Archiv f. Kinderheilk., 1887, p. 401. * Deut. med. Woch., 1886, No. 29. Sittmann. Deut. Arch. klin. Med., Bd. 53, p. 323. Thomson. Cited by Schmidt. Vincent. Semaine'Med., 1893, p. 228. Weiss. Jahrb. f. Kinderheilk., Bd. 35, p. 146. Wlaew. Cited, Baumgarten's Jahresb., 1894, p. 400. St. Petersb. med. Woch., 1894, Lit. beil., p. 43. Zancarol. Revue de Chirurgie, 1893, T. XIII. 23 CHAPTER XXI. LUNGS, HEART, KIDNEYS. DISEASES OF THE LUNGS. Respiratory Changes in the Blood. — During its passage through the lungs the blood undergoes certain changes which in their normal progress belong to the physiology of the blood, but when disturbed by pathological conditions may considerably aiFect both its chemistry and its morphology. The absorption of O^, which results partly by simple absorption by the plasma, but largely by chemical union with Hb, raises the volume of this gas in arterial blood to 21 percent as against 12 percent for venous blood, while the discharge of COg from the tissues into the plasma, where about two-thirds of it exist as a bicarbonate and one- third in the red cells, increases the CO2 content of venous blood to 46 volume percent from 38 percent, the average content of arterial blood. The character of the diet considerably influences the ratio of COg excretion to O2 absorption, which with carbohydrate food main- tains a proportion of about 1 to 1, while with highly albuminous diet more oxygen is absorbed and devoted to oxidation, and the ratio is .74 to 1. (Pembry.) Arterial blood was long supposed to be richer than venous in solids and Hb (Kruger), but Cohnstein and Zuntz showed this apparent thickening to be due to stasis caused by the procedures followed in venesection ; and with Kruger and others these observers demonstrated that the blood in larger arteries and veins is equally rich in red cells, Hb, and dry residue. An important diiference between venous and arterial blood was demonstrated by Hamburger, who showed that under the influence of CO2 the isotonic tension of the red cells is moderately increased, so that chlorides and water pass more readily from plasma into cells, while albumens, alkalies, and probably also phosphoric acid, pass from cells to plasma. A simultaneous increase of the sugar and fat of the serum was also noted, but its origin not fully determined. As a result of these alterations in osmotic characters, the red cells of venous blood were found to swell and assume a more spheroidal shape. Limbeck, who also verified this change in the osmotic qualities of the red cells in venous blood, referred the excess of dry residue of the plasma not to withdrawal of albumens from the cells but to absorption of the water of the plasma by the red cells, which he found to suffer an in- crease of volume in venous blood. DISEASES OF THE LUNGS. 355 With the increase of CO2 in venous blood those observers who iden- tify alkalinity with C02-content found an increased acid-neutralizing power in venous blood. This anomalous result, viz., the increase of alkalinity of a fluid by the addition of an acid, they referred to the withdrawal of alkaline salts from the tissues for the neutralization of the acid. It is of prime importance in considering the pathogenesis of symp- toms in various anemias to call attention to several studies which have demonstrated that Hb has a strong chemical affinity for oxygen and that this attraction, rather than the percentage of O2 in the air, chiefly regulates the supply of oxygen to the tissues. Lukjanow first showed that by increasing the pressure of the O^ in the lungs he could not add to the quantity of absorbed oxygen, although arterial blood is not sat- urated with oxygen, while Lowy found that neither by increasing the pressure to 1,400 mg. Hg, nor by lowering it to 40 mg. (normal 760 mg.), could the blood be made to absorb more or less O2. Beyond these points hovvever the absorption of O^ begins to vary. Turning their attention to the respiratory processes in anemia, Kraus, Chvostek, Bohland, and especially Biernacki, found that the quantity of oxygen that can be exhausted from blood remains about the same although the Hb may vary greatly, that in simple anemia the system instead of using less oxygen employs rather more than in health, and only in the severest forms of anemia is the percentage of oxygen in the blood dis- tinctly reduced. It therefore appears, as Grawitz concludes, that the oxygenation of the blood in the lungs is not regulated by the laws of diffusion, and is not closely dependent upon the atmospheric pressure nor upon the amount of Hb in the blood, but is governed by a special chemical activity residing in the Hb. The laws governing the respiratory changes in the blood, thus briefly sketched, principally apply to chemical processes, but have important bearing also upon certain morphological changes seen in conditions of dyspnea, asphyxia, and in the peculiar state of the blood observed at high altitudes. The EXHALATION OF VAPOR DURING RESPIRATION, while not fol- lowed under normal conditions (Dastre) by any concentration of the blood, may, if we accept the results of Grawitz's observations in patho- logical states, lead to considerable loss of water and increase of red cells and dry residue. Grawitz examined the blood of an hysterical woman before and after a prolonged attack of hysterical dyspnea, in which the respirations rose to 100 per minute, and found the red cells increased from 4.9 to 5.1 1 millions, the dry residue of the whole blood from 19.78 percent to 22.08 percent, and that of the serum from 9.54 percent to 9.85 percent. A similar result was noted after he had voluntarily in- creased his own respiration to 40 per minute during a period of one and three-fourth hours. Asphyxia. — During dyspnea from disease of heart or lungs there is proportionate increase of COj in the blood. (Kraus, Chvostek.) In ex- 356 LUNGS, HEART, KIDNEYS. treme asphyxia Zuntz collected the results of nineteen analyses, which gave for O., 0.96 volume percent and 49.53 percent of COg. In the attempt to"discharge CO2 from the tissues and to supply O^, respira- tion is accelerated, and the same result is attained by the polycythemia which results from venous stasis. Among the results following the ex- cess of CO in the blood is an alteration in the red cells which renders the Hb more readily diffusible in the plasma. This change, while not apparently leading to hemoglobinemia during life, is evident after death in the reddish tinge of the serum. One of the prominent signs in the blood of asphyxia is the failure of coagulation, which is often very complete and is probably referable to the presence of CO2. Ottolenghi, examining the blood of asphyxiated rabbits, found it to be thinner and more diffusible, while the red cells were reduced in number, resistance, and specific gravity. In the peripheral blood of the human subject, during rapid asphyxia, there are all the evidences of extreme concen- tration. In the slower forms of asphyxia, which are accompanied by pallor, the blood has apparently not been compared with that of the more rapid cases. Emphysema. — The blood in emphysema varies considerably accord- ing to the state of the circulation and respiration. During the intervals ivhen the circulation is at its best and cyanosis is absent, an anemic condition may be demonstrated in many cases, which is referable to the malnutrition of the patients or to complications, especially cirrhosis of liver, nephritis, and chronic gastritis. When cyanosis becomes chronic most observers have found more or less poly- cythemia and increase in specific gravity (Peiper,^ Grawitz^), but in Leichtenstern's cases this concentration was apparently insufficient to obscure the anemia, as he reported a loss of Hb in patients suffering from cyanosis, edema, and heart failure. An increase in the volume of red cells up to 8 tj. has been described by Vaquez,^ in cases of chronic cyanosis from various causes. During asthmatic attacks with cyanosis, the blood becomes still further concentrated and the expressed drop may appear almost black. The leucocytes are usually of high normal average, but during the asthmatic attacks there may be distinct leucocytosis. When bronchitis is added the leucocytes may be markedly increased. With or with- out leucocytosis a considerable proportion of the leucocytes may be of the eosinophile variety. Gabritschewsky, who called attention to this excess of eosins in the blood corresponding to their increase in the sputum of many asthmatic patients^ found in 3 cases from 11 to 22 percent of eosins, while Fisk reported 14.6 percent in one case, and Billings found as many as 53.6 percent out of 8,300 cells. From the observations of Leyden, v. Noorden, and especially of Schwerschew- ski, it has been shown that excess of eosins in the blood of asthmatics occurs principally or only at the time of the paroxysm, that they in- crease with repeated attacks, diminish shortly after the paroxysm and with the appearance of expectoration, and in the intervals largely dis- appear. Since there is no such tendency to eosinophilia in the suffoca- DISEASES OF THE HEART. 357 tive attacks of endocarditis, nor in the severe dyspnea of Bright's dis- ease, the examination of the blood may serve to distinguish bronchial asthma from other forms of dyspnea, and may perhaps indicate the ap- proach of a paroxysm. Teichmuller and Fuchs, however, find many eosinophile cells in other forms of bronchitis. Bronchitis. — Acute catarrhal bronchitis of ordinary severity and af- fecting the larger bronchi only, has little effect upon the red cells and seldom raises the number of leucocytes. In some severe cases the writer has found as many as 15,000 white cells, when the patient's temperature was 104°. In some of Cabot's cases the leucocytosis was slightly higher. Capillary bronchitis, with high temperature and marked prostration, may strongly resemble lobar pneumonia in clinical features, and like pneumonia, is accompanied by marked leucocytosis. In three fatal cases with high temperature, dyspnea, and cyanosis, the writer found 18,000-37,000 leucocytes. At autopsy there were signs of severe general bronchitis, no consolidation, but well-marked chronic nephri- tis. In two cases without autopsy, Cabot found as high as 41,000 leucocytes. DISEASES OF THE HEART. The majority of observations on the blood of chronic endocarditis have shown more or less continuous polycythemia and concentration of the blood. The more important of these studies have been con- tributed by Nasse, Naunyn, Toenissen, Bamberger, Lichtheim, and Reinert,^ who have referred this concentration to stasis from local vasomotor or general cardiac paresis, to transudation of serum, or (Naunyn) to a physiological effort on Nature's part to furnisli more oxygen by increasing the cells and Hb of the blood. The opposite condition, of hydremia, has been found more usual or regarded as more important, by Leichtenstern, Oertel, Oppenheimer, Stintzing and Gumprecht, and many others, who hold that an excess of lymph from various sources passes into the vessels as a result of lowered blood pressure. Schneider and Limbeck find that while mitral lesions are usually associated with polycythemia, aortic insuffi^ciency more often leads to hydremia. All observers are agreed that the valvular lesion and the usual causes leading to it exert little or no effect upon the blood, and while differing as to their significance all admit that the changes in the blood vary considerably at different periods of the disease. The conflicting views have been carefully studied and partially uni- fied by Grawitz, who finds that the blood shows three different states, according to the stage of the disease, and determined chiefly by the de- gree of compensation established in the heart. 1. When valvular lesions are fully compensated by cardiac hyper- trophy, and when there are few or no symptoms of disease of this or other organs, the state of the blood depends entirely on the constitu- tion and state of nutrition of the individual. 358 LUNGS, HEART, KIDNEYS. 2. When disturbances of compensation begin, when the heart muscle fails, and the pulse is accelerated, and dyspnea and its train of symp- toms appear, changes occur in the blood which are clearly referable to the diminished force of the heart. The whole blood loses in specific gravity and in dry residue, while its content of water increases, and these changes are more marked in the veins than in the superficial capillaries. The red cells are reduced in number but they suiFer no changes in size or in Hb-content. The leucocytes show no characteristic changes. The blood serum shows the most marked changes, always becoming more watery. These altera- tions Grawitz refers solely to the lowered blood pressure, which is fol- lowed by dilatation of capillaries and passage of tissue fluids into the blood stream. Oliguria may result from the same factors but is not the cause of the hydremia. 3. When chronic venous stasis is established and dyspnea, cyanosis, and edema exist, the changes in the blood are more complicated. Under these circumstances the blood loses water, becomes richer in red cells, and more so in the capillaries than in the veins. These ef- fects Grawitz believes to result from transudation of fluids through the capillaries into the tissues, but increased exhalation from the congested lungs, and possibly also increased evaporation from the skin, contrib- ute to the same result. When the balance of pressure between capil- laries and tissues has been adjusted, any added failure of the heart may be followed by still lower pressure and temporary hydremia. Grawitz's deductions have appeared to most critics to be only partially valid. In the first stage it is generally agreed that the blood is usually nor- mal, and it is also fully demonstrated that a fall of blood pressure leads to relative hydremia, but his explanation of the origin of the polycythemia of chronic endocarditis has not proven fully satisfactory. As Limbeck points out, many patients always show some reduction of red cells whether compensation is complete or not ; while others show considerable variations in red cells without apparent relation to the action of the heart, or the pres- ence or absence of edema, or the activity of the kidneys. Limbeck, there- fore, attributes less importance to the interchange of fluids between the blood and tissues, and believes that the polycythemia of heart disease is partly referable to the same obscure factors which increase the red cells in high altitudes. That there is an actual new formation of red cells in the polycythemia of endocarditis is held by Marie, Reinert,^ and others. Without entering further into the details of individual opinions, it appears certain from the studies already cited and from others contrib- uted by Siegl, Peiper,* Schmaltz, Banholzer, Maxon, Stintzing and Gumprecht, that : 1. In advanced endocarditis with failing compensation, there is a distinct tendency toward concentration of the blood, which at times in- creases the red cells to 8 millions or more. 2. This polycythemia is the combined result of venous stasis, dysp- nea and cyanosis, transudation of blood serum, and possibly also of other obscure factors which lead to polycythemia in high altitudes. 3. The state of the blood in chronic endocarditis responds to a I DISEASES OF THE HEART. 359 limited extent, but not invariably to diuretics, diaphoretics, purges, and heart tonics, 4. The constant tendency toward anemia is usually masked by the peculiar condition of the circulation. 5, The result of the examination of the blood must be interpreted only with strict regard to the general condition of the patient. Results of Blood Examinations in Chronic Endocarditis. — The forego- ing statements are fully borne out by a review of the reported exami- nations of the blood in endocarditis. In some series of cases the polycythemia is very uniform, as found especially by Oppenheimer and Reinert, while the Hb commonly runs from 90— 110 percent. Usually the red cells do not exceed 6 millions, but some cases with cyanosis reach 7 or even 8 millions. On the other hand a slight reduction of cells, not below 4 millions, appears in most cases reported by other observers. Maxon's statement that the blood of heart disease com- monly shows slight variations above or below the normal in red cells, specific gravity, and albumens, is the only rule of general application. That aortic insufficiency is usually found with less concentrated blood is also apparent in nearly all comparative studies. (Schneider, Grawitz, Limbeck, Menicanti, Reinert, Hayem, Sadler.) Yet even here the red cells seldom fall below 4 millions. Variations from the above rules are not unknown, as some cases with cyanosis have shown less than 4 million red cells, and aortic in- sufficiency may lead to polycythemia. Much depends upon the par- ticular circumstances surrounding each patient, which must always be regarded with care. The LEUCOCYTES in uncomplicated cases show no important varia- tions from the normal, but with concentrated blood they are apt to be rather above the average normal figures. During the febrile periods which mark the terminal stages of many cases, leucocytosis of con- siderable grade is usual. The complications of endocarditis which cause leucocytosis are numerous, and ante-mortem leucocytosis is usually prolonged and pronounced. Malignant Endocarditis. — The more acute cases of bacterial endo- carditis furnish some of the most typical examples of septicemia, in which the blood shows rapid loss of cells, albumens, and Hb, leucocy- tosis, and very often bacteria in demonstrable numbers. The anemia of malignant endocarditis is of rapid progress and usually becomes one of the characteristic features of the disease. Be- ginning in patients with practically normal blood, the septic process often reduces the red cells within two to three weeks to 3.5 or 3 mil- lions, or even lower. In some cases, usually those of longer duration, the anemia is less marked (over 4 million cells), or is masked by the febrile process and by venous stasis. In all the markedly septic cases, however, the anemia becomes severe. In the later stages of severe cases the evidences of destruction of red cells become marked, the Hb is very deficient and may become dissolved in the plasma, shadow cor- puscles appear in the fresh specimen, and various forms of degeneration 360 LUNGS, HEART, KIDNEYS. of red cells are demonstrable in the dry specimen. In such cases the red cells may not number much over one million. After death the deposits of blood pigment in the viscera are very abundant and closely simulate those of pernicious malaria, for which they may readily be mistaken. Leucocytosis is present probably in all cases, but is often inter- mittent and single observations may fail to discover it. In this, as in other forms of sepsis, there may be a tendency to slight leucocytosis with high proportion of polynuclear cells, and in general the increase of white cells is not very marked, considering the condition of the patient. Roscher and Cabot report fatal cases with 8,000 and 8,900 leucocytes shortly before death, while Krebs found ante-mortem leuco- cytosis of 44,200, and Grawitz reported 168,000. Baeteriological examination of the blood is often required in the di- agnosis of obscure cases. From the rather limited number of recorded examinations by reliable methods it appears that bacteria are always to be found in the blood in a certain group of cases of ulcerative endo- carditis. These cases include the examples of cryptogenic infection which run an acute septic course and end fatally within a few weeks or months. In the cases of ulcerative endocarditis which terminate with septic fever and in which an acute process is often added to the chronic lesion, cultures of the blood are usually negative. Grawitz' claim that repeated negative cultures are sufficient to rule out ulcerative endocarditis cannot, however, be admitted ; for the evidence rather favors Kuhnau's view that the difference between " simple " ulcerative and malignant endocarditis is one of degree, not of kind, and that in many cases of subacute or chronic ulcerative endocarditis of bacterial origin the blood is sterile. These groups of cases are illustrated by the reports of several investiga tors. Grawitz 2 examined 7 cases of suspected ulcerative endocarditis, all of which gave negative cultures but only one of which proved to have an ulcerative lesion. In three pronounced cases of malignant endocarditis he obtained in two Staphylococcus pyogenes aureus, in one Diplococcus lanceolatus. Kraus examined the blood in 7 cases, finding the streptococcus in one, negative cultures in six. Petruschky, examining the blood from a wet cup, obtained the Streptococcus pyogenes in one, and a negative result in the other, of two cases. Kuhnau had only one positive result {Staphylococcus pyogenes aureus) in 12 cases examined. Cohn in 2 acute cases found in one both Streptococ- cus pyogenes smd Staphylococcus aureus, and in the other Staphylococcus aureus alone, while in two chronic cases the cultures were negative. James and Tuttle obtained positive cultures in 3 of 4 fatal cases, once finding the Dip- lococcus lanceolatus. White obtained negative results in two cases 3-4 days before death, but isolated Staphylococcus pyogenes aureus from both just be- fore death. The writer, using 5-10 cc. of blood drawn from the median basilic vein, obtained pure cultures of Streptococcus pyogenes in two typical cases of malignant endocarditis. These patients gave no history of cardiac disease until the development of a septic febrile process which continued without intermission for 1-6 months, and until death. In four cases of chronic ulcerative endocarditis ending with septic fever cultures of the blood were negative. Gonorrlieal Endocarditis. — The gonococcm has been obtained in pure DISEASES OF THE KIDNEY. 361 culture from the circulating blood of cases of endocarditis by Thayer and Blumer, and Thayer and Lazear, and Halle. In all cases the blood was extremely anemic, red cells under 2 millions, while the leucocytosis was slight, 12,000-14,000 in one case, and 8,500-18,000 in the others. Many cases of gonorrheal septicemia collected by these authors indicate the frequency with which this germ is probably connected with endocardial lesions. Columbini, also, obtained the gonococcus in cultures of the circulating blood in a case of gonorrheal septicemia with endocarditis, and dem- onstrated its identity and pathogenic qualities in the human urethra. McCallum and Hastings report a case of acute endocarditis from which they obtained from the circulating blood cultures of a somewhat peculiar coccus resembling Streptococcus pyogenes. Significance of the Bacteriological Examination of the Blood in Endo- carditis. — 1. A positive result when obtained under proper precautions, and long enough before death to avoid ante-mortem secondary infec- tions, places beyond doubt the infectious nature of the process. 2. A negative result does not indicate that the vegetations do not contain bacteria in their substance or on their surface. A large num- ber of negative bacteriological examinations of the blood have been recorded in cases showing, post-mortem^ various bacteria in and on the inflamed heart valves. From personal experience of moderate extent the writer has drawn the impression that when malignant endocarditis follows the type of pure septi- cemia with cardiac symptoms in the background, bacteriological examina- tion of the blood is usually positive, but when cardiac symptoms are or have been prominent, bacteriological examination of the blood is usually negative. Congenital Heart Disease. — The pronounced cyanosis from which most of these patients suffer leads to extreme degrees of concentration of the blood, as the following cases will show : Author. Red cells. Leucocytes. Hb. S. G. Krehl 8 mil. 8.9 9.44 8.47 6.70 8.10 8.82 7.54 8.43 Vaquez^ Banholzer 160% 110 92 1.071 Gibson 12,000 12,000 16,000 Carmichael Toenisson , Cabot Townsend has recently reported 13 cases in which the red cells varied between 5.6 and 11.8 millions. DISEASES OF THE KIDNEY. The existence of marked hydremia in cases of nephritis was dem- onstrated in several earlier studies of the chemistry of the blood 362 LUNGS, HEART, KIDNEYS. (cited by Gorup-Besanez), but Schmidt's analyses first clearly showed the more exact nature of this hydremia. In three patients with albuminous urine and marked edema, he found considerable loss in specific gravity of the whole blood (1.043-1.051), and of the serum (1.018-1.024), while the red cells remained nearly normal in gravity (1.081-4) but were much reduced in bulk (34.2 from 44.9 percent). These changes were believed to result largely from the loss of albumens of the serum, while the Hb of the red cells and the salts of the plasma were but slightly affected. Scherer's analyses in six cases demonstrated a de- crease in the solids of the whole blood, and a loss of 5 percent of the albu- mens, although the fibrin was sometimes increased. The red cells were re- duced but their dry residue relatively increased. Although the serum was very watery, the dry residue averaging 6.9 percent, the salts of the serum were increased. Very low gravity of the serum, 1.019-1.023, was also noted by Becquerel and Rodier. From numerous other analyses by Frerichs, Gorup-Besanez, Hinter- berger, and others, it began to appear that the composition of the blood varies considerably in different stages of the disease, and later studies have shown that the two main clinical and pathological types of the disease, the chronic exudative and the chronic productive without exudation, are attended with distinct changes in the blood. 1. Chronic Exudative (Parenchymatous) Nephritis. — The albu- minuria and edema of this group of cases are associated with well- marked anemia, which, however, is subject to great variations on ac- count of the intermittent course of the disease and of the frequent disturbing effects of treatment. The usual condition of the blood is one of moderate chlorotic anemia. The majority of cases show between 3 and 5 million red cells and 40- 80 percent of Hb. The reported series of cases in the literature show wide variations in the condition of the blood. Considerable polycy- themia is not infrequently encountered. Leichtenstern and Sorensen found little reduction of Hb and an average of 4.74 million red cells, while Laache and Reinert ^ found a low Hb-percentage with slight loss of cells. These changes represent the condition most commonly found. Cases with less than 2 million cells are reported by Grawitz, Sadler, Cabot, and are not rare. In these and occasionally in other cases the writer has sometimes found a high or increased Hb-index, and at autopsies in cases of pernicious anemia has several times found the lesions of chronic productive nephritis with exudation. These variations are well illustrated by Cabot's synopsis of 35 cases from the Massachusetts General Hospital. Red cells. Cases. 6 7 millions 3 5-6 " 6 4-5 " 12 3-4 " 11 2-3 " 2 1-2 " 1 The attempt to analyze the causes of the variations of the anemia in chronic nephritis is met with difficulties, and numerous factors must be considered. Polycythemia results from cyanosis, disturbance of cir- DISEASES OF THE KIDNEY. 363 culation, and rapid transudations, and each of these conditions may and frequently does obscure pronounced anemia. An astonishing degree of pallor of the face may thus be found with 5 million red cells and 100 percent of Hb. The progressive anemia of the average case is clearly referable to the loss of albumens of the serum and general malnutrition. Yet the real advance of the anemia is often disturbed by the intermittent losses of albumen and attacks of edema. While Benczur and Czatary, and v. Jaksch ^ believe that the hydremia is not proportionate to the edema, yet Bogdanow, and Stintzing and Gum- precht have shown a positive connection between hydremia and edema, and have noted an improvement in the blood following improvement in general symptoms. The relatively high Hb-index is apparently the natural result of a loss of albumen which affects principally the serum. The gravity of the red cells has been found uniformly high by most analysts since the time of Scherer. Grave or pernicious anemia develops in a small proportion of cases of chronic parenchymatous nephritis, and appears at times to result directly from nephritis. It must be referred to the repeated losses of albumen from the blood and to the general disturbance of nutrition, but it is probable that the very severe grades of anemia result from the combined effects of lesions in several viscera, including chronic gas- tritis, cirrhosis of liver, arteriosclerosis, etc. Chemistry. — Chemical analysis gives the clearest insight into the changes in the blood of nephritis. The specific gravity is regularly reduced. Peiper^ found a gravity of 1.026 in a very anemic child, but after marked improvement four weeks later, the gravity was 1.055. Similar variations, usually between these limits, are reported by various observers. The changes in the serum demonstrated by the older ob- servers have been verified by many later studies. The gravity of the serum is much reduced (as low as 1.013, Bostock), and its volume remains high. While considerable relation has been found to exist between the changes in the blood and the albuminuria and edema (Stintzing, Gumprecht, Bogdanow), the immediate effects of transudates and albuminuria are not always evident, (v. Jaksch.) Hammar- schlag, who found the gravity of the serum between 1.018 and 1.030, concluded that edema has more effect upon the gravity of the serum than has albuminuria ; and that when edema is absent the gravity of the serum is usually about normal. The leucocytes in chronic nephritis are usually normal or subnormal in number. Cabot found no leucocytosis in 31 of 40 unclassified cases, while in 14 of 19 uremic cases there was leucocytosis, reaching 44,- 000 in one eclamptic patient. Other reports show similar variations, and it is evident that the behavior of the leucocytes in chronic nephritis depends upon accidental conditions and complications. Chronic Interstitial Nephritis. (Contracted Kidney.) — The ab- sence of edema and marked albuminuria in cases of chronic nephritis without exudation allows the blood to remain practically normal, at 364 LUNGS, HEART, KIDNEYS. least during the latent progress of the disease. Many patients are carried oiF by acute uremia and other terminations of interstitial ne- phritis before the blood is markedly altered. If at any time in the progress of the disease there are exacerbations marked by exudation, albuminuria, and edema, the blood suffers for a time as in chronic ne- phritis of more distinctly exudative type. After the exudative period the blood is partially restored, but a certain grade of chlorotic anemia is apt to persist. That chronic nephritis may run its course without giving notable changes in the blood, or indeed any other pronounced symptoms, is a well attested fact, of which the writer has seen at autopsy at least one instance. This subject was an able-bodied policeman dying of acute colitis, whose kidneys were shrunken to an extreme degree. The blood and urine were normal shortly before death. On the other hand, very severe grades of anemia, usually of micro- cytic type, are sometimes associated with chronic diffuse nephritis with- out exudation. Some of these cases are partly referable to lead poi- soning, which has initiated the renal lesion and damaged the blood. In others, there is advanced arterio-sclerosis with frequent hemorrhages from the nose, kidneys, or other regions. In still others no reason- able explanation of the anemia is to be found. Grawitz recognizes two stages in this disease, one in which the heart and circulation, and the blood, are normal and a second in which compensation fails in the hypertrophied heart and the blood suffers the same changes as in uncompensated valvular disease. The importance of the condition of the heart in determining the blood changes in nephritis is very great and has already been considered in the references to edema and cyanosis, but failure of the left ventricle with feeble pulse is more common with the large white than with the small contracted kidneys. The writer is, therefore, unable to recognize the above stages in the course of the contracted kidney, finding them more evident in the cases of chronic diffuse nephritis with exudation, and in patients with albuminuria and edema. Acute Nephritis. — The changes in the blood in acute nephritis with albuminuria and edema resemble those of the same type of chronic nephritis. Peiper found no diminution in specific gravity in an acute case. Laache and Bogdanow observed more rapid, severe, and more variable changes in the acute than in the chronic cases, while v. Jaksch concluded that in both forms the changes were about equally variable. Hayem found no considerable loss of red cells except in hemorrhagic cases. Many other isolated reports show all but the very severe grades of anemia developing within a few weeks, sometimes very rapidly. Leucocytosis (maximum 22,000) has been observed in a considerable proportion of cases by Hayem, Sadler, and Cabot, the latter author re- ferring the persistent increase to loss of blood by the kidney. None of these observers has attempted to connect the behavior of the leucocytes with the character of the lesion in the kidneys. It would seem that the productive inflammation could run a fatal course without leucocy- tosis, that simple acute exudative nephritis might be accompanied by a DISEASES OF THE KIDNEY. 365 moderate increase of white cells, and that acute idiopathic or secondary purulent nephritis must nearly always be accompanied by marked leucocy- tosis. Unfortunately there are no reports at hand on which to base these conclusions. Profuse hemorrhage, uremic attacks, and complicating infec- tions, may be responsible for occasional leucocytosis, but Sadler's negative cases show that much albumen, many granular casts, and blood cells, may be present in the urine when the leucocytes in the blood are not increased. Uremia. Chemistry. — Very numerous chemical analyses of the blood in uremia have failed as yet to demonstrate the true nature of this intoxication. Urea in the blood was believed by Frerichs to be the particular toxic agent in uremia, and although this principle has been demonstrated in abnormal quantities in the blood of uremic coma by Spiegelberg, Hoppe-Seyler, Bartels, and others, it was shown by Landois that intravenous injections or local applications to the medulla of large quantities of urea exert no toxic influence. Neither has its de- rivative, ammonium carbonate, been found in the blood, nor shown to exert a toxic influence on the nervous system (Kuhne, Strauch). Kee- ATININ, believed by Schottin and Perls to be the active agent, has like- wise been set aside although this substance is increased in the blood of uremic subjects. Potassium has been found in considerable excess in the blood by Feltz and Ritter, and by Astaschewsky, who regard uremia as a form of potassium poisoning. Although Sneyers and Hor- baczewsky failed to find an excess of potash salts in eclamptic and uremic subjects, this theory has received more recent support frow Bouchard, Roger, Rovighi, and others, who oflFer evidence to shom that potassium is one of several poisonous substances which accumu- late in the blood in uremia. Limbeck has shown, however, that while an excess of potash can be obtained from the blood of uremic dogs, he could get no such excess if the blood was examined before death. Diminished alkalescence was first noted in uremic blood by v. Jaksch,^ and subsequently by Peiper,^ Rumpf, Mya and Tassinari, Limbeck, and others, who have shown that the alkalescence diminishes greatly on the approach of uremic symptoms. The attempt to estab- lish this theory of acid intoxication has, however, not been successful. Although uric acid may be in excess in uremia, it is also quite as much increased in many other conditions. Fatty acids, and phosphoric acids were not found in excess by Limbeck, who indeed was unable to de- monstrate a deficiency of CO2 in all cases of experimental uremia. Diminished alkalescence appears more probably to be only a secondary condition of the blood in uremia. The present tendency is to regard the uremic seizure as the result of toxic action of a variety of nitrogenous metabolic products which are supposed to be retained in the system or at times thrown off" in excess in the urine. Bouchard and others have isolated from the urine a variety of toxic principles, including ptomaines and urotoxines, some of which produce convulsions, others are narcotic, while still others lower temperature, contract the pupil, or produce salivation. Although the actual existence of these principles has been denied (Stadthagen), and 366 LUNGS, HEART, KIDNEYS. the urine has not always been found so toxic as Bouchard claims it to be (Fleischer), the autotoxic theory is generally regarded as approach- ing most nearly to the true explanation of uremia. One of the chief objections to it consists in the fact that uremic attacks frequently arise at periods when the excretion of urine is much improved. Bibliography. Diseases of Heart, Lungs, Kidneys. Astaschewsky. St. Petersb. med. AVoch., 1881, No. 27. Bamberger. Wien. klin. Woch., 1888, No. 1. Banhoher. Cent. f. inn. Med., 1894, No. 23. Bartels. Ziemsenn's Handb., Bd. IX., p. 1. Benczur, Czaiary. Deut. Archiv klin. Med., Bd. 46, p.*478. Biernacki. Cent. f. inn. Med., 1895, No. 14. Billings. N. Y. Med. Jour., Vol. 65, p. 691. Bogdanow. St. Petersb. med. Woch., 1875. Bohland. Berl. klin. Woch., 1893, No. 18. Bostock. Cited by Bartels. Bouchard. Lecons s. 1. Autointox., Paris, 1887. Carmichael. Edin. Hosp. Reports, 1894, p. 298. Chvostek. Wien. klin. Woch., 1891, Nos. 6, 7. Cohn. Deut. med. Woch., 1897, p. 136. Cohnstein, Zuntz. Pfliiger's Archiv, Bd. 42. Columbini. Cent. f. Bact., 1898, Bd. 24, p. 955. Copeman. Brit. Med. Jour., 1891, I., p. 161. Baslre. Semaine Med., 1893, p. 580. Feltz, Bitter. De I'Uremie exper., Paris, 1881. Fink. Diss. Bonn, 1890. Fleischer. Verb. Cong. inn. Med., 1885, p. 307. Frerichs. Nierenkrankheit., Braunschweig, 1851. Fuchs. Deut. Archiv klin. Med., Bd. 63, p. 427. Gabritschewsky. Archiv f. exper. Path., Bd. 28, p. 83. Gibson. Lancet, 1895, I., p. 24. Gorup, Besanez. Archiv f. phjs. Heilk., 1849, p. 514. Grawitz. ^Charite-Annalen, Bd. 19, p. 154. Halle. Annal. d. Gynecol., Bd. 50, p. 179. Hamburger. DuBois Reymond's .\rchiv, 1892, p. 513. SuppL, Bd. 1893, p. 167. Hammarschlag. Zeit. klin. Med., Bd. 21, p. 475. Hinterberger. Archiv f. physiol. Heilk., 1849. Hoppe-Seyler. Physiol. Chem., 1881. Horbaczeicski. Med. Jahrb. Wien., 1883, p. 385. V. Jaksch. 'Zeit. f. klin. Med., Bd. 23, p. 187. ^Zeit. f. klin. Med.,Bd. 13, 1887. James, Tuttle. Presbvterian Hosp. Reports, N. Y., Vol. 3, p. 44. Kraus. Zeit. f. Heilk., Bd. 17, p. 117. Zeit. f. klin. Med., Bd. 18, p. 160; Bd. 22, p. 449. Krebs. Inaag. Diss. Berlin, 1893. Krehl. Deut. Archiv klin. Med., 1889. Kruger. Zeit. f. Biol., 1890, Bd. 8, p. 452. Kuhnau. Zeit. f. Hygiene, Bd. 25, p. 492. Kuhne, Strauch. Cent. f. inn. Med., 1864, No. 36. Laache. Die Anemie, 1883. Landois. Die Anemie, Wien, 1891. Leichtenstern. Untersuch. u. d. Hb-gehalt, etc., Leipsic, 1878, p. 88. Leyden. Deut. med. Woch., 1891, p. 1085. Liehiheim. VII. Cong. inn. Med., 1888. Loury. Pfliiger's Archiv, Bd. 58, p. 409. Lukjanow. Grundriss ein. allg. Path. d. Gefasssystems, Leipsic, 1894. McCallum, Hastings. Jour, of Exper. Med., 1899, p. 521. Marie. Semaine Med., 1895, No. 4. Maxon. Deut. Archiv klin. Med., Bd. 53, p. 399. Menicanti. Deut. Archiv klin. Med., 1892, Bd. 50. Mya, Tassinari. Archiv p. 1. sci. med., 1886, No. 20. BIBLIOGRAPHY. 367 Naunyn. Corresp. Schweizer Aertze, 1872, p. 300. V. Noorden. Zeit. klin. Med. , Bd. 20, p. 98. Oertd. Deut. Archiv klin. Med., Bd. 50, p. 293. Oppenheimer. Deut. med. Woch., 1889, Nos. 42-44. Ottolenghi. ( Ref. ) Allg. med. Central-Zeitung, 1894, p. 724. Peiper. iCent. f. klin. Med., 1891, p. 217. ^Virchow's Archiv, Bd. 116, p. 337. Pembry. Text-book of Physiol. (Schafer), I., p. 717, Perls. Berl. klin. Woch., 1868, No. 19. Petruschky. Zeit. f. Hygiene, Bd. 17, p. 59. Beinert. iZahlung d. Blutkorp, 1891, p. 199. "Miinch. med. Woch., 1895, No. 15. Rogers. Action du foie sur 1. poisons, Paris, 1887. Roscher. Inaug. Diss. Berlin, 1894. Rovighi. Rivista clin., 1886. Rumpf. Cent. f. klin. Med., 1891, p. 441. Sadler. Fort. d. Med., 1891, Suppl.-Bd. Seherer. Haeser's Archiv, Bd. X., p. 121. Schmaltz. Deut. med. Woch., 1891, No. 16. Schmidt. Zur Charac. d. epidem. Cholera. Schneider. Diss. Berlin, 1888. SchoUin. Archiv f. physiol. Heilk., 1853, p. 170. Siegl. Wien. klin. Woch., 1891, p. 606. Sneyers. Path. d. Nephritis chron., Brussels, 1886, p. 144. Sorensen. Cited by Grawitz. Spiegelberg. Archiv f. Gynecol., 1870, p. 383. Stadthagen. Zeit. f. klin. Med., Bd. 15, p. 383. Stintzing, Oumprecht. Deut. Archiv klin. Med., Bd. 53, p. 465. Swerschewski. Cent. f. inn. Med., 1895, p. 183. Teichmuller. Deut. Archiv klin. Med., Bd. 60, p. 576; Bd. 63, p. 444. Thayer, Lazear. Jour. Exper. Med., 1899, p. 81. Toenissen. Diss. Erlangen, 1881. Townsend. Boston Med. Surg. Jour., Vol. 142, p. 426. Vaquez. 'Compt. Rend. Soc. Biol., 1895, p. 142. ^Traite de Med. (Charcot), V., 298. White. Jour, of Exper. Med., 1899, p. 425. Zuntz. Hermann's Handb., IV., 2, p. 43. CHAPTER XXII MALIGNANT TUMORS. CARCINOMA. The impression obtained by individual observers of the effect of carcinoma upon the blood has largely depended upon the class of cases encountered. In active hospital wards the cases are usually advanced, cachexia is distinct, and routine examination of the blood shows as a rule, marked reduction of red cells, low Hb-index, and moderate but distinct leuco- cytosis. In dispensaries the earlier stages of the disease are encoun- tered, cachexia and palpable tumors are not always noted, many tenta- tive diagnoses stand with positive cases, and the blood very often fails to show any pronounced alteration. In a considerable group of cases the progress of the usual anemia is interrupted, however, for unknown reasons, and while the patient emaciates, the blood appears to maintain its standard. Hampeln, Neubert, Dehio, and others, therefore speak of an anemic, and a marantic type of carcinoma. Neubert's cases, 5 esophageal and 4 gastric, were especially adapted to illustrate the differences, but both types may occasionally be seen in other forms of the disease, e. g., can- cer of uterus. Many careful studies have shown that in the earlier recognizable stages of the disease, and in somewhat more advanced phases of visible growths, the blood remains practically normal. Few published series, how- ever, show so large a proportion of cases with red cells above 4 millions as does Cabot's, in which 34 of 72 cases of carcinoma of stomach gave over 4 million, and 19 over 5 million red cells. Yet this series may well serve to emphasize the fact that carcinoma com- monly exists in otherwise healthy subjects, and for a time fails to reduce the number of red cells. The significance of a normal red cell count is of course limited by the fact that in many situations the growth tends to retard the ingestion and absorption of fluids and thus to concentrate the blood. Yet in some instances the malignant growth, like tubercu- losis, appears to exert some obscure concentrating influence on the blood, and even at death, although the body is emaciated, the blood may be unusually deep red. Osterspey found 5 millions of red cells and 98 percent of Hb three months after the symptoms of a gastric cancer were developed, w^ile in a patient who had lost 100 pounds weight within a year from gastric cancer, he found 4,544 red cells and 82 percent of Hb. Laache, who noted the same apparent CARCINOMA. 369 immunity of the blood against the effects of a bleeding uterine carcinoma, concluded that there is an individual insusceptibility to the effects of malignant tumors. It is, nevertheless, true tliat in the majority of carcinomata the appear- ance of cachexia is accompanied by a corresponding oligocythemia. Cancers of the stomach are among the more active in reducing red cells, and some of them, even while remaining of small size, lead to the changes of pernicious anemia. Grawitz counted only 500,000 red cells in one extreme case of gastric cancer. Of complications, hemor- rhage and ulcerations usually affect the red cells promptly. Rapid growth and numerous metastases also have a similar tendency but prob- ably less marked. Morphologically, the red cells in carcinoma, while presenting the usual degenerative changes of secondary anemia, do not show any special peculiarities. The usual changes are those of well-marked chlorotic anemia with considerable deficiency of Hb, but with mod- erate changes in size and shape, Grawitz found granular degeneration of many cells in 10 cases of cancer of the stomach or esophagus, but none in two cases of uterine cancer. The writer finds that in those oases in which the appearance of the blood suggests pernicious anemia, the deficiency of Hb, even in the majority of megalocytes, is nearly always quite distinct. Mouisset has already called attention to this fact, finding the Hb-index in these cases to be quite as low as in chlorosis. In a case of Daland's, the hematocrit gave rather fewer red cells than the hematocytometer, but in per- nicious anemia the volume of the red cells is usually much above normal. (Cf. Moraczewsky.) Nucleated red cells are commonly seen when the anemia is severe, and in some cases of gastric cancer they may be extremely abundant. Usually these cells are of normal size, but megalocytes appear in the severest cases. It has been suggested that the predilection for the bone marrow as the seat of metastases is responsible for the large num- ber of nucleated red cells seen with many advanced carcinomata. (Epstein.) They are sometimes seen, however, in the early stages of the growth and when anemia is slight. The Hemog'lobm begins to suffer much earlier than the number of red cells. The writer has never seen an uncomplicated visceral carci- noma, positively identified, which had failed to reduce the percentage of Hb and, barring clear cases of polycythemia, can find none fully re- ported in the literature. Fully normal Hb appears to be a very re- liable negative indication against carcinoma at any stage. There are however reports of 90-100 percent of Hb in cases with increased numbers of red cells, and the loss is doubtless slight during the early stages of most cases. Usually a malignant epithelial tumor rather promptly reduces the Hb, so that during the greater part of its prog- ress the Hb is below 75 percent. In Bierfreund's and Reinbach's series of 57 surgical cases, it varied between 18 and 80 percent, while few registered over 70 percent, and in comparison with benign tumors 24 370 MALIGNANT TUMORS. the Hb in cancer was distinctly lower. After operation the restora- tion of the Hb required about a week longer with malignant than with benign tumors, and the restoration was never quite complete. With internal visceral carcinomata the Hb has usually been found even lower, but there are numerous exceptions to this rule, usually de- pending on the site of the tumor. A rather uniformly low percentage of Hb in visceral carcinomata was found by Laker, Eiclihorst,Haberlein, Laache, Dehio, Leichtenstern, Moraczewsky, and others, while Sailer's and Taylor's series of 21 cases gave the unusually low average of 25 percent. Osterspey's and Cabot's cases were somewhat higher, often approaching the normal. Yet Eaberlin concluded that, in steno- sis of the pylorus, cancer is contradicted if the Hb is over 60 percent. The Hb-index is uniformly low. In some cases it has approached that of chlorosis, and although it is usually somewhat higher, carci- noma offers a closer resemblance to chlorosis in this respect than almost any other cause of secondary anemia. As the disease progresses the anemia itself becomes more severe, and when chronic cachexia has been long established the blood shows the characters of secondary pernicious anemia, with variations in the size of the cells and increasing Hb-index. Yet with rare exceptions the relatively great loss of Hb remains char- acteristic. Leichtenstern 2 long since called attention to a rapid increase of Hb, to 100 percent or more, seen in some cases of gastric cancer shortly before death, and to the tarry appearance of the blood at autopsy. Patrigeon, in a gastric case, found the red cells normal nine days before death. The writer has ob- served this condition in patients who had failed to take any food and little drink for some days before death. Special Factors in the Anemia of Carcinoma. — The special con- ditions leading to severe anemia are numerous, and some are obscure. Hemorrhage and ulceration are among the frequent complications recog- nized as rapidly impoverishing the blood, but occasional escape from their natural effects has been recorded. The site of the tumor may be such as to interfere with nutrition, as do some gastric carcinomata. A previous condition of anemia may exist before the development of the cancer, as when a round ulcer of the stomach becomes malignant. Rapidly forming and numerous metastases are usually associ- ated with severe and progressive anemia, but there are numerous ex- ceptions to this rule, some emaciated subjects showing a tendency to concentration of the blood. In two cases with extensive bone meta- stasis the writer found only moderate anemia and but few nucleated red cells. Individual idiosyncrasy appears to influence the course of the cachexia both favorably and unfavorably. The histological char- acter of the tumor appears to have no relation to the anemia except in so far as the structural type favors hemorrhage or ulceration. The chronic toxemia of carcinoma may apparently affect the blood in a variety of ways. Muller offers evidence to show that the blood of cancerous patients contains toxic principles which destroy albumens and diminish alkalescence. Grawitz injected the alcoholic extract of CARCINOMA. 371 a cancerous tumor into rabbits, and finding a considerable loss of gravity and dry residue, concluded that the toxines of carcinoma tend to dilute the blood by inducing excessive flow of lymph. Maragliano has demonstrated an increased globulicidal activity of the serum in carcinoma. The Leucocytes. — Leucocytosis in carcinoma was very early recog- nized as a nearly constant condition, having been observed post-mortem by Andral, in 1823, and in the living blood by Lucke, and by Virchow, about 1867. The white cells were long regarded as derivatives of the tumor, an error which was slowly relinquislied when leucocytes were submitted to more careful classification by Schultze. That the carcinomatous process has in itself any capacity to draw leucocytes to the blood may still be doubted, although the edges of the tumor are always inflamed and adjacent lymph nodes invariably show inflammatory hyperplasia before metastasis. In tumors of ordinary size and character this inflammatory process is usually too limited to aflFect the leucocytes, but with very large growths, as in the liver, lung, uterus, etc., the frequency of leucocytosis without complications, indi- cates that a large actively growing tumor, merely through its local irritation, may induce leucocytosis. Thus Hayem saw a leucocytosis of 21,700 disappear after the removal of a scirrhus of the breast, only to return again with the recurrence of the tumor. Hayem believes that recurrences may be predicted by the appearance of a gradually increasing leucocytosis. On the other hand, advanced cachexia may exist with hypoleucocytosis. (Sailer, Taylor.) Rapid growth is much more eff^ective in raising the number of leucocytes, probably by excit- ing more active local reaction, and by more sudden disturbance of function in the part affected. Superficial ulceration, or other inflammatory complications, are the usual causes of well-marked leucocytosis with carcinoma. Hemorrhages usually accompany ulceration, and add distinctly to the excess of leu- cocytes commonly seen in bleeding and ulcerating tumors. The type of the tumor has a distinct relation to leucocytosis, which, however, cannot always be traced. The cellular carcinomata tend to irritate, ulcerate, soften, and bleed, while fibrous tumors are of slow growth, and non-vascular. The writer has observed two very large gelatinous carcinomata of the peritoneum without leucocytosis. The diffuse cancers of the stomach regularly excite less increase of leucocytes than does the adenomatous type. The situation of the tumor often determines the frequency of leucocy- tosis, from special liability to ulceration or bleeding. Many cancers of the stomach and uterus illustrate this rule. Epithelioma of the esophagus rarely induces leucocytosis unless there is extreme ulcera- tion and extension. Unusually high leucocytoses are reported in a case of carcinoma of thyroid, by Hayem, and in cases of unidentified tumors of the kidney, by Cabot, and in several cases of gastric cancer, (q. v.) Leucocytosis in Cancek of the Breast. — Alexander found an average 372 MALIGNANT TUMORS. of 11,400 leucocytes, and variations between 2,360 and 21,700, in 14 cases of scirrhus. In three cases of alveolar carcinoma with extensive metastasis the leucocytes numbered 10,075, 11,625, and 12,400. Hayem (p. 947) found a distinct reduction of leucocytes following 4 operations for scirrhus. It will be seen that the leucocytosis of carcinoma is referable largely to complications, and these complications are such as appear very con- stantly in rapid or advanced cases. These conclusions accord with the o-eneral experience that most cases of well-established carcinoma are accompanied by leucocytosis. The great variety of these compli- cations render it unwise to draw any narrow diagnostic conclusions from the presence or absence of leucocytosis. The varieties of leucocytes in the blood in carcinoma are found in much the same proportions as in other forms of cachectic leucocytosis. With marked leucocytosis the polynuclear cells usually form a high percentage, 77-89 percent (Reinbach), 74-96 percent (Cabot). Yet in a case of the pernicious anemic type of gastric cancer Sailer and Taylor reported 45,000 leucocytes, of which 46 percent were mono- nuclear, and these authors found the large mononuclear cells to regu- larly exceed the small lymphocytes. Unusually high leucocytoses have been reported in the anemic type of gastric cancer by several other ob- servers, some instances of which apparently represent ante-mortem leu- cocytosis, while in others the excess was noted long before death. Eisenlohr, Mayer, and Lebert reported a ratio of one white to fifty red cells ; Potain, 1-48 ; Welch, 1-20. Braun has described a case of cancer of prostate with pernicious anemia (1 million red cells, 10,700 leucocytes) in which the majority of white cells were lymphocytes, but a few eosinophile myelocytes were present. Eosins usually persist in low normal proportions, from 1.2, 1.5, to 2 percent. (Cabot, Sailer, Taylor, Reinbach.) Myelocytes are found in a large proportion of cachectic cases, and sometimes in considerable numbers. They are most abundant in the ante-mortem leucocytoses, when the excess of various forms of leuco- cytes, with normoblasts and anemic red cells, may suggest leukemia. Sailer and Taylor found 9.3 percent of these cells in a very anemic case. Leucocytes holding fragments of red cells or showing various degen- erative changes are described by Hayem and Escherich. Specific Gravity. — The specific gravity of the blood in well estab- lished carcinomatous cachexia is remarkably low. This fact is illus- trated in Dieballa's series in which cases of gastric cancer with 1| to 2 million cells show the same gravity, 1.030-1.032, as cases of pernicious anemia with only .5 to 1 million cells. Peiper also found very low gravity in 4 advanced cases, while the lowest reports are those of Moraczewsky, 1.012 and 1.016. While Hammarschlag obtained nearly normal gravity for the serum in these cases, Grawitz, Strauer, and v. Jaksch found exceptionally low percentages of dry residue and albu- men, indicating that in carcinoma the albumens of the serum suffer in a peculiar degree. Biernacki located the marked loss of albumens SARCOMA. 373 principally in the red cells, but concluded that the iron is not always markedly diminished. In the less advanced cases of carcinoma the gravity of the blood does not differ from that of other secondary anemias. (Devoto, Schmaltz, Scholkoff, Hammarschlag, Stintzing, Gumprecht.) The alkalescence of the blood has been found considerably dimin- ished, as in other forms of secondary anemia. Klemperer and v. Lim- beck, Peiper and Rumpf, and Moraczewsky, using different methods, found very low grades of alkalescence in advanced cases, v. Noorden refers this change to the liberation of sulphuric, phosphoric, acetic, and oxybutyric acids, from the destruction of albumens. An excess of sugar was found by Freund in a series of cases, and while Trinkler confirmed these results, Matrai had previously shown that it is neither constant in, nor peculiar to, carcinoma. A peculiar form of coma, described by v. Jaksch ^ and others, some- times terminates the course of gastric cancer, and has been referred by Senator to an obscure toxemia resulting from disturbed metabolism. Herter's demonstration of a great excess of bases over acids, and a great diminution of ammonia, in the urine in chronic gastritis, strongly indi- cates that an acid intoxication exists in these cases, as in diabetic coma. (See Chronic Gastritis.) SARCOMA. The early changes in the blood in sarcoma closely resemble those of carcinoma, but some comparison of details will be of interest. Red Cells. — The usual variations in the progress of anemia are to be observed, including the absence of oligocythemia, or even the pres- ence of polycythemia in the early stages (Alexander, Reinbach, Cabot), and very slow or very rapid impoverishment of the blood according to the rate of growth and character of the complications. In the cases recorded by Hayem, Alexander, Laker, Limbeck, Sad- ler, Rieder, Reinbach, Bierfreund, Cabot, it is impossible to find any uniform difference in the grade of anemia from that observed with car- cinoma, and this conclusion accords with the writer's experience. Reinbach, however, finding only 4 normal blood specimens in 20 cases of sarcoma, and 6 among 16 cases of carcinoma, concluded that the anemia of sarcoma is usually greater than in carcinoma. There are also some pathological grounds on which Reinbach's con- clusions may be supported, in the special relation of sarcomatous growths to blood vessels and marrow, their greater tendency to hem- orrhage and degeneration, and their more frequent association with febrile processes, leucocytosis, or even leukemia. The lowest number of red cells observed appears to be that recorded by Hayem at 663,400, while Rieder found only 6 percent of Hb (?) in a dying subject. Limbeck says that normal blood is more fre- quently found with early sarcomata than with early carcinomata, while Reinbach has never seen in early carcinoma the extreme alterations of 374 MALIGNANT lUMORS. the blood which he referred to an early sarcoma. Alexander reports an extensive osteosarcoma of sternum and vertebrae with over 6 mil- lion red and 52,000 white cells, and in his and other series of osteo- sarcoma the average anemia is not extensive. In Cabot's 16 cases the averao-e of red cells was 4.4 millions, with extremes between 2.6 and 6.2 millions. Nucleated red cells are rather less common than in carcinoma. The chano-es of progressive pernicious anemia were present in the blood in cases of osteosarcoma reported by Grawitz, Ehrlich, Hosier and Gast, Fede, and Hausler. Leucocytes. — The relatively greater frequency and extent of leuco- cytosis in sarcoma appears to be the most striking difference in the blood between the two groups of malignant tumors. The great major- ity of sarcomata when first observed have shown a distinct leucocytosis, which moreover tends to persist and often to increase, with the advance of the tumor, till at death the upper limits of inflammatory leucocytosis are frequently reached. Martin and Matthewson have described leu- cocytosis of this extreme type and pointed out some of the difficulties in diagnosis which may result therefrom. Indeed, there are several recorded instances in which the lymphocy- tosis of sarcoma terminated in lymphatic leukemia. The writer observed such a case in 1899, and others have been reported by Palma, Sadler, and Strauss. Palma' s and Sadler's cases were round -celled sarcomata. In the writer's case various parts of the tumor and its metastases showed gradations be- tween round and large spindle-celled sarcoma. The recorded transforma- tions of pseudo-leukemia into leukemia are doubtless of similar significance. (Fleischer, Penzoldt, v. Jaksch,2 Westphal, Hosier, Senator.) The types of leucocytes show greater variety in sarcoma than in car- cinoma. The majority of leucocytoses being of inflammatory origin, an excess of polynuclear cells is the rule, and even with a normal number the proportion of these cells may be excessive. Lymphocytosis, though not relatively frequent with sarcomata, has at times reached the grade of lymphatic leukemia, as in the cases mentioned. In lym- phosarcoma, as in pseudo-leukemia, the lymphocytes are not usually excessive, and in many cases are very scanty. Eosinophilia has been regarded by Neusser as one of the diagnostic signs of sarcoma of bone marrow. A marked and persistent excess of these cells, 48 per- cent, was found in one of Reinbach's cases of lymphosarcoma compli- cated by phthisis, and bone metastases were found at autopsy. Three other cases, without autopsy, showed from 8 to 12 percent of eosins, but these cells were normal in most of his cases, and absent in five. There seems to be some diff'erence in the extent of leucocytosis ob- served in various types of sarcoma. In 16 cases of osteosarcoma col- lected by Cabot from several authors, the average was 17,000, in 12 lymphosarcomata 20,000, and in 7 melanotic sarcomata 25,100. With- out further details than the reports of these cases furnish the signifi- cance of these data is limited. The greater leucocytosis of melanotic BIBLIOGRAPHY. 375 sarcoma accords with its recognized malignancy, and some marked leucocytoses observed by Alexander and Limbeck (52,000, 32,000) in osteosarcomata may perhaps go with myeloid tumors, which are more malignant than the periosteal. Myelocytes were noted in Reinbach's case, cited above, which also showed eosinophilia. Cabot found 7 percent of large and small myelo- cytes in a case of general sarcomatosis, and an occasional myelocyte in three other cases. Bibliography. Cancer, Sarcoma. Alexander. These de Paris, 1887. AndraL Archiv gen. de Med., June, 1823. Bierfreund. Langenbeck's Archiv, Bd. 41, p. 1. Biernachi. Zeit. f. physiol. Chem., Bd. 19. Braun. Wien. med. Woch., 1896, pp. 482, 582. Baland. Fort. d. Med., 1891, No. 20. Behio. St. Petersb. med. Woch., 1891, p. 1. Bevoto. Zeit. f. Heilk., 1891, p. 176. Dieballa. Deut. Archiv klin. Med., Bd. 57, p. 302. Ehrlich. Charit^-Annalen, 1878, Bd. 5. Eichhorst. Spec. Path. u. Therap., Bd. II. Eisenlohr. Deut. Archiv klin. Med., Bd. 20, p. 495. Epstein. Wien. med. Presse, 1894, p. 2050. Escherieh. Berl. klin. Woch., 1884, p. 145. Fede. Cited by Ehrlich, Charit^-Annalen, Bd. 5. Fleischer, Penzoldt. Deut. Archiv klin. Med., Bd. 26, p. 368. Freund. Cong. f. inn. Med., 1889. Grawitz. 2 Virchow's Archiv, 1876, p. 323. Haherlin. Miinch. med. Woch., 1888, No. 22. Hammarschlag. Zeit. f. klin. Med., Bd. 21. Hampeln. Cited by Neubert. Haussler. Diss. Greifswald, cited by Muller. Herter. Trans. N. Y. Path. Soc, 1900. V. Jaksch. » Wien. med. Woch., 1883, pp. 473, 512. ^Zeit. f. klin. Med., Bd. 6, p. 423. Klemperer. Charit^-Annalen, Bd. 15, 1890. Laache. Die Anemie, 1883. Laker. Wien. med. Woch., 1886, Nos. 18-20. Lebert. Krankh. d. Magens. Tubingen, 1877, p. 481. Leichtenstern. lUnters. u. d. Hb-gehalt., etc., Leipsic, 1878. ^Ziemsenn's Handb., Bd. 8, p. 344. Martin, Matthewson. Brit. Med. Jour., Dec. 5, 1896. Matrai. Pest. med. chir. Presse, 1885. Mayer. Aerzt. Intelligensbl., 1870, No. 21. Moraczewsky. Virchow's Archiv, Bd. 114, p. 127. Mosler. Virchow's Archiv, Bd. 114, p. 461. Mosler, Gast. Deut. med. Woch., 1885. Mouisset. Eevue de Med., 1891, p. 885. F. Muller. Zeit. f. klin. Med., Bd. 16. Neubert. Inaug. Diss. Dorpat, 1899. St. Petersb. med. Woch., 1889, No. 32. V. Noorden. Lehrb. d. Path. d. Stoffwechsels., p. 461. Osterspey. Berl. klin. Woch., 1892, pp. 271, 308. Palma. Deut. med. Woch., 1892, p. 784. Patrigeon. These de Paris, 1877. Peiper. ^Cent. f. klin. Med., 1891, p. 217. « Virchow's Archiv, Bd. 116, p. 337. Potain. Gaz. d. Hop., 1888, p. 525. Beinbach. Langenbeck's Archiv, Bd. 46, p. 486. Bumpf. Cent. f. klin. Med., 1891, p. 441. Sadler. Fort. d. Med., 1892; Suppl. Sept., p. 38. Sailer, Taylor. Internat. Med. Mag., Vol. 6, p. 404. 376 MALIGNANT TUMORS. SchmaUz. Deut. Archiv klin. Med., Bd. 47, p. 145. Scholkoff. Diss. Bern, 1892. Senator. Berl. klin. Woch., 1882, No 19. Stintzing, Gumprecht. Deut. Archiv klin. Med., Bd. Strauer. Inaug. Diss. Greifswald, 1893. Strauss. Charit^-Annalen, 1899. TrinkUr. Cent. f. med. Wissen., 1890. Welch. Pepper's Syst. Med., Vol. 2, p. 552. Westphal. Diss. Greifswald, 1887. 53, p. 265. PART VI. ANIMAL PARASITES. CHAPTER XXIII. MALARIA. I. TECHNICS. (a) The Examination of Fresh Blood. — By far the readiest method of determining the presence of malarial infection is by the examination of the fresh blood. For this purpose thoroughly cleaned and polished slides and cover-glasses are the chief requisites, and a thin layer of blood may be secured by touching the exuded drop to the cover-glass, which is immediately laid upon the slide and examined with a one- twelfth immersion lens. The quantity of blood should be small, and in the resulting layer the red cells should be separated from each other. Besides the facility in preparation, this method permits the study of the vibratory motion of pigment, of ameboid motion of the parasite, of some limited phases of the natural development of the parasite, such as exflagellation, and of some artificial changes, such as the escape of the parasite from the cell, and occasionally of the forma- tion of vacuoles. By mixing the blood with ascitic fluid deeply tinged with methylene- blue, Celli and Guarnieri secured excellent stained specimens of the parasite in the fresh condition. The ascitic fluid was prepared by dis- solving powdered methylene-blue in the fluid and filtering. The blood was mixed with this fluid and spread under a cover-glass in the usual way. Their drawings of parasites seen under these conditions appear to have been the first definite representation of the nuclear body of the malarial organism. No other notable modification of the method of examining malarial blood in the fresh condition appears to have been perfected. Concerning the value and reliability of these methods opinions are somewhat at variance. Most of the early study of malarial blood was made exclusively by this method, and it still remains, of course, the only method permitting the minute study of many important changes in the form of the parasite, of the motility of the pigment and of some degenerative processes. On the other hand, exclusive reliance on this method appears to have been responsible for much confusion regarding the varieties of the parasite, and its minute structure, as well as for 378 MALARIA. many erroneous interpretations of exceptional or artificial changes in the organism. As a ready means of diagnosis, in the hands of an expert, the ex- amination of fresh blood is all that is required in the average case, and when the parasites are moderately numerous and when crescents are present, even an inexperienced observer can hardly err. Under any other conditions, except for special purposes, reliance upon this method appears to the writer inadvisable. When the parasites are scarce, especially when they are of the small unpigmented form, a prolonged search through fresh blood has frequently proven negative, in the writer's experience, although a few minutes sufficed for the discovery of one or more minute parasites in the stained specimen. Practically, this fact demands that a negative result with a fresh specimen be in- variably controlled by the examination, with mechanical stage, of a stained specimen. Moreover, any one who is familiar with the host of appearances in fresh blood simulating the malarial parasite must regard with some suspicion the report, from anyone but a recognized expert, of the discovery of one or two " hyaline bodies." From these consid- erations the writer would limit the use of fresh specimens to the study of a few special features of malarial parasites, and would urgently rec- ommend that as a diagnostic procedure reliance be placed only upon the examination of dry specimens stained, preferably, by Nocht's method. The study of flagellate bodies may be conducted in fresh specimens prepared in the ordinary way, but placed, if possible, on a warm stage. Mannaberg: alone claims to have found the flaorellate bodies in consid- erable numbers in the fresh human blood immediately after shedding. Usually they appear only after the lapse of 10-20 minutes, when they may form in variable numbers from the larger tertian or crescentic estivo-autumnal organisms, less frequently from quartan parasites. The addition of a little water or salt solution may facilitate the es- cape of the parasite from the red cell and the formation of flagella. The successful action of moisture has been obtained by several ex- pedients. Marshall added about an equal quantity of water to a small drop of blood containing many crescents and saw the almost immediate change of crescents to spheroidal bodies, followed by exflagellation. Manson recommends that the blood under the cover-glass be kept moist by exposure to steam exhaled from a hot moist sponge. After a few minutes the cover may be carefully removed, the specimen dried, and the flagellate bodies stained. The writer finds that the moist chamber may be secured in a Petri dish with tightly fitting vaselined cover. Wet blotting paper placed in the dish furnishes the necessary moisture. Specimens spread on slides or covers may be kept moist for 10-20 minutes in such dishes, and flagellation proceeds with moderate rapidity. A simple method is as follows : Cut an opening | x 1 inch in a piece of thick blotting paper and moisten the paper in hot water. Spread two glass slides rather thickly with fresh blood, lay the blotting paper STAINING METHODS. 379 ou one slide, cover the cut opening by the other, specimen side down, and slip a rubber band about both. After 15—20 minutes the slides and paper may be separated and the two specimens dried. (b) Preparation of Dry Specimens. — Smears of the blood may be prepared by any of the usual methods. After drying thoroughly in the air, the malarial parasite stains best after fixation in 95-97-per- cent alcohol for 15—30 minutes. Five minutes fixation will suffice, but if left in alcohol over night the staining qualities of the organism will sometimes be found slightly altered. When slides have been kept in alcohol, or in the air, after fixation by heat or alcohol, for periods varying from one month to three years, there is a progressive loss of the capacity to unite with eosin, while methylene-blue gives a deep diffuse stain. This tendency may be gradually corrected by passing the slides rapidly through a one-tenth-percent solution of nitric acid. Fixation by heat is less satisfactory for the present purpose, and the addition of ether to the alcohol secures no advantage. In a critical study of the malarial parasite full reliance cannot be placed on specimens that have been fixed longer than a few weeks, after which period there are often distinct alterations in the staining charac- ters of the organism. Staining Methods. Eosin and Methylene -Blue. — For ordinary purposes, staining by eosin and methylene-blue may be generally recom- mended, and many of the detailed studies of the parasite have been based upon specimens stained in this way. The solutions required are : (1) a saturated alcoholic solution of Ehrlich's blood eosin, diluted with an equal quantity of 95-percent alcohol, and (2) a saturated watery solution of Ehrlich's rectified methylene-blue, at least one week old. A light staining by eosin, such as is given by the diluted solution of eosin, is essential for the clear demonstration of the parasite by methylene-blue, and in specimens containing only the small signet- ring forms, heavy staining by eosin may almost entirely prevent the subsequent action of methylene-blue, and these minute parasites may be overlooked. By intense staining with old solutions of alcoholic eosin which have absorbed water, fragments of the red cell probably containing traces of Hb may often be demonstrated in the meshes of the parasite, of the presence of which no indication is given by the usual method. Methylene-blue fails to stain the young ring forms, especially of the tertian type, as clearly as is desirable, and more powerful basic stain- ing fluids may well be employed for this purpose. Nocht's method may be recommended over any other, as it facilitates the identification of the sniall rings by means of a large densely stained nucleus, but when this method cannot be employed, one may resort to the method modified by Futcher and Lazear from the suggestions of Benario and of Marchoux, as follows : "Fix the specimens 5 minutes in 95-percent alcohol, to 100 cc. of which have been added 1 cc. of formalin. Stain 1—3 minutes in the 380 MALARIA. following mixture : Saturated alcohol solution thionin, 20 cc; 20-per- cent carbolic acid, 100 cc. The fixing solution must be used fresh, and the staining fluid must be at least one week old. The rings are then densely stained and the specimens do not fade." The sharpest demonstrations of minute ring-shaped parasites secured by the writer were obtained by staining one hour in diluted Gage's hematoxylon before treatment by eosin and methylene-blue. Hema- toxylon stains the nucleus of the ring and makes the body of the para- site blacker after methylene-blue. Such specimens are specially suit- able for photography. The Nocht-Romanowsky Method. — The method originally devised by Romanowsky, and which has given such uncertain results in the hands of many investigators, has now been modified most successfully by Nocht, whose procedure gives positive results without much dependence upon the quality of dyes or the time of staining. (The interesting history of this stain may be found in the writer's article on Malarial Parasitology, Journal of Experimental Medicine, 1901.) Nocht's modification consists in the addition of a few drops of neutralized Unna's polychrome methylene-blue (Griibler) to the 1-per- cent solution of ordinary methylene-blue. The usual specimen of poly- chrome methylene-blue is distinctly alkaline, and to be rendered effectual for the present purpose Nocht found that it requires neutralization, preferably by acetic acid. This may be done by adding drop by drop of dilute 2-3-percent acetic acid till the commercial fluid polychrome blue no longer turns red litmus blue above the zone coming into im- mediate contact with the dye. The writer has never failed to secure a good result by the following procedure : 1. To 1 ounce of polychrome methylene-blue (Griibler) add 5 drops of 3-percent solution of acetic acid (U. S. P., 33 percent). 2. Make a saturated 1 -percent watery solution of methylene blue, preferably Ehrlich's (Griibler) or Koch's, dissolving the dye by gentle heat. This solution improves with age, and should be at least one week old. 3. Make a 1 -percent watery solution of Griibler's watery eosin. The mixture is prepared as follows : To 10 cc. of water add four drops of the eosin solution, 6 drops of neutralized polychrome-blue, and 2 drops of 1 -percent methylene- blue, mixing well. The specimens, fixed in alcohol, or by heat, are immersed, specimen side down, for 1 to 2 hours, and will not overstain in 24 hours. The density of the blue stain may be varied to suit in- dividual preferences. The above proportions need not be rigidly fol- lowed, but the polychrome solution should be accurately neutralized, and the staining mixture should be deep blue. Nocht later reports that the two solutions of methylene-blue may be replaced by a 1-percent solution of Ehrlich's methylene-blue, alkalin- ized by J percent of NaOH and kept a few days in a thermostat at 50° C. This is the ordinary laboratory method of improvising polychrome methylene-blue. To 2 cc. of water add 2-3 drops of 1- THE TERTIAN PARASITE. 381 percent watery eosin, and drop by drop of the alkalinized methylene- blue till the original red color of the eosin has almost disappeared. In this fluid specimens stain in 5-10 minutes. This method is less reliable than the former. The rationale of the Nocht-Romanowsky method is not yet fully under- stood, but it appears most probable that a staining agent which unites selec- tively with chi'omatin exists ready- formed in polychrome methylene-blue, and may be developed in specimens of methylene-blue in various ways, among which is slow digestion with an alkali and heat. Nocht refers to this principle as " red from methylene-blue." It is not the commercial methy- lene-red, but may be extracted from polychrome methylene-blue, etc., by chloroform (Nocht), and it is a reasonable expectation that it can be put on the market in pure form. Goldhorn has recently succeeded in digesting methylene-blue with satu- rated solution of lithium carbonate so as to develop in it a large proportion of the red chromatin-staining principle. This fluid, neutralized by acetic acid, not only stains the chromatin rapidly (15-60 seconds), but demonstrates, l)etter than has yet been done, early and extreme granular degeneration of the infected and other red cells. Goldhorn's fluid, ready for use, can be obtained from New York dealers. (See Trans. N. Y. Path. Soc, February, 1901.) Nocht's method furnishes so much information regarding the minute structure of the parasite, and renders its identification so complete and positive, that it must be recommended above all other methods of staining the malarial parasite. Moreover, it has a large field of ap- plication in the study of nuclear structures in various other micro- organisms. II. MORPHOLOGY. The Tertian Parasite. 1. The youngest form of the tertian parasite seen in the red cell is identical in appearance with the spore of the parent rosette. It is a compact spheroidal or slightly oval, or irregular body, about 2 // in diameter. It shows an outer rim of basophilic protoplasm inclosing a single large nuclear body which is achromatic to methylene-blue but stains readily in hematoxylon or by Nocht's method, and which is usu- ally inclosed or accompanied by a clear achromatic substance, termed by Gautier " the milky zone." (Plate XI., Fig. 1.) In the fresh condition these bodies are noticeably refractive, espe- cially the nucleus, change their position but rarely their shape, and are never pigmented. From the earliest period of infection the red cell is often swollen. 2. The Young Tertian Ring. — Within a few hours after the chill the parasite is usually found to have assumed a somewhat characteristic ring shape which it commonly maintains in some definite form up to the pre-segmenting stage. These bodies measure from 3-4 /i in diameter, and the regular ring form is retained without marked increase in bulk at any point, for 6-8 hours. Sometimes the ring is elongated, one arm reaching across the 382 MALABIA. cell, while a thin bow persists. Occasionally the ring appears to un- fold, and the parasite stretches clear across the swollen cell, with the nucleus at one end. The tertian ring is rarely as geometrical or deli- cate as the estivo-autumnal signet-ring. The development of pigment is inconstant, some large rings failing to show pigment, but usually one or more fine grains are to be seen in the medium-sized and smaller forms. The ring always encloses a considerable mass of hemoglobin. The nuclear body of the tertian ring is its most characteristic feature, appearing as a rather large, achromatic, highly refractive body, after methylene-blue, but staining intensely with hematoxylon and by Nocht's method. Significance of the Ring Form. — In regard to the formation and signifi- cance of the ring opinions are at variance. Most of the Italian writers hold that the ring form is not really a ring, being bridged across by a transparent and vesicular nucleus. There are many considerations favoring this view, especially the usual appearance of the parasite in the fresh condition, and the fact that the chromatin usually lies within the ring, eventually filling it. On the other hand, Mannaberg and Ziemauu claim that this body is a true ring, formed by the thinning of the central portion of the body of the para- site, and state they have seen the ring develop in this manner in the fresh blood. From the examination of the rings themselves the writer has been unable to convince himself as to which view is correct, but there are some early forms of the parasite which strongly indicate that the ring does not represent a vesicular nucleus. In one such form the ring isunfolded and the nuclear body of the parasite lies naked in the hemoglobin. Moreover, elon- gated forms of the young parasite are often seen in which the ring is absent and the nuclear body lies bare at one end. These forms vividly recall the appearance of the ameba dysenterise in which the nucleus remains at the hinder end during active movements of progression. Further, it is difficult to associate the relatively huge size of the ring with any nuclear structure, which would require the young malarial parasite to have a nucleus which is larger than that of the adwlt cuneba dysenterix. Again, secondary rings some- times form from the union of pseudopodia, and these are identical iu appear- ance with the primary ring, but lack the chromatin granules. In specimens stained by Nocht's method the chromatin is usually found within the ring, sometimes lying in an isolated position in the center, but very often the chromatin is found outside of the ring, connected by a very fine thread of protoplasm. If the ring represents a vesicular nucleus, we have here the anomaly of a complete separation of chromatin from the vesicular portion of the nucleus, which is opposed to some rigid histological principles. Even more frequently the chromatin is found to be inclosed by bluish staining protoplasm which shuts it off" entirely from the ring. (Plate XL, Fig. 7.) From various biological studies it appears that the nuclei of the protozoa are usually widely different from the nuclei of metazoa. Many protozoa do not have a vesicular nucleus, with cell membrane, linin, nucleolus, etc., but possesses the so-called " distributed nucleus'''' composed of a number of gran- ules lying free in the body of the parasite. The study of the malarial para- site by Nocht's method indicates that the nucleus of this protozoan is of the distributed type, which does not exhibit a vesicular appearance nor possess a nuclear membrane. (See Calkins on "Protozoan Nuclei," Annals N. Y. Acad. Sciences, Vol. IX., Part III., 1898.) On these grounds the writer is inclined to agree with Mannaberg and others who hold that the form in question is a true ring, a form usually but not necessarily assumed by the parasite, and does not represent a vesicular nucleus. PLATE X Developmental Cycle of Benign Tertian Parasite. Fig. I. Veryearly form of parasite, showing chromatin granule, "milky zone," and spheroidal body. Figs. 2, 3. Typical young ring-shaped parasites. Figs. 4, 5. Subdivision of chromatin, development of body, and appearance of pigment in later ring-forms. Fig. 6. Double rings, in single parasite. Figs. 7,8. Turban-sliaped parasites. Secondary rings, eccentric jiosition of chromatin. Fig. 9. Double infection of cell. Figs. 10, II. Complex ameboid figures in doubly infected cells. Fig. 12. Full grown form, with large eccentric nucleus. Figs. 13, 14. Protrusion of chromatin granules and milky substance in body of full grown parasite. Figs. 15, 16. Division of chromatin granules into groups in reticulated presegmenting bodies. Fig. 17. Tertian rosette. THE TERTIAN PARASITE. 383 Marchiafava and Bignami still hold that the ring is bridged over by an invisible membrane. They do not insist, as do many, that this membrane is a nuclear structure, but consider the central area to be a digestive vacuole. Comparison of the Tertian and Estivo-autumnal Rings. — From the study of the ring-shaped tertian parasite and the estivo-autumnal sig- net-ring forms, in typical cases of these infections, the writer believes that these parasites, with very rare exceptions, can be fully distin- guished from each other in this early stage, and on the following pe- culiarities : 1. The nuclear body and chromatin mass of the young tertian para- site is achromatic to methylene-blue, which densely stains the nucleus of the estivo-autumnal organism. The writer has been unable to find in the literature any specific reference to this diagnostic peculiarity, which may be readily verified by comparing specimens of the two parasites stained by methylene-blue and by Nocht's method. (The dense staining of a nuclear body in the young estivo-autumnal para- site has often been noted, and in 1894 Okintschitz mentioned the fact that the nucleus of the young tertian parasite fails to stain by methy- lene-blue.) 2. The shape and contour of the tertian ring is usually coarse and irregular, but the estivo-autumnal ring is geometrically circular, more delicate, with an extremely fine bow, and usually with a typical signet- like swelling. (Cf. Plate XI., Figs. 2, 3, and Plate XIII., Figs. 6, 7.) 3. One or two grains of pigment are almost invariably found in the early tertian ring, but are, with nearly equal constancy, absent from the estivo-autumnal signet ring. 4. The writer's specimens confirm the statement of Gautier that the tertian ring is usually pigmented before the chromatin becomes subdi- vided, while the chromatin of the estivo-autumnal ring is always sub- divided before the appearance of pigment. In some cases, however, the chromatin of the tertian ring divides before pigmentation. 5. The infected cell is usually swollen from the moment of infection by the tertian spore, and commonly shrunken w4ien harboring the estivo-autumnal ring. All of these characters are usually apparent in ordinary specimens, but naturally are most distinct in flatly spread and rapidly dried cells. The writer has met with no exception to the above rules in cases infected by the tertian parasite in New York and cases of estivo-autumnal infec- tion from Cuba. In many of the irregular relapses, in cases showing tertian organisms, encountered among volunteer soldiers long returned from Cuba single ring-shaped parasites not admitting of positive iden- tification have sometimes been seen. The significance of this observa- tion will be considered later. (See section on Plurality of Species among malarial parasites.) 3, Large Tertian Rings. — After a period of 6-8 hours, the tertian ring is usually found to have developed an outgrowth which is actively ameboid in the fresh condition and appears in stained specimens as a tongue-like protrusion or turban-shaped mass attached to one seg- 384 MALARIA. ment of the ring. (See Plate XI.) The nuclear body meanwhile increases slightly in size, projecting into the ring, and the chromatin divides into several large granules. At this period occurs the greatest ameboid activity of the parasite and in some severe tertian infections the organism may be found fixed in the height of its ameboid excursions, when it presents in stained specimens the peculiar appearance depicted in Plate XI., Figs. 9, 10. There the ring persists but the body of the parasite is strung out into a number of slender threads with nodal thickenings. At times, especially in cases taking quinine, the number and delicacy of the threads greatly exceeds those seen in the sketch, which represents an average instance. A close inspection of cells harboring such parasites may disclose the presence in one cell of two distinct nuclear bodies, indicating the co- existence of two parasites. Frequently twin parasites entirely separate from the other are found in the same cell, each of which shows a ten- dency to develop the long threads. (See Plate XI., Fig. 9.) When, however, the threads are numerous and very thin, it is usually impos- sible to find any break in their continuity, while in many instances the two parasites are distinctly united. The significance of these forms will be considered later. (^See Conjugation,) 4. Tertian Spheroidal Bodies. — During the second quarter of the cycle, the body and the nucleus of the parasite develop rapidly in size, ameboid motion and ameboid figures gradually diminish, and pigment is abundantly deposited in the form of fine dark brown or yellowish grains showing in the fresh state active vibratory motion. The in- fected cell continues to increase slightly in size, and its hemoglobin is progressively diminished. Depending upon the character of ameboid activity the variety of figures seen during this period is very great. Eventually, toward the end of 24 hours or possibly somewhat later, the parasite occupies three-fourths of the swollen cell, in the form of a spheroidal or ellip- tical, homogeneous body, the outer portion of which contains most of the pigment and is rather more deeply stainable than the zone imme- diately surrounding the nucleus. (Plate XI., Fig. 12.) The nucleus gradually increases in size, growing into the ring. It no longer has the appearance of a small highly refractive achromatic spot (after methylene-blue) but takes a light bluish tinge with 1-percent methy- lene-blue, and stains less deeply than before with hematoxylon. At the end of this period the nucleus completely fills the ring, stains rather distinctly with methylene-blue and sometimes exhibits a deli- cate bluish network. After Nocht's method the changes in the nucleus are found to consist in the gradual subdivision of the chromatin granules, which finally become rather numerous, of minute size, and more difficult to stain. Usually these chromatin granules lie on the inner circumference of the bow of the ring, projecting within the ring, and partly surrounded by a " milky " unstained zone. This milky zone is often absent in young para- THE TERTIAN PARASITE. 385 sites in cells thinly spread and rapidly dried, but in older parasites it is al- ways present. Various other positions may be assumed by the chromatin mass, as follows : A subdivision of the granules into two distinct groups, separated by a strand of bluish stained protoplasm ; an excentric position entirely apart from the ring ; a position midway between two rings found in the same par- asite ; a position in the center of the ring entirely apart from any bluish protoplasm ; a circular arrangement about the periphery of the ring. Some- times the smaller granules are grouped about a central larger granule, as has been noted in other protozoa whose nuclei are of the intermediate type. (Microglena. Euglena.) (Plates XI., XIII.) 5. Full Grown Tertian Parasites. — The third quarter is occupied bj the continued growth of the parasite in the form of a large homogene- ous richly pigmented body which finally occupies at least four-fifths of the swollen cell, and by certain nuclear changes which it is diffi- cult to follow in specimens stained by methylene-blue or hematoxy- lon, but are fully demonstrated by Nocht's method. The exact limits within which the parasite may be termed " full grown " can be sharply fixed only with great difficulty, but there appears to be a period of at least twelve hours, during which there is little change in the structure of the organism and during which the body stains homogeneously and the nucleus occupies the entire ring. The writer would place this period between the 24th and 40th hour of the cycle. A portion of the period is occupied by nuclear changes belonging to the reproduc- tive phase of the parasite's development. After the appearance of a faint intranuclear network most authori- ties agree that the nucleus largely disappears in specimens stained by methylene-blue or hematoxylon, and is next seen in the form of highly refractive achromatic spots in the meshes of the reticulated pre-seg- menting body (methylene-blue), and these again stain deeply with hem- atoxylon. Nocht's method, however, fully demonstrates the nuclear changes which occur in the full-grown parasite. A considerable area, usually the entire original ring, is now occupied by a " milky " or slightly bluish staining sub- stance, in which lie a considerable number of very fine chromatin granules. These granules are usually difficult to stain, and being of very minute size they are difficult to see. This fact has led Ziemann and Gautier to admit the possibility that the chromatin may actually disappear at one stage of the development, especially since they have found some large parasites in which no chromatin was demonstrable. In the writer's specimens there were a very few large tertian parasites in which no chromatin granules appeared, but these were not more numerous than younger forms which were devoid of chromatin and which must therefore be regarded as sterile. Similarly, the larger forms devoid of chromatin the writer would class with the sterile forms rather than accept the view that the chromatin may entirely disappear at one stage of the fertile parasite, a view which is at variance with biological principles. After the subdivision of the chromatin has reached a limit the next change, observed in a considerable number of parasites, appears to con- sist in the extrusion of a portion of the milky substance and its chromatin granules into the body of the parasite. (Plate XI. , Fig. 13.) At the same time the granules of chromatin increase in size and diminish in number. Other forms may be seen in which the ' ' milky substance ' ' and chromatin granules occupy an elongated space within the body of the parasite, and in such 25 386 MALARIA. cases the beginning concentration of pigment and deeper stain of the para- site indicate the presence of the pre-segmental stage. (Plate XI., Fig. 14.) 6. Pre-segmenting bodies usually begin to appear in the blood 8-10 hours before the chill. In specimens stained by methylene-blue the first demonstrable indi- cations of the division of the parasite are seen in a deeper staining capacity and tendency toward reticulation which appear throughout the whole or in a part of the body of the parasite. Occasionally these changes may be noted in one half the parasite, while the other half retains the homogeneous appearance of the " full grown " organism. Usually the process is found to have affected the entire organism, giv- ing the very characteristic forms sketched in Plate XI., Figs. 1 5, 1 6. In the pre-segmenting bodies the pigment is gathered in a reduced number of coarse grains or spindles which lie in the body of the parasite, in a position determined by that of the new multiple nuclei. These bodies were first described by Golgi in fresh blood and properly in- terpreted as belonging to the process of segmentation. Later they were de- scribed by Marchiafava and Celli as vacuolated parasites, the highly refrac- tive nuclear bodies, appearing in the fresh condition very much like vacuoles. Still later, Celli and Guarnieri sketched them from specimens stained in the fresh condition, regarding some as showing partial segmentation, others as vacuolated parasites, although they accurately described the appearance of the nuclear particles invariably found within these " vacuoles," while others they supposed to be groups of confluent parasites, i. e., true plasmodia. Mannaberg's descriptions (1899), referring only to fresh blood, do not include these bodies, nor have they found a distinct place in his plates, although some of the figures in Plate IV. indicate that they have not escaped his ob- servation. In Thayer and Hewetson's careful study of the parasite in fresh blood, the pre-segmenting bodies are not described as such. Laveran (1898, page 62) refers to the similarity in appearance between a nuclear body and "vacuole," but he neither describes nor depicts the pre-segmenting reticu- lated parasite. Ziemann describes the pre-segmenting bodies as they appear after his or Romanowsky's staining methods, but the plates would not enable one unfamiliar with the subject to identify these forms in specimens stained by eosin and methylene-blue. The reticulated pre-segmenting tertian parasites may be seen in every case examined within 6-8 hours preceding the chill, and often in be- lated parasites shortly after the chill. Many transitional stages be- tween the homogeneous adult parasite and the perfect rosette may be seen in rich infections. They are well demonstrated by eosin and methylene-blue, especially as regards the increasing density of stain and reticulation. After hematoxylin the multiple nuclei stain deeply. By Nocht's method a series of interesting nuclear changes may be fol- lowed in the pre-segmenting forms. After the mass of enlarging chromatin granules and milky substance has flown out into the elongated form de- scribed above, the chromatin granules leave the central clear space and make their way in groups out into the body of the parasite. Various stages of this process may be followed in specimens taken at suitable periods, and some ob- served phases are seen in Plate XI. , Figs. 13-15. Considerable differences in the number of such groups may be noted in different cases. Usually a large number of ill-defined groups are seen, before the central mass of granules is THE TERTIAN PARASITE. 387 exhausted. (Plate XIII., Fig. 14.) In some specimens the compact nuclei of the young spores appear to form in one segment of the parasite before the main mass of granules has become exhausted. Each of the groups appears always to be surrounded by a milky zone, and the mass of granules is often of a peculiar triangular form. During these changes the pigment granules increase in size, diminish in number, and are distributed in the meshes of the now distinctly reticulated body of the parasite. 7. Tertian rosettes are usually seen in the circulation 3 or 4 hours be- fore the chill, most abundantly just before the chill, and a few are often to be found for one hour or longer after the chill. These limits may occasionally be much wider, as Marchiafava and Celli have seen rosettes two to six hours before the chill and six to seven hours thereafter, and indeed, when the different broods of para- sites are not very distinct, there is no reason why occasional rosettes should not be found at any period of the main cycle. Of the three types of sporulation described by Golgi, the second type, ac- cording to which the entire parasite is divided into spores leaving nothing but pigment, is undoubtedly the usual process. As regards Golgi's first type, after which only the peripheral portion of the parasite divides, leaving a distinct central globular pigmented body, most stained specimens fail to show convincing evidence that the physiological segmenting process may be subject to such an important modification, nor does it appear in recent litera- ture that the existence of this variety of segmentation has been fully veri- fied. In a few cases taking quinine, the writer has seen rare segmenting bodies which resembled those described as above by Golgi, but never in fresh cases. Golgi's third type of "partial segmentation," together with the lateral circumscribed sporulation of Celli and Guarnieri, may frequently be seen in rich tertian infections in fresh blood, but according to the evidence of stained specimens, must be classed with the pre-segmeuting forms. The various morphological differences seen in the segmenting forms of the separate species serve to distinguish those species, but do not constitute different types of this process. The tertian rosette is usually distinguished by its large size and considerable number of spores, 15-20. From the writer's specimens it does not appear, however, that the identification of the tertian rosette can always be based upon the number of spores, as these are sometimes found to number under fifteen, though rarely over twenty- one. Marchiafava and Bignami, however, have described tertian rosettes with 40—50 spores. With the smaller number of spores the rosette is always distinctly larger than either the quartan or the estivo-autumnal body. Although rosettes sometimes appear to be extra-cellular, when seen in fresh blood, in stained specimens the writer has never seen a clearly extra-cellular segmenting tertian body. Almost invariably there is an unbroken ring of hemoglobin about the parasite, and very often traces of hemoglobin may be found scattered through amass of spores, where they may be demonstrated by dense staining with epsin. The nuclear changes demonstrated by Nocht's method in the tertian rosette consist principally in the gradual fusion of the new formed groups of chromatin granules into one compact globule, which is partly surrounded by a '* milky zone." 388 MALARIA. While the rosette is still compact the vesicular shape of the spore is distinct. The outer segment of the ring is usually thickened, the nucleus tends to lie near the inner pole, and between the nucleus and outer segment is a small milky zone. The pigment is usually collected into a central block or mass of granules, but may be found variously scattered among the spores, or along the periphery of the rosette. (Plate XI., Fig. 17.) The Quartan Parasite. The earliest form of the quartan parasite as seen in the stained red cell is practically indistinguishable from that of the tertian organism, but its true character may usually be suspected from the slightly shrunken appearance of the infected cell. In fresh specimens the higher refracted quality of this body is often, however, sufficiently characteristic for its identification. After a very slight increase in size the quartan parasite becomes rather easy to distinguish in both fresh and stained specimens, for it early takes the form of a ring, of the gen- eral character of the tertian ring, but smaller, more compact, and more richly and coarsely pigmented. As with the tertian organism the nuclear body is found projecting into the ring. In fresh specimens at this period, the higher refractive quality and slower ameboid motion are additional diagnostic characters from the tertian parasite. (Plate XII.) The growth of the quartan ring is very similar in all important re- spects to that of the tertian, while its distinguishing features, especially the abundance of coarse pigment grains, are uniformly retained. During the pre-segmenting stage the characters of the quartan para- site are markedly different from those of the tertian. On account of the slower progress of sporulation and from the greater tendency of the quartan parasite to complete its cycle in the general circulation, quartan pre-segraenting bodies are relatively much more numerous in the stained specimen than are the similar forms of the tertian organism. In some specimens taken several hours before the chill the majority of organisms found may present the markedly reticulated structure indic- ative of approaching division. The multiple nuclei being less numer- ous and the pigment more abundant, the meshes of the reticulum are much coarser and the pigment is often found in irregular partly radi- ating rows. These coarsely reticulated, relatively small, and richly pig- mented bodies, lying in markedly shrunken cells, are very character- istic and not readily confused with any other form of malarial parasite commonly found in the peripheral blood. In some severe estivo- autumnal infections, showing many parasites of all stages in the peripheral blood, somewhat similar spheroidal or pre-segmenting forms may be found in considerable numbers, but, as will be seen by reference to Plate XIII., the character of the pigment in the estivo-autumnal parasite is very different, while such cases are very rare, and readily recognized on clinical grounds, being almost invariably of the perni- cious type. PLATE XII Cycle of Quartan Parasite. Fig. I. Very early non-pigmented form. Figs. 2, 3, 4. Small quartan rings, with large chromatin masses and abundance of pigment. Fig. 5. Turban-shaped ring, with subdivided chromatin. Fig. 6. Subdivision of ring and of chromatin granules. Fig. 7. Coarse quartan ring with central chromatin granules. Fig. 8. Full-grown c[uartan parasite, with eccentric chromatin, hyaline body, and abundance of pigment. Fig. g. Extra-cellular reticulated body. Figs. 10-13. Quartan presegmenting forms. Fig. 14. Quartan rosette. Fig. 15. Pigmented mononuclear leucocyte. THE ESTIVO-AUTTJMNAL PARASITE. 389 The quartan segmenting bodies are usually more abundant in the peripheral circulation than are rosettes of any other type, and are read- ily identified by the small number (6-12) and comparatively large size and geometrical arrangement of the spores. (Plate XII.) The Estivo- Autumnal Parasite. The following description applies to a group of organisms, which, according to the Italian school, comprises two or three varieties of ma- larial parasites. Waiving for the present the question of a plurality of species, the entire group will be described as one, and the grounds for their separation into two or more species v/ill be considered later. The earliest form of the estivo-autumnal parasite seen in the red cells is very similar to that of the tertian and quartan parasites, but is of slightly smaller size than either, and is often distinguishable from the tertian by the shrinkage of the cell, and from the quartan by its dis- tinctly smaller dimensions. (Plate XIII., Fig. 1.) In fresh speci- mens the young ameboid body usually shows a low refractive index as compared with the tertian and quartan parasites. It is never pig- mented. Associated with the inter-cellular spores there are frequently seen in the plasma small spheroidal bodies exhibiting an active roll- ing motion and occasional blunt projecting points (pseudopodia ?), and which on becoming arrested by contact with the red cells are found to be indistinguishable from compact intra-cellular bodies. The positive identification of these extra-cellular bodies, however, appears to the writer a very hazardous undertaking. (Cf. Ziemann, p. 49.) In dried specimens stained by Nocht's method, however, the young extra-cellu- lar parasite may be positively identified from the presence of a mass of chromatin. In the writer's specimens such extra-cellular bodies were very rarely encountered. The Estivo-autumnal Signet-ring. — At a very early period of its de- velopment the estivo-autumnal parasite assumes a very characteristic ring shape. Many of these rings early develop a thickening of one segment, and to these bodies of various sizes the term " signet-ring " very aptly applies. (Plate XIII., Figs. 5—7.) It is to be noted that in some cases the rings fail to exhibit this thickening but remain of a uniform but very fine caliber throughout. (Plate XIII., Fig. 4.) The period during which the rings retain this uniform caliber has not been determined, but bodies of this type may be seen measuring at least 3 /i in diameter. They nearly always pre- sent two nuclear bodies, lying at opposite poles or close together. Oc- casionally such rings are found to have unfolded, and to be stretched like a thread clear across the cell, the nuclei appearing at inconstant intervals. In other cases no rings of this type are seen, all showing the thick- ening of the signet and a single nuclear body. In the majority of cases rings of both types are associated in variable numbers. Multiple infection with the young rings is very common in the 390 3fALARIA. red cells of most cases of estivo-autumnal malaria, and, as a rule, its frequency is proportionate to the severity of the disease. In the periph- eral blood three parasites are often found in the same red cell, and occa- sionally four, while in smears of the marrow of a fatal case, seen by the writer, infection of one cell with four rings was common, five par- asites were occasionally seen in the same cell, and one slightly swollen red cell was encountered containing seven well-formed rings. These observations accord with the reports of Ziemann (p. 49), who found often three and four parasites, and once as many as five in one cell. It appears in the description of the hemameba immaculata, which is said to sporulate without producing pigment, that most of the rosettes contain comparatively few spores, averaging from 6 to 10. (Marchiafava, Bignami, Ziemann, Marchoux, Grassi and Feletti.) The close resemblance to a non- pigmented rosette presented by some of these red cells harboring 5, 6, or 7 young parasites is very striking. In the writer's specimens (see Plate XIII,, Fig. 2) there could be no doubt as to the proper interpretation to be placed on these examples of multiple infection. Multiple infection of the red cell appears in rather rare instances to lead to the development of a peculiar form of the young estivo-au- tumnal parasite on which Mannaberg bases his unique theory of the development of crescents. This body consists in the apparent union of two rings by a fusion of their nuclear bodies. (See Plate XIII., Fig. 3.) Mannaberg depicts all transition forms between these bodies and the fully developed crescent. The signet-ring forms frequently reach a diameter of 4 /^ while still retaining the peculiar thickening of one segment, the thin geometrical bow, and a very distinct nuclear body staining with methylene-blue and surrounded by a narrow achromatic zone. (Plate XIII., Fig. 6.) The distinguishing features between the estivo-autumnal and the tertian rings have been enumerated under the description of the latter parasite. Beyond this size, when persisting in the finger blood, the growth of the parasite produces an irregular body in which the outline of the ring becomes more or less obscure. The full development of the large signet-rings appears to require about 24 hours. In the majority of ca.ses the ring forms seen in the peripheral blood fail to show any trace of pigment, especially in the patients showing distinctly intermittent quotidian or tertian paroxysms. In a consider- able number of instances, however, especially in very severe and fatal infections, the largest rings exhibit a very few minute pigment grains, usually associated with older pigmented forms. Later Stages of the Estivo-autumnal Parasite. — The later forms of the estivo-autumnal parasite are rather rarely seen in the peripheral circulation. Most of the Italian writers speak of their occurrence in the finger Mood as being very unusual but not unknown. SaccharofF in two cases of estivo- autumnal infection saw many rosettes in the peripheral blood. Ziemann reports that in malignant tertian cases occurring in Italy he could follow in PLATE Xiri. Cycles of Kstivo-autuiTinal Parasite. Fig. Fig. Fig. Fig. Fig. Figs Figs Figs Figs Fig. Fig. Figs Fig. Figs Fig. Fig. Fig. Fig. 1. Very j'oung form. 2. Infection of one cell with seven young parasites. ( Drawn from a marrow smear.) 3. Triple infection. Two parasites joined by single chromatin mass. Double infection. Peculiar rings with two chromatin grains at opposite poles. Double infection. Small ring adherent to cell. 7. Signet-ring forms. Sub-division of chromatin. 9. Later ring forms, with sub-divided chromatin and few pigment grains. 4- 5- 6, . 8, . 10-12. Full-grown forms with finely sub-divided chromatin and gradual concentration of pigment. 13, 14. Stages of presegmenting forms, with concentrated eccentric pigment. 15. Double infection with separate presegmenting bodies. 16. Estivo-autumnal rosette. .17, 18. Young crescent and ovoid. 19. " Pulsating " crescent. . 20-22. Various forms of crescents. 23. Two bows about single crescent. Fully developed crescent ; two masses of chromatin ; achromatic substance ; double wreaths of pigment. Diagrammatic flagellating body. Extra-cellular sterile body. ' 24. 25- 26. THE ESTIVO-AUTUMNAL PARASITE. 391 the blood the complete cycle, but that in cases occurring in Kamerun the later forms were not found in the finger blood. Plehn describes a variety of parasite which he believes is peculiar to hemoglobinuric fever, and of which the later forms are of very small size but abundantly represented in the peripheral blood. In five cases examined by the writer the entire de- velopmental cycle of the estivo-autumnal parasite could be followed in the peripheral blood, and on the forms observed in these cases is based the present description of the later phases of this parasite. After the ring has reached its full size (4 //, 24 hours +) the swollen segment begins to increase in bulk and to involve a larger portion of the circumference, yielding forms seen in Plate XIII., Figs. 8 and 9. Some of these forms closely resemble the turban-shaped rings of the tertain parasite (Plate XIII., Fig. 8), but are much smaller. A few fine pigment grains were usually found scattered along the periphery of the growing segment. Forms corresponding to the full-grown ter- tain parasite with homogeneous body are rarely seen in the peripheral blood. When encountered they are found to occupy three-fourths of the shrunken cell, stain homogeneously with methylene-blue, and fail to exhibit a distinct nuclear body after methylene-blue or hema- toxylon. Most of the larger forms of the parasite seen in the peripheral blood give evidence of approaching segmentation, exhibiting a distinctly reticular structure and a condensation of pigment into one or two clumps. (Plate XIII., Fig. 10.) In many of these bodies the original ring persists at one segment of the parasite, but appears to be of re- duced size, and is sometimes subdivided by strands of protoplasm. The nuclear body at this period fails entirely to stain with methylene- blue and is indistinct after hematoxylon, resembling in this respect the full grown homogeneous tertain organism. The presence of a distinct achromatic spot adjoining the clump of pigment is very frequent in these forms, and this spot is found by Nocht's stain to be composed of chromatin granules. The further development of the pre-segmenting forms is represented in Plate XIII., Figs. 11-15, two separate parasites in the same red cell being occasionally seen. In them the reticular structure becomes more distinct, the pigment is still further concentrated, and the subdivided nuclear bodies appear as small achromatic spots in the meshes of the reticulum and again stain distinctly with hematoxylon. Estivo-autumnal rosettes appear in the peripheral blood of rare cases only and in moderate numbers, and exhibit a very uniform struc- ture. The pigment is grouped in a central granular clump, or, rarely, somewhat scattered. The spores seem to be arranged in two rows, but this appearance is probably an optical effect produced by the flattening of the more or less spheroidal body of the rosette, the spores origi- nally lying in the central axis of the rosette, falling in the hardening process, within those lying on the periphery. When admitting of accurate enumeration their numbers are found to vary between 18 and 21. The same number of spores was repeat- 392 MALARIA. edly counted in favorable specimens, made by the writer from the marrow of a fatal case. In sections of the tissues of fatal cases, how- ever, the number of spores appears to vary between wider limits, i. e., 8 to 20, but as the entire rosette need not always be included in the section the observations made in smears are the more reliable. A rim of hemoglobin invariably surrounds the rosette and strands of hemo- globin are frequently fouifd running between the spores for a variable distance, sometimes within the outer row. These rosettes differ from the tertian segmenting forms in the smaller size of the body and shrunken appearance of the cell, and in the small size, but not in the number, of the spores. The Changes in the Chromatin of the Estivo-autumnal Parasite. — In the young ring forms the early subdivision of the chromatin has been noted by Gautier, and in the writer's specimens was a prominent differential character from the tertian rings. A great variety of appearances is pro- duced by the irregular subdivision and distribution of the chromatin in the young estivo-autumnal parasite, many of which have been sketched or de- scribed for the tertian rings. The grains are usually quite small and are sometimes apparently fused into a spindle-shaped mass, lying within the ring. Other peculiarities which may be noted are : a markedly unequal size of the grains, a widely separate position, a frequent concentration in the center of the ring, and very rarely, a complete absence of chromatin. After 24 hours' growth, the chromatin granules become more numerous and extremely minute, and are inclosed in a trace of the milky substance, as in the tertian parasite. The changes in the chromatin in the pre-segmenting estivo-autumnal body are similar to those of the tertian parasite. In some of the writer's specimens the chromatin granules were found in radiating lines stretching from the parent mass to the new peripheral groups. In many specimens the peripheral groups of granules were well formed while the central portions of the body contained many diffuse granules. The relative quantity of chromatin in some of these bodies appears surprisingly large. The spores in the mature rosette usually contain single compact grains of chromatin which stain readily by methylene-blue, but in some rosettes two large gran- ules of chromatin are seen in a few spores although the rosette seems ready to burst. The double nuclei seen in many young estivo-autumnal rings may per- haps be referred to the incomplete fusion of the chromatin in the rosette. Characters of the Pigment of Estivo-autumnal Parasites. — When any considerable quantity of pigment gathers in the estivo-autumnal para- site it is usually found in one or two groups, but rarely is diffuse. When the parasite has reached the full-grown homogeneous stage the pigment is commonly found concentrated in a single compact mass. This early concentration of pigment is one of the chief features which distinguish the estivo-autumnal from the tertian parasite, in the pre- segmenting stage. This fact has been fully emphasized by Gautier, and was very uniformly illustrated in the writer's cases. The Plurality of Species in the Estivo-autumnal Group of Para- sites. — The probability that several species of parasites are concerned in the severe types of malarial fever prevailing in tropical countries, especially in the summer and autumn, has been chiefly maintained by THE ESTIVO-AVTUMNAL PARASITE. 393 the Italian school, who divide this group of parasites into two species : (1) the quotidian and (2) the malignant tertian. 1 . The Quotidian Parasite. — The typical fever curve of this variety is rather rarely seen, more frequently in relapses than in initial seiz- ures, while a postponement of paroxysms is usually observed, and a continuous fever is very common. The typical attack is short, the fever lasting 6-8, rarely 12 hours, the temperature then falling to 37° C. The descriptions of the morphology of this parasite unfortunately refer only to its appearance in fresh blood. During the rise of the tem- perature, the sweating stage, and the first hours of apyrexia, the blood is found to contain a variable number of red cells infected with one or more very actively motile, or non-motile, parasites of discoidal or ring shape. During the afebrile period the parasites increase in size, the ameboid motion diminishes or ceases, and fine pigment grains are de- posited along the periphery of the organism. Later, in the larger forms, the pigment gathers in a single clump or heap of grains. Dur- ing the entire development, the infected red cell diminishes in size and presents a " brassy " color as a result of " acute necrosis " induced by the parasite. Rosettes are seldom encountered in the finger-blood, segmentation occurring principally in the internal organs, as seen in the aspirated splenic blood. Rarely, segmentation occurs before pig- mentation, but usually the numerous round or oval spores are found grouped about a central pigment mass, the rosettes being much smaller than those of the quartan or mild tertian parasites. Contrary to the rule in malignant tertian infections, the young para- sites are found in the blood from the beginning of the paroxysm, and, except in very mild cases, there is no period in the cycle when the parasites are absent from the finger blood. 2. The malignant tertian parasite is distinguished on both clinical and morphological grounds. Clinically the typical paroxysm begins with a sharp elevation to about 40° C, the febrile period lasts 24 or 36—40 hours, is marked by a pseudocrisis and preoritical elevation, the fever describing in the three hourly chart a very characteristic course which differs from that of the mild tertian paroxysms. A tendency toward various irregularities is common. In the blood the parasites may be scarce or even entirely absent at the beginning of the paroxysm. At the height of the fever the red cells contain small non-motile ring or disc-shaped bodies, or irregular ameboid bodies, which begin to show pigmentation toward the ap- proach of the afebrile period. Most of the parasites then disappear from the finger blood, and rosettes are rarely seen except in some very rich infections. The pre-segmenting forms are round or ovoid, one- fourth to one-half the size of the red cell, and the pigment is gathered in a single clump or in a mass of vibrating granules. The rosettes occupy about two-thirds of the red cell, and exhibit two rows of spores which usually number 10-12, rarely 15-16. The infected cells are markedly shrunken and present a " brassy or golden " appearance. 394 MALARIA. The authors distinguish the malignant tertian parasite from the com- moner or mild variety on the following features. 1. The malignant tertian parasite is smaller in all stages. 2. Its pigment is less abundant and often non-motile, while in the other the pigment is very abundant and always in vibratory motion. 3. The rosettes are smaller, contain only 10-12 (rarely 16) spores, and are rarely seen in the finger blood. 4. The infected cell is shrunken in- stead of being swollen, as with the mild tertian infection. From the quotidian parasite the malignant tertian is distinguished on the following grounds. 1. The tertian ameba is, in corresponding stages, larger and less transparent than the quotidian. 2. In the malignant tertian parasite the ameboid movement is lire- lier, so that the resting discoidal forms are less frequent than with the quotidian parasite. The larger pigmented tertian forms, also, are often ameboid, this property persisting for 24 hours or longer. 3. The pigment in the tertian parasite is often vibratory, but never in the quotidian. 4. In the quotidian rosettes pigment is sometimes wanting. 5. The appearance in the finger blood of a new generation of ter- tian parasites is seen some hours after the beginning of the paroxysm, therefore much later than with the quotidian infection. Marchiafava and Bignami admit that the similarity between the malignant tertian and the quotidian parasites is very great and that the differential diagnosis is difficult and possible only from the full-grown forms seen just before the paroxysm. They apparently do not feel quite certain that the quotidian and malignant tertian parasites are separate species, as is indi- cated by the following extract from their discussion on this point. (Syd. Soc. Transl., p. 88.) "The remarkable points of resemblance between the quotidian and ma- lignant tertian parasites make it very difficult to solve the question whether we have to do with different sorts of parasites in the strict sense, or with one and the same parasite which varies greatly in the time of its develop- ment, 24-48 hours — and there are all intermediate degrees. On this latter theory it becomes easy to ascribe the morphological differences to the vary- ing length of the cycle. But various facts oppose this hypothesis. First, the clinical types of the quotidian and tertian are clearly distinct from each other, and have a certain stability which is maintained in relapses and re- currences. Second, we have never met with intermediate forms or transi- tional cases, although it is very difficult to interpret the irregular fevers. Granting that the question cannot at present be solved definitely * * * we are inclined to adopt the view that the ameba of the quotidian and the ameba of the summer tertian are closely related varieties of one and the same parasite. ' ' This view is not materially altered in the authors' last treatise (1900). Mannaberg, and Grassi and Feletti accept the views of Marchiafava and Bignami and describe a quotidian parasite, but do not furnish original observations tending to confirm their opinions. Thayer and Hewetson, Ziemann, and Gautier, all of whom have studied very mi- nutely the estivo-autumnal parasites, the latter authors from extremely THE ESTIVO-AUTUMNAL PARASITE. 395 rich material in tropical regions, fail to find sufficient grounds for the subdivision of the group. From the writer's study of cases at Montauk among soldiers re- cently arrived from Cuba in 1898, it was concluded that it is impos- sible from the observation of parasites in the peripheral blood to dem- onstrate the existence of a 24-hour cycle, since the infecting broods are seldom compact. Quotidian paroxysms always seemed to be refer- able to infection with two broods of malignant tertian amebse. Two features were noted, however, which offered some hope of dis- tinguishing a quotidian from the tertian parasite. 1. The exclusive presence in a few instances of rings with double nuclei and without the signet. 2. The appearance of large numbers of very small adult parasites in the peripheral blood. Neither of these features proved sufficient for the positive identifi- cation of a quotidian parasite. The peculiar rings were usually asso- ciated with the ordinary signet-ring forms, and the rosettes seen in peripheral blood were identical in appearance with those found in marrow-smears from fatal cases of malignant tertian infectio n. It therefore seems necessary to conclude for the present that a spe- cial quotidian parasite has not yet been demonstrated and probably does not exist. As stated by Ziemann, it seems likely that estivo-autumnal fever is caused by a single species of parasite whose developmental cycle usually requires 48 hours, but may possibly be completed in 24, or, as in one of his cases, may be prolonged to 72 hours. The Hemameba Immaculata. — Grassi and Feletti, Marchiafava, Celli and Bignami, Guarnieri, Sacharoff, Marchoux, and Ziemann re- port cases in which rosettes were found in the blood or viscera which were free from pigment. Most of these authors, while admitting that the parasites may occasionally sporulate without producing pigment, are not inclined to regard the hemameba immaGulata as a separate species. Grassi and Feletti claim to have observed in a bird exclusive infection with a variety of parasite which fails to produce pigment, and regard the appearance in the human subject of rosettes without pigment as evidence of infection by a distinct variety of parasite. Mannaberg also accepts this classification. In the report of the examination of the viscera of this bird no mentiou is made of the presence or absence of pigment, and it is impossible to deter- mine whether or not the infection had failed to produce pigment in the vis- cera as well as in the peripheral blood. In all cases in which pigment-free rosettes have been found in the blood of human subjects there have been found the usual pigment de- posits, and pigmented rosettes, in the viscera. That there is consider- able variation in the quantity of pigment produced by the parasite in fatal cases is shown by the reports by Marchiafava and Bignami of fatal cases in which a microscopical examination was required to show the presence of very scanty deposits in the viscera. 396 MALABIA. Ziemann mentions that he has seen a pre-segmenting body of the benign tertian type which was entirely free from pigment. The writer believes that some examples of multiple infection of red cells have been mistaken for pigment-free rosettes, but this explana- tion can hardly apply to the reports of many such rosettes in the cere- bral capillaries. The majority of observers, including Marchiafava and Bignami believe that there is no pigment-free variety of human malarial organism but that the estivo-autumnal parasites may occa- sionally fail to elaborate pigment. The Crescentic Bodies. On the 4th-6th days of any but initial paroxysms of estivo-autum- nal infection the peripheral blood may contain red cells infected by spheroidal, oval, elliptical, or small crescentic bodies which represent the early forms of the sexual cycle of the parasite. The relation of these forms to the young ameba is not clear and it is not known whether the crescents develop directly from the spores or from some parent body, such as is seen with various coccidia, and in which the young 'parasites are of crescentic form before their discharge from the mother cell. The latter view is the more probably correct, being favored by analogy, but the earliest forms of crescents are very fre- quently spheroidal or elliptical. On the other hand, neither direct observation nor analogy favor the view that they spring directly from the ordinary ameboid parasites of the pyrogenous cycle, as claimed by Marchiafava, Celli, and Bignami. Mannaberg regards the crescents as the syzygia developing from the union of two young parasites, and while this view is lacking in sup- port from the actual demonstration of two parasites uniting to form the crescent, and is entirely contrary to the recognized mode of origin of homologous forms in other protozoa, there are some morphological features which are somewhat in its favor. These are the appearance, frequently, of two groups of pigment, occasionally of two masses of chromatin, and the isolated observation of two halters with one crescent. Appearing first as bodies somewhat shorter than the diameter of the red cell, through various changes in length, breadth, and shape (see Plate XIII. ), the adult crescent is developed, appearing in the blood usually after the fifth to the seventh days of the paroxysm. The average adult crescent measures about 9-12 /i in length, by 2-3// in breadth, but in some cases very large or giant crescents have been observed, the writer having seen specimens measuring 20 /i x 5//. While the young crescent contracts and swells, altering its shape, the older forms are quiescent, and neither show true ameboid properties. The ends are either pointed or blunt, many old specimens appearing with swollen ends. The pigment is in the form of coarse golden yellowish grains, rods, or possibly at times in rod-shaped crystals. It is the coarsest pigment FLAGELLATE BODIES. 397 elaborated by the parasite and is always more abundant than in the ameboid bodies of equal size. It is arranged in a single central mass or circle, or as a double circle resembling a figure 8. Occasionally it is found in scattered groups, especially in very large crescents. Although a membrane cannot be said to exist about the crescent, yet its outer border may be stained reddish by eosin, and it has therefore been supposed that a remnant of Hb surrounds the crescent on all sides. The membrane or thickened outer border of the red cell is closely applied about the convex side of the crescent, while across the concavity it stretches loosely like a halter. The writer has seen two of these halters spanning equal segments of the concavity of a single large crescent (Plate XIII., Fig. 23), while Marchiafava and Bignami have seen two adult crescents within a single cell. In the center of the crescent is a sharply marked, light blue staining, or achromatic area of variable size, containing the chromatin and usually also the pigment. Marchiafava and Bignami describe the nucleus of the crescent as vesicular, but for reasons already stated the writer is unable to regard the nucleus as consisting of any other struc- ture than the mass of chromatin granules, which can always be found in living crescents. In the young crescent tliese chromatin granules are of larger size than in the adult body, in which they become sub- divided, and, when obscured by pigment, extremely difficult to detect. Sometimes there are two separate groups of chromatin granules. Marchiafava and Bignami expose most specimens containing cres- cents in a moist chamber for a few minutes, thereby causing the pig- ment to separate, the nucleus to swell, and the chromatin granules to be more apparent. With Nocht's method this expedient is seldom necessary. Although crescents may show transverse segmentation and occasion- ally lateral budding, these changes are probably degenerative, as is also their vacuolation. Former views regarding their multiplication by various methods are uoav known to be erroneous. That the marrow is a special seat of the development of crescentic bodies is believed by Councilman, Bignami, and Bastianelli, who have found an excessive number of young forms in this tissue when they were scarce elsewhere. The long persistence of crescents in the blood and their resistance to quinine are matters of common clinical observation. Yet Leukowicz, who describes several varieties of crescents, denies that they are any more refractory to quinine than are the ameboid forms. Flagellate Bodies. When blood containing crescents is allowed to stand in the air or under a cover-glass for a few moments some of the crescents slowly assume the spheroidal form, active vibratory oscillations of the pig- ment granules begin, and soon, from one or more points, pscudopodia shoot out with active lashing movements. These flagella continue their 398 MALAEIA. movements for some time, changing their position actively, their shape slowly, while some may be seen to break off from the body and swim off through the plasma. The formation of flagellate bodies represents the second stage in the sexual cycle, probably never occurring in the human body. These flagellate bodies are found, on staining, to be composed of a spheroidal pigmented mass, usually surrounded by a remnant of the red cell, of 2 to 4 flagella of variable type, of one or more lateral buds, and of chromatin, which has now undergone a remarkable transforma- tion. The bulk of chromatin is apparently much increased, and most of it is usually found in the form of long slender threads within the flagella. Occasionally short threads of chromatin may be found within the sphere. The flagella are composed of an outer protoplasmic cover- ing, their ends are blunt or pointed, their borders even or bulbous, and while most of them contain chromatin threads, some are entirely lack- ing in this constituent. Not all crescents develop flagella, some remain in the semilunar form, exhibit slow undulatory movements, and tend to stain more deeply with methylene-blue. It is believed that these crescents are penetrated by a motile flagellum, after which they become actively locomobile, and it is further supposed that these motile fertilized crescents or " vermiculi " are capable of penetrating the wall of the mos- quito's stomach, and there developing the encysted bodies, which are the next stage of the bi-sexual growth of the parasite. The evidence on which this relation of two varieties of crescents is indi- cated is largely derived from analogous phenomena, known to occur among various coccidia. Simond and Siedlecki, in the coccidia, have seen the fer- tilization of one crescentic body by the motile flagellum of another, and MacCallum, examining the parasites of crows, has seen the entrance of a flagellum into a crescent with the development of a motile vermiculus. (Cf. Solley and Carter.) The same evidence indicates that the flagella are mo- tile chromosomes of a karyokinetic nucleus, and that their function is that of a male fertilizing element. It thus appears that the crescents and their flagellated derivatives constitute a divergent developmental series destined to perpetuate the species in the body of the mosquito. Tertian flagellate bodies develop from the full-grown tertian parasites in mucii the same way as from crescents. Of full-grown tertian para- sites three varieties must be separated : (1) Bodies which are destined to segment, and which the writer believes are often produced by con- jugation. (2) Large hyaline forms which develop flagella. (3) Large hyaline forms with little chromatin which probably become fertilized by detached flagella. This last form has apparently been observed in shed blood as a large vacuolated, apparently sterile, body, but its fertilization has never been observed. In the tertian parasite which is developing flagella, the chromatin increases in quantity, becomes arranged in a basket network, breaks up into several coarse filaments which may be found protruding from the edge of the sphere, and finally these filaments are discharged in the form of flagella, of which there are usually six to each parasite. THE MALARIAL PARASITE IN THE MOSQUITO. 399 Quartan parasites develop flagellate bodies very similar in appear- ance to those of the estivo-autumnal type. III. The Development of the Malarial Parasite in the Mosquito. It is principally owing to tlie labors of Eoss, at Hanson's suggestion, that the life cycle of the parasite in the mosquito has been elucidated. Eoss followed the development of the proteosoma of Labbe in the tissues of mos- quitoes which had fed upon the blood of birds containing this malarial para- site. He traced the development of an encysted body in the wall of the intestine, and the formation therein of large numbers of so-called germinal rods. Finally, when the germinal rods were found abundantly in the sali- vary glands, the complete cycle of the parasite in the mosquito, and its natural mode of access to the human body were made clear. The demon- stration was rendered doubly positive when Bignami succeeded in transfer- ring the estivo-autumnal infection from one human being to another by pass- ing it through Anopheles claviger, a species of mosquito found in the Eoman Campagna. Later, tertian infection was transferred in the same manner. The details of development in the mosquito have been worked out principally by Grassi, Bignami, and Bastianelli, while much valuable information regarding the few dangerous varieties of mosquitoes, their habits, and the means of identifying and destroying them, have been contributed by Ross, Nuttall, the Italian writers, and by the Jenner School of Tropical Medicine. Developmental Forms. — Two days after the mosquito has fed upon blood containing crescents, there are to be found in the intestinal sub- mucosa spindle-shaped, oval, or spheroidal bodies, resembling the spindle-shaped bodies of the blood. They contain a single nucleus, scattered pigment, and their protoplasm appears vacuolated. The actual development of these bodies from the crescents and the pene- tration of fertilized crescents into the intestinal wall has not been demonstrated. By the third to fourth days the bodies are much enlarged, distinctly encapsulated, their protoplasm is reticulated and their pigment has largely disappeared. By the fifth to the sixth days, the encysted bodies have greatly in- creased in size, measuring up to 70 ju in diameter, and project into the celomic cavity. They are now found to contain numerous nuclei and globules resembling fat. By the seventh day fully formed germinal rods are found in the cyst in enormous numbers. These rods average about 14 /i. in length, their ends are pointed, and at the center of each is a granule of chro- matin. They are arranged in radiating or undulating masses about numerous centers. After the seventh day broken capsules are found which have dis- charged their rods into the celom, from which region they make their way into the salivary glands. In addition to encysted bodies which develop rods, some contain " brownish bodies " of various sizes, the nature of which is not known. 400 MALARIA. Very similar phases have been described for the development of the tertian parasite, while Bignami and Bastianelli have recently succeeded in tracing the similar phases of the quartan parasite. IV. Conjugation of Malarial Parasites. In 1897 the writer's attention was attracted by a specimen of blood from a rich tertian infection in which there were very numerous twin parasites of a younger brood while all the older parasites were single. A remarkable specimen of this type was secured at Montauk, in which both broods were very compact and very numerous, and nearly all the younger parasites were twinned, while all the older forms were single. Extreme length and variety of ameboid processes also characterized the younger parasites in both specimens. Later the writer was able to follow such a case through one complete cycle and found that the young twins were succeeded by full-grown single parasites of larger size. Finally, the application of Nocht's stain furnished convincing pictures of all stages of union, first of the bodies, later of the nuclei of the twin parasites, and left no room for doubt that under some con- ditions malarial parasites undergo a form of conjugation. The phases of this process are sketched in Plate XIV. Usually the members of the conjugating pair presented distinct diiferences in appearance. One was a thick compact parasite with abundant cytoplasm, one large and sometimes one smaller granule of chromatin, and was seldom seen in the ring form. The other was a ring-shaped parasite of delicate contour, of larger diameter and with rather less chromatin. The separate development of each of these parasites was followed in these cases up to the large spheroidal bodies. The compact parasite usually retained its compact form throughout, producing a parasite re- sembling in size the quartan forms, and staining very densely with methylene-blue. In some instances it developed 6 or 8 separate masses of chromatin granules, but the writer could not be certain that it sporulated. The delicate rings, developing singly, produced medium- sized, pale-staining parasites of full-grown appearance, but beyond this stage it was not followed. The conjugating parasites produced large segmenting forms with 16-20 spores. The extent and significance of this form of conjugation it is difficult to determine. In the cases in which it can be profitably studied, para- sites are usually very abundant, and in most cases few indications of the process can be detected. It is, therefore, probably not essential to sporulation, but when parasites are scanty the chances of finding typical examples of conjugating pairs are greatly reduced, and the peripheral blood may not be a complete index of the processes in the internal viscera. The large size of the sporulating forms developed from conjugating pairs suggests that this process is intended to espe- cially favor the multiplication of the species in the human host. It seems probable that conjugation ocours in the first generations of PLATE XIV. Conjugating Cycle of Tertian Malarial Parasite. Fig. I. Single compact body with double chromatin masses. Fig. 2. Conjugating rings of unequal size. Fig. 3. Double infection with a coarse ring, double chromatin granules, and a thin ring form. Figs. 4, 5. Early stages of conjugation of a thin ring and a compact body. Fig. 6. Early ameboid figures of conjugating rings. Fig. 7. Double nuclei in ameboid parasite. Fig. 8. Union of nuclei, and subsidence of ameboid motion in older conjugating parasites. Figs. 9, 10. Stages of union of bodies and of three chromatin masses, of two conjugating parasites. Fig. II. Complete union of bodies and nuclei. Figs. 12, 15. Comparative sizes of full-grown forms developed with and without conjugation. PARASITES IN THE CIRCULATION. 401 the infection and becomes less frequent as the disease progresses, the infection in the iiuman host thereby tending to limit itself. Several sources of error in the interpretation of these appearances have to be considered. The presence of two masses of chromatin has been noted in single young parasites. Yet the existence of the bodies of two parasites can usually be determined with certainty, while some conjugating forms contain three large masses of chromatin granules, i. e., when one of the conjugating forms supplies an extra granule. The conjugating forms must, therefore, be identified from the presence of the bodies as well as the chromatin masses of two parasites. The death and extrusion of one, or, as suggested by Thayer, of both of the twin parasites cannot account for the absence of older twinned forms, since no traces of such extrusion were encountered. Twin parasites sometimes proceed to full development without conjugating, but this event is comparatively rare. The existence of these three types of full-grown tertian parasites naturally suggests that the single compact form develops into the microgametocyte or male flagellating body, the single ring-shaped form developing into the macrogametocyte or female form, and the conju- gating form into the sporulating body. V. The Occurrence of Parasites in the Circulation. In all well-marked initial attacks of malarial fever the parasite can be found in the blood if examined tvithin 18 hours after the chill. The writer found this rule to hold even in patients who were taking large doses of quinine, but in one case examined 24 hours after the chill parasites could not be found during a search of two hours through stained specimens. In the great majority of ciuchonized cases of acute estivo-autumnal fever, parasites can be found within a week after the beginning of the paroxysm, while crescents may persist for two or three weeks or more. In cases treated by quinine the writer found the ameboid parasites to largely disappear from the blood on the third day. If no crescents appear, the blood then becomes entirely free of parasites, but usually crescents enter the circulation on or after the fourth day and the ex- amination of the blood still continues positive. On the other hand, with very mild or with cinchonized cases a prolonged search is sometimes required for the detection of parasites. In a few relapses of estivo-autumnal infection, examined during the chill, the writer had to search 35-50 minutes before discovering a sin- gle parasite, and in several other cases only young crescents could be found, while if the examination was delayed for l