absence of epithelial sheets has been considered characteristic of exudates, but the author's results do not support this view. There is a slight tendency to greater abundance of neutrophiles in ascites, due to portal stasis, than in tuberculous processes, but not sufficiently marked or constant to be of diagnostic value. The cellular elements present in ascites due to hydremia does not necessarily differ from those in tuberculous peritonitis or cirrhosis, except that they are usually less abundant. The predominance of neutrophiles in ascitic fluids may mean tuberculosis, acute peritonitis, or cirrhosis. It may be that the predominance of neutrophile cells always indicates the presence of inflammation, primary or secondary, but there is nothing in the specific gravity or amount of albumin to indicate that this is the case. No characteristic cells were found in cases of carcinoma of the peritoneum, but the epithelial cells were much more abundant and showed more fatty degeneration than in any other fluids examined, except a case of pseudoleucemia with involvement of the mediastinum, in which the finding resembled closely that of the carcinoma.

A review of pleural fluids shows the frequent presence of sheets of epithelial cells in exudates. This, Miller thinks, can readily be explained by the inflammatory process involving only a portion of the serous surface, the presence of the fluid on the uninflamed surface causing marked desquamation. It may be true that it is only in tuberculous pleurisies, secondary to lung involvement, that the neutrophiles are found in abundance, but this is of little practical value, as it is usually impossible to detect a tuberculous focus in a lung compressed by fluid. It is apparent that the lymphocytes predominate in tuberculous pleurisies, but the same may be true in a hydrothorax, and when we may find epithelial sheets in both the picture is identical. To be sure, as a rule, the cellular elements are more abundant in tuberculous pleurisies than in transudates, but not, however, of a sufficient degree to be of diagnostic value. In the acute non-tuberculous pleurisies the neutrophiles are usually abundant, and, as a rule, the predominating cell. When fluids of this type have remained in the cavity for a long time they present a lymphocytosis. One case of this type is found in my series. The presence of a large percentage of neutrophiles may mean either an acute simple pleurisy or a tuberculous pleurisy secondary to a lung focus. From a study of these cases the presence both relatively and absolutely of a large number of lymphocytes is only found in primary tuberculous pleurisy, but the absence of abundant lymphocytes does not mean that the process may not be tuberculous.

Regarding the cerebro-spinal fluids, too few cases have been studied to draw any conclusions. Significant, however, is the absence of increase in lymphocytes in the single case of cerebrospinal syphilis examined.

The following are the conclusions of the studies:

I. The color of the fluid, its reaction, and size of the erythrocytes are of little diagnostic value.

2. Nucleoalbumen is usually much more abundant in exudates than transudates. However, many exudates fail to show more than a faint trace.

3. On account of the constancy of the salts present, the specific gravity and amount of albumen bear intimate relation.

4. Fluids from different body cavities of the same individual, or fluids of the same origin in different individuals, or successive tappings of the same cavity, may show marked variation in the amount of albumen, probably explained by the permeability of the capillaries, the degree of irritation of the serous surface, the condition of the patient's blood, tension of the fluid, the amount of absorption taking place, and possibly the blood-pressure.

Methods of estimating the bulk albumen, as Esbach's or Purdy's, are not applicable to serous fluids.

6. Reuss' formula is the most accurate simple method for determining the amount of albumen in serous fluids.

7. The diagnostic value of the specific gravity, or amount of albumen, is greatly lessened by the frequent presence, simultaneously, of several processes. A hydremia, as a result of anemia or transudation from pressure, may lower the specific gravity of an inflammatory fluid. Secondary inflammation, high tension of the fluid, or absorption taking place, may increase the specific gravity of a transudate.

8. At present there is no absolutely reliable method for determining the tuberculous character of a serous fluid.

9. The specific gravity of a fluid, when taken for successive tappings, may be of prognostic value.

10. Epithelial cells in sheets are not infrequently present in fluids of inflammatory origin.

II.

The predominance of lymphocytes may be associated with a long-standing, simple pleuritis, a hydrothorax, or a tuberculous pleuritis.

13. The presence of relatively large numbers of neutrophiles in pleuritic fluids may be associated with an acute, simple, or a secondary tuberculous pleurisy.

13. The presence of numerous cells, chiefly neutrophiles, suggests an acute, simple pleuritis.

The presence of numerous lymphocytes, with few, if any, neutrophiles, indicates that the process is probably a tuberculous one.

15. In ascitic fluids there is little that is characteristic. Neutrophiles are usually more abundant in fluids, due to portal stasis from cirrhosis, than in tuberculous peritonitis.

16. An eosinophilia in one body cavity may occur along with a neutrophilia in another.

17. A study of the cellular elements in fluids from the pleural or abdominal cavities is of comparatively little diagnostic value either in differentiating transudates from exudates or in determining the character of the latter.

THE JOHNS HOPKINS HOSPITAL MEDICAL

SOCIETY.

MEETING HELD NOVEMBER 21, 1904.

Experimental Streptococcus Arthritis.-Dr. Rufus I. Cole reviewed this subject and reported some work he had done during the past year with the purpose of throwing some light, if possible, on the nature of the arthritis of rheumatism. The tendency, he said, was nowadays to regard acute articular rheumatism as an acute infectious disease, and the following four theories had been advanced to explain the condition: 1. Rheumatism is due to a specific organism as yet unknown; 2. It is due to a specific diplococcus or streptococcus; 3. It is a mild pyemia resulting from infection with the ordinary streptococci or staphylococci; 4. It is due to an ordinary bacillus. Those who have argued for the specific nature of the disease have been able to isolate a diplococcus from the joints, the heart's blood, and the exudates of many cases. Meier, noticing the close connection between angina and rheumatism, has been able to isolate from the throats of anginal patients streptococci which produced acute articular rheumatism in animals. Mentzer has held that the angina and the anginal rheumatism in these cases were due not to a specific organism, but to the ordinary parasites of the mouthmost often to streptococci. Diplococci have, however, been isolated from many cases (most often from severe ones), the most recent report having been that of Lewis of Philadelphia.

The Specificity of the Diplococcus of Rheumatism.-Dr. Cole said that those who claimed for the organism isolated a specific nature based this claim chiefly on its morphology; a paired coccus, sometimes in chains, the pairing being most marked in recent cultures, and the general features quite similar to those of the streptococcus-these were its main characters. Marmorek's test was said not to be reliable in classifying the organism, and the dark discoloration in growths on blood agar-at one time thought to be characteristic-was said to occur also with other organisms. By inoculation of the organism isolated the production of arthritis, of endocarditis, and of chorea had been reported.

Experimental Arthritis.--Dr. Cole's own work was begun with a streptococcus isolated from the blood of a patient suffering with endocarditis, septicemia, and joint pains. Inoculations were made into the ear veins of several rabbits, and an arthritis was produced in practically every case. Another series of experiments was then carried out in which streptococci were used for the inoculations taken from patients suffering with various non-rheumatic conditions and giving no rheumatic history. Six races of streptococci were then used, no one of them being the streptococcus of rheumatism. In practically all of the rabbits so inoculated a typical arthritis was produced. The English technique was followed, an emulsion of the culture being injected into the ear veins. The rabbits first became somewhat languid, and later became lame first in one joint and then in another. At autopsy a thick, tenacious fluid was found in some joint of the body, and

here the cartilage was smooth, but the synovial membrane injected. Smears from the exudate showed large numbers of diplococci with flattened sides. In two rabbits a typical endocarditis developed, and two showed twitching, inco-ordinated, possibly choreic movements. In other words, with streptococci from seven different sources results followed inoculations identical with those following inoculations of the so-called diplococcus of rheumatism, and the latter is therefore not specific. Dr. Cole called attention to the fact that the cases reported in the literature as having furnished specific organisms have been usually the severe cases with pericarditis and other complications. This fact, together with the frequency of a secondary invasion of streptococci, argued, he thought, against the specificity of the organisms obtained. Dr. Osler mentioned the specimens which Paynter obtained in his experiments, and called particular attention to the significant nature of the experimental subcutaneous fibroid nodules. He felt certain that the disease is an acute infection, but the organism, he thought, still undetermined. Dr. Cole said that Paynter and Paine had been able to get cultures from rheumatic joints, but that the attempts made at the Johns Hopkins Hospital had never been successful. Philip, who has reported elaborately on the subject, has also never succeeded in growing organisms from rheumatic joints. Dr. Bloodgood said that it has recently been noticed that in cases in which no organisms could be found in arthritic exudates the villi of the joint (particularly in gonorrheal cases) often contained bacteria.

The Mosquitoes of Maryland.-Dr. Kelly exhibited a case presented to him by Professor Smith of Rutgers in which the commoner varieties of mosquitoes-male, female, and larval forms-were beautifully mounted. He asked that further work be done to find out what forms infested Baltimore, and to eliminate the malarial bearers if possible. Dr. Thayer said that the work done thus far on the subject showed that three forms of anopheles were common in Maryland-the maculapennis or claviger, the punctipennis, and the crucians. The main source of the mosquitoes of Baltimore was said to be the drainage wells, which often go uncleaned for years.

Multiple Carcinoma of the Ileum.-Dr. Bunting showed the specimens from two cases of this condition. Primary malignant epithelial tumors of the small intestine were, besides, exceedingly rare. Only one has previously been seen in the Johns Hopkins Hospital, and only 30 could be found by Lubarsch in the whole literature. The first patient reported was a negro who had died with cardiac vascular symptoms. In the upper ileum six nodules were found which turned out, on section, to be carcinomata with alveoli consisting of small polymorphous cells and firm fibrous stroma with hyaline degeneration here and there. The tumors—though they answered only two of Billroth's requirements-were probably all primary. These tumors-and the ones reported in the literature have been similar—were small and occurred well along in the carcinoma age, grew slowly, and were relatively benign. They resembled the small skin carcinomata seen on the scalp and arising from the Malpighian layer (the Basalzellen Karsinom of Krohnpeker). The second patient showed symptoms of obstruction and a palpable mass in the right iliac fossa. Inoperable carcinoma was found at operation, and a colostomy was done. Autopsy showed multiple primary carcinomata of the small intestine, with no peritoneal involvement. Dr. Bloodgood said that these skin carcinomaļa were more frequently seen than

formerly, due in part to earlier diagnosis and in part to a realization by the laity that growths anywhere are not to be neglected. Clinically they are only slightly malignant, and might almost be called benign epitheliomata.

BALTIMORE CITY MEDICAL SOCIETY.

SECTION OF CLINICAL MEDICINE AND SURGERY.

MEETING HELD DECEMBER 2, 1904.

Gastroenterostomy and Pyloroplasty.-This subject was reviewed by Dr. Finney, and a report of his own operation, with record of results to date, was presented. At the time of the first published account of the Finney pyloroplasty five cases were reported. Since then 80 cases have been recorded in the literature, and the results have been very satisfactory. The procedure was said to be really an anastomosis between stomach and duodenum, with division of the intervening structures, and it should more properly be called a gastropyloroduodenostomy. (Dr. Finney had, he said, recently operated on his twenty-second case.) In this series there had been two deaths—one in a diabetic, who died in coma, healing of the pyloroplasty being found to be perfect at autopsy, and the other a death from intestinal volvulus five days after operation, the healing again being perfect. This case emphasized the fact that the so-called "vicious circle," so often occurring in anastomosis between stomach and intestine and characterized by persistent vomiting, may in reality be an intestinal obstruction. Dr. Finney said that his series warranted him in reporting progress, though the operation was only to be recommended in benign stenosis. He had, however, gradually extended the class of cases in which he performed pyloroplasty. Most often it was done for cicatricial obstruction. In two patients chronic indigestion without cause was the indication, and both were relieved. It has also been efficacious for persistent uncontrollable vomiting. In two patients with active gastric ulcer the operation was done successfully.

Objections to Finney Pyloroplasty.-The objections raised to the operation were said to be, first, that it was difficult in cases where adhesions were present. Dr. Finney said this had not been the case in his series. Second, that it was not of use in the presence of active ulceration. Two cases of those reported had, however, shown this condition, and had done well. Third, that the operation did not give a low enough pyloric opening, but Dr. Finney said this could be made as low as desired.

Advantages of Pyloroplasty.—These, Dr. Finney said, were (1) absence of regurgitation of bile; (2) restoration of practically normal anatomical relations; (3) slight mutilation of tissue; (4) absence of opportunity for formation of peptic ulcer.

Gastroenterostomy.-The history of this operation was reviewed and the defects of each procedure noticed. Dr. Finney then described a method which he had recently used, similar to the procedure described by Scudder of Boston. The portion of bowel chosen was at the junction of duodenum and jejunum. Here the intestine runs directly downward, and anastomosis at this point provides direct passage for the food, allows for a posteriooperation, uses the most dependent portion of the stomach or leaves a short