writers have pointed out the acid smell of the peritoneal fluid in perforation, and C. Brunner has directly determined the acidity as due to free hydrochloric acid.

Microbic elements are introduced with the food, their number varying with the state of the buccal cavity, and especially with teeth. According to Miller, they may reach the duodenum or intestines. In a recent work, Coyon has found thirty varieties of bacteria in the gastric contents. That gastric juice has a certain bactericidal power is shown by the fact that the relative number of organisms decreases with the increase of acidity to digestion point, and also with certain albumenoid substances with which acid can enter into chemical combination. This varies also with the species of bacteria. In the pathological stomach, microbic elements are numerous, increasing if insufficient mobility favours gastric fermentation. C. Brunner's experiments on influence of degree of acidity are interesting. Stomach contents were collected after a test breakfast of Ewald, and containing free hydrochloric acid. This was injected successively, in quantities of 1 to 10, 15, and 20 c.cm., into the peritoneal cavity of rabbits. Four only succumbed to peritonitis, in two cases due to a streptococcus. The experiments were repeated with gastric fluid from a case of hysteria with neryous vomiting, from a nervous dyspeptic, from a stenosis of pylorus with gastric stasis, then from gastric ulcer with a gastro-enterostomy performed, of pyloric spasm and hyperchlorhydria with gastro-enterostomy. Forty-four rabbits

were injected with from 4 to 40 c.cm. of different liquids, and eleven died. A dose of 4 or 5 c.cm. was nearly always supported, but one in sixteen died. With 10 c.cm. the number of recoveries were double the deaths, but one animal resisted a dose of 35 c.cm. When the experiment was repeated with gastric fluid deficient or wanting in acid obtained from subjects of stomach disease, non-cancerous, all died, even after injection of a small dose, to 10 c.cm., the pathological agent being in most cases a streptococcus. Experiments with case of stomach cancer: a recovery after c.cm., but nine deaths after 1 to 6 c.cm., still due to a streptococcus, in one case

associated with the bacilli communis. Thus Brunner contests that hypo-acidity is incontestably more virulent than normal acidity, and that the acid, while aiding digestion, diminishes the virulence of the infective microbic agents, so that it becomes possible to inject them into the peritoneal cavity of rabbits without causing fatal peritonitis. He has also shown that the effect produced depends in the same subject on the time after Ewald's test meal. Thus the contents are more pathogenic a quarter or half an hour after meal than one and a half hours. Thus, of seventeen animals, all recovered after 4 to 10 c.cm. of fluid of one hour; but six out of seven died after injection of fluid of a quarter to half an hour from the same case (a case of spasm of pylorus); the same result being attained with that from a healthy subject, two out of three dying where fluid was of a quarter to half an hour, but of seven injections of one and a half hours' fluid there were no deaths. Thus experimentally the conclusion is arrived at that in pathological states where the total acidity is diminished or absent, pathological organisms which enter the stomach do not lose their primary virulence; that they multiply when the contents of the stomach accumulate or are stationary; and that they become, by their number and virulence, more dangerous than when they have been exposed to the actions of acid; and whilst in the perforation of cancer a small quantity of extrusion is very infectious to the peritoneum, in the perforation of ulcer the pathological virulence is much less, and will also vary according to the different phases of digestion. Clinical experience is entirely in accord with this. The number of cures recorded in which, in spite of general extravasation, general peritonitis does not follow, is considerable, and is explained by the fact that the gastric contents are acid in the case of ulcer, and that the peritoneum is capable of resisting even an extended contamination. There is one point, however, which cannot be allowed for, and that is that the successful cases are invariably recorded, while the unsuccessful ones may or may not be. Still, the conclusion is correct that in perforation by gastric ulcer the infective agent, with its microbic contents, always varies both quantitatively and qualitatively.

Bacteriological researches are yet few, but they have been collected by C. Brunner, 1903. Most frequently the agent is poly-infections, and rarely a mono-infection. The chief agents are Streptococci, pyogenes and lanceolatus, and the bacillus coli. Hydrochloric acid has no effect upon the former, which resists fluids rich in acid. Besides the streptococci, the pneumococcus, the streptococcus lanceolatus, and more particularly the streptococcus lanceolatus of Gamaleia, the micrococcus of salivary septicemia of Miller-a microbe of the mouth-all play a certain part.

Out of 22 observations, nine times streptococcus pyogenes (twice alone and seven times with bacillus coli), staphylococcus aureus, bacillus subtilis, bacillus mesentericus, an anærobic bacillus ; the pneumococcus or streptococcus lanceolatus four times, always with the bacillus coli; bacillus coli, ten times (nine times with streptococcus pyogenes, with a staphylococcus or pneumococcus). Bacillus subtilis is fre

quent in the stomach, and has been found in the peritoneal exudate. In perforations, diplococci appear frequently. Anærobic bacteria do not appear to have any importance.

Finally, C. Brunner cites two observations of Warren and Allen (Boston), where the cultures made show the peritoneal fluid to be sterile. An interesting experiment of Brunner showed that rabbits successfully resisted injections of 10 c.cm. of gastric fluid collected during an operation for perforated gastric ulcer where patient died of peritonitis, and where there was a large amount of extruded fluid in belly. The exudate showed the bacillus coli and streptococcus lanceolatus. Thus, in spite of a relatively feeble virulence of the gastric fluid, since rabbits resisted 10 c.cm., the patient died, showing that it is not always the quality, but the quantity, which constitutes the danger.

(To be continued.)

THE GLOMERULUS IN ACUTE NEPHRITIS.

IT might be thought that in such a common condition as acute nephritis, all that there was to observe microscopically would have been described long since. Better methods of technique and more accurate observations, as well as advance in knowledge as to the nature and development of cells, however, render it essential for us to revise our opinions upon the most ordinary of subjects from time to time.

The glomerulus of the kidney has always been somewhat of a puzzle, both to the normal and to the morbid histologist. The exact nature of the epithelial cells covering it, and of those lining Bowman's capsule the question as to whether or not it contains cells of a connective tissue origin—as well as the function of the structure itself, have occupied the attention of observers for many years. Intimately bound up with the structure and function of any particular tissue, as well as with the question of the changes which the tissue undergoes under pathological conditions, is the developmental history of the cells which compose it. Much has been done on the question of the development of the glomerulus from the time of Remak onwards. More recently Hamburger, Bornhaupt, and Ribbert have turned their attention to the question, but none have done more to elucidate the early stages in the process than Herring (Journal of Path. and Bact., July, 1900). The conclusions of the last-named author may be summed up thus :-The kidney arises from two distinct structures, viz., the kidney blastema and the ureter. The former is of mesoblastic origin, and consists of a mass of cells closely related to the blastema of the Wolffian body. From the kidney blastema arise the Malpighian bodies, convoluted tubules, the connective tissue framework, and capsule. Each Malpighian body, with its tubule, arises as a solid mass of cells at the periphery of a lobule, either under the capsule or in the inter-lobular septa, in close relation to the dilated extremity of a ureter branch or collecting tubule. The solid cell mass acquires a lumen, and takes on

the shape of an S. The lower limb of the S becomes a Malpighian body, the upper and middle limbs form convoluted tubules, the bend between them constituting Henle's loop, while the extremity of the upper limb joins the collecting tubule and forms the greater part of the junctional tubule. The Malpighian body forms in the lower limb of the S-shaped tubule, and is the first part to show differentiation. A splitlike lumen appears early, and attains considerable size. The convex side of the limb develops into the epithelium which lines Bowman's capsule, the concave side into the epithelium which covers the glomerulus. The epithelium on both sides is similar at first, but only for a very short time; the cells lining the capsule becoming thin and flattened.

The glomerulus is formed, not by an invagination, as many still suppose, but by a thickening of the epithelium. The capillaries of the glomerulus develop in situ, and are covered by the epithelium. The division of the glomerulus into lobes comes about by the penetration of the epithelial cells between the capillary loops. When the glomerulus is fully formed, the cells diminish in size, and become flattened like the cells lining the true capsule.

According to Herring, the glomerulus contains, in addition to endothelial and epithelial cells, cells of a connective tissue nature. This is denied by Fichera and Scaffidi. The homogeneous basement membrane of Bowman's capsule does not exist during the process of development. The cavity of Bowman's capsule may be regarded as of a nature similar to the peritoneal cavity, and the cells which line it as analogous to the endothelium of the peritoneum. That the cells covering the glomerulus have a secretory function is possible, especially as they are so closely related to the epithelium of the convoluted tubules developmentally.

Turning now to the bearing of these developmental facts upon inflammatory lesions of the kidney, we find the most recent views well summarised in a paper by Fichera and Scaffidi (Virchow's Archiv., 1904, Bd. 177, p. 63). These authors are in agreement with Herring on most developmental points except in this: that they deny the presence of