A simple ulcer is far more prone to perforate the wall of the stomach than a cancerous growth. According to our clinical statistics this accident occurs in 7 per cent, of all chronic ulcers,1 but only in 3 per cent, of the cases of malignant disease. The disproportionate tendency to perforation which is thus exhibited by the simple disease is further emphasised when post-mortem evidence is solely relied upon, for we found that out of 678 necropsies in which an open ulcer was present, perforation with general peritonitis had taken place in 153, or in 22.5 per cent., while in 1,062 fatal cases of gastric cancer a similar condition existed only in thirtythree, or in 3 per cent. Consequently, as a cause of death, perforation is more than seven times as frequent in ulcer as in cancer. The explanation of this phenomenon is to be found in the different pathology of the two complaints. In simple ulcer the necrotic process is limited to a small area of the gastric wall, and any peritonitis that may develop around the base of the disease is strictly circumscribed. It therefore happens that by the time the peritoneum is exposed the sole obstacle to perforation lies in the fortuitous adhesion of some contiguous organ. A malignant growth, on the other hand, is usually attended by extensive adhesions as soon as the muscular coat has become involved, while, pari passu with the loss of substance internally, the base of the ulcer is thickened by the growth of fresh material. Indeed, it is only when rapid sloughing occurs and the process of destruction outruns that of repair that the danger of perforation becomes imminent.

1 Ulcer of the Stomach and Duodenum, p. 200. This estimate includes scars as well as open ulcers.

The seat of election of the two diseases also exerts an important influence upon their respective proclivity to perforation. An ulcer affects the anterior wall in about 8 per cent, of all cases in which it occurs, but in cancer the percentage incidence of the growth in this position does not exceed 2 ; and since this part of the stomach is more liable to perforation than any other, owing to the almost invariable absence of protective adhesions, it follows that the accident must be most frequent in the simple complaint. For a similar reason general peritonitis is the usual result of the perforation of the stomach by a simple ulcer, while in carcinoma a localised abscess develops in about one-third of the cases. With regard to the establishment of internal fistulse, it is worthy of note that a gastro-colic fistula is twice as frequent in cancer as in ulcer, and that abnormal communications between the stomach and other parts of the intestinal tract are practically confined to the malignant complaint. On the other hand, the more frequent incidence of simple ulcer upon the anterior wall of the viscus renders this disease far more liable to produce an external fistula, and also permits the occasional perforation of the diaphragm.