According to Lebert ulceration is met with in three-fifths of all cancers of the stomach; and with this statement our own observations closely coincide, since we find that it was present in 64*5 per cent, of our cases. It occurs most frequently among the soft medallary growths, but it is also common in the fungating forms of adenocarcinoma and in localised scirrhus, though it is rare in colloid. The ulcer itself varies considerably in appearance in different cases. In the soft exuberant growths it usually takes the form of a deep crater-like excavation, the edges of which are thickened, irregular, and overhanging, the walls shaggy, and the base studded with villous or fungoid processes. When the tissues of the stomach are affected with a diffuse form of infiltration, the inner surface often presents several discrete ulcers, which are somewhat oval in shape, with their long axes parallel to the great curvature. Occasionally they are quite superficial, and resemble simple abrasions of the mucous membrane ; but as a rule the edges are slightly thickened and everted, while the base, which is situated in the submucous or muscular coat, is hummocky or terraced, andcovered with fine papillae or with loops of blood-vessels. In colloid carcinoma the ulcer is usually superficial, and has a characteristic reticulated appearance.

Although carcinoma is most frequent in the pyloric region, the tendency to ulceration appears to be greatest when the growth occupies the walls of the viscus in the vicinity of the curvatures.

This is probably due rather to the type of the disease than to any influence of locality, since growths of the central regions of the viscus are usually of the medullary or villous type, which are particularly prone to ulcerate.

The size of the ulcer varies considerably; in some cases it hardly exceeds the dimensions of a split pea or a hemp-seed, while in others tracts of tissue several inches square may be involved, or almost the whole of the inner surface of the organ may be affected (fig. 12, p. 14). The depth of the ulcer varies in different cases. In the scirrhous and colloid forms of the disease it is comparatively superficial and rarely extends beyond the submucous coat, while in the softer growths more or less destruction of the muscular layer is met with in nearly 40 per cent., while exposure of the peritoneum occurs in about onequarter (27 per cent.) of the cases.

The different features which are thus presented by a cancerous ulcer depend to a great extent upon its mode of formation. In scirrhus the solution of continuity often owes its origin to stretching of the mucous membrane by the subjacent growth, and to its partial deprivation of arterial blood by the pressure exerted upon its nutrient vessels. The gradual clevitalisation which is thus induced renders the tissue unable to withstand the solvent action of the gastric juice, which consequently gives rise to superficial erosion. In other cases the mucous membrane is itself invaded by the disease, and undergoes a gradual necrosis as the result of the retrograde changes that occur in the new tissue.

The extensive ulceration which so often attacks the softer varieties of tumour is caused either by softening and disintegration of the growth, or by gangrene arising from thrombosis of a nutrient artery. In the former case the process is a gradual one, and while the central portion of the tumour is destroyed, rapid proliferation takes place in the surrounding parts. In the latter, large masses are apt to slough off, with imminent danger to the integrity of the gastric wall and to the large blood-vessels that supply it.

Simple ulceration of an acute character occasionally occurs along with carcinoma. As a rule it develops in the immediate vicinity of the disease, and owes its origin to an extension of the arterial thrombosis to which reference has just been made. In other cases the ulcer appears at some distance from the growth, either near the cardiac orifice or in the first part of the duodenum. In such it is usually found that the patient had suffered from the symptoms of septicaemia for several weeks before death, and occasionally the mitral valve shows signs of recent endocarditis. It is probable, therefore, that the disease originated in septic embolism, as in ordinary cases of pyaemia. In one of our cases of gastric cancer which succumbed to fatal haemorrhage the coronary artery was found to have been eroded by a small simple ulcer on the lesser curvature ; while in another fatal peritonitis ensued from the perforation of an acute ulcer of the duodenum. On the other hand, a chronic simple ulcer, if it precedes the carcinoma, is apt to be invaded by the malignant disease ; while the so-called ' simple ulcer,' which occasionally develops opposite the growth, is always found on microscopical examination to possess a cancerous structure.