The formation of adhesions around the seat of the disease is of considerable importance both as regards diagnosis and also surgical treatment. The adhesions themselves vary considerably in different cases, in some being of recent formation and so soft that they are easily broken down by the finger, while in others the abdominal viscera are so firmly matted together that it is impossible to separate one organ from another. In one form or another they exist in the vast majority of the cases in which the disease has run its usual course, but the frequency with which the more important variety is met with has been variously estimated by different writers. Debelut found well-marked adhesions in one-half of the cases he collected, and Osier in about 56 per cent, of those which came under his observation. Gussenbauer and V. Winiwarter noted their existence in 370 out of 542 cases of pyloric cancer (68 per cent.), while, in 300 cases in which laparotomy was undertaken for the relief of the disease, Guinard states that the pylorus was only perfectly movable in fourteen, or about 5 per cent. In our own series of 265 fatal cases, adhesions were sufficiently numerous to attract the attention of the pathologist in 213 instances, or in 80 per cent. It may therefore be accepted that in about four-fifths of all cases the affected region of the stomach is adherent to some neighbouring organ at the time of death.

The situation of the growth exerts a certain amount of influence upon the formation of adhesions. This is shown in the following table.

In order that plastic perigastritis should be set up it is necessary that either the neoplasm itself or its toxic products should gain entrance to the subserous lymphatics, and it is consequently found that adhesions are most frequent and extensive in those cases in which the disease has infiltrated the whole thickness of the gastric wall or has undergone ulceration. Both conditions are best exemplified in the soft spheroidal-celled cancers, which rapidly penetrate the muscular coat and produce deep ulcers. The cylindrical-celled variety, although at first chiefly confined to the submucous tissue, is apt to slough, and is therefore rarely unaccompanied by adhesions after the lapse of a few months. A localised scirrhus of the pylorus, on the other hand, often grows very slowly, and not infrequently gives rise to fatal vomiting without having excited sufficient inflammation of the peritoneum to fix the part to a neighbouring viscus.

Disease of the pylorus and the lesser curvature usually gives rise to adhesions with the under surface of the right lobe of the liver, while a growth of the cardia either involves the same organ or the pancreas. When the posterior wall is the seat of ulceration, the pancreas is almost always affected, and in many instances the stomach also becomes united to the vertebral column. A neoplasm of the great curvature is particularly apt to invade the transverse colon, and in those rare cases in which the fundus is primarily affected the spleen often becomes attached to the base of the disease. It is comparatively rare for only one organ to be involved in adhesions, since the inflammation of the peritoneum soon spreads to the other abdominal viscera. In our series of cases two or more organs were firmly united in about 70 per cent, of those in which adhesions occurred ; while in about 10 per cent, all the viscera were so matted together that it was impossible to separate one organ from another.