Carcinoma of the stomach may be said to be a disease of maturity. Scheffer found that only 2 per cent, of the cases" which he collected were less than thirty years of age; Brautigam's estimate, made on similar lines, was 2.5 per cent., Reichert's 3 per cent., while our results indicate a ratio of 2'5 per cent. In the following table we have arranged 2,604 fatal cases (necropsies) according to the age of each individual at the time of death, and in the next one 882 cases which were diagnosed at various hospitals in London.

Table 15.-An Analysis Of 2,604 Necropsies Upon Gastric Carcinoma, Showing The Age At The Time Of Death

Table 16.-Ax Analysis Of 882 Clinical Cases, Showing The Age At The Time Of Death

It will be observed that in the main the results expressed in the two tables are in accord with one another. In both the greatest number of cases in any decade occurs between fifty and sixty; while the aggregate between forty and seventy amounts in each instance to about 80 per cent, of the whole. On the other hand, the series of necropsies which were collected from different cities of Europe present a greater proportion of cases over seventy years of age than do the clinical figures from London. This is readily explained by the fact that in London the aged poor who are afflicted with incurable diseases are more often admitted to workhouse infirmaries than into general hospitals, while in other countries this distinction is not observed.

These two sets of figures, although of interest in their own way, merely indicate the period of life at which the majority of the cases of gastric carcinoma come under medical observation, and obviously afford no evidence as to the relative liability of the population to the complaint at different periods of life. We have therefore appended the following tables, which have been compiled for us by Dr. Tatham.

In the first table, which refers solely to cases met with in hospital practice, the maximum liability to the complaint appears to occur between fifty-five and sixty-five years of age, and thus corroborates the evidence obtained from the postmortem and clinical statistics previously quoted. When, however, the deaths from carcinoma of the stomach occurring in the whole county of London are taken into consideration, it at once becomes evident that the disease increases in frequency up to seventy-five years of age, and that the maximum liability really occurs between sixty-five and seventy-five. The fallacies inherent in hospital statistics regarding the influence of sex in disease have already been noted, and it now becomes evident that, owing to the infrequent admission of old persons into general hospitals, the effect of age has also been misjudged. We therefore conclude that, contrary to the usual teaching, the tendency to gastric carcinoma steadily increases with each decade of life until about the age of seventy-five.

Table 17.-Death-Rate From Cancer Of The Stomach (Fatal Cases In London Hospitals)

Table 18.-Death-Rate From Cancer Of The Stomach (Registration Returns For London, 1901)

It was formerly believed that carcinoma never attacked the stomach before puberty, but there is now sufficient evidence to prove that the disease does occasionally develop during childhood. Six cases have been reported in children less than ten years of age, but in most of them the details given of the necropsy are hardly conclusive as to the existence of primary cancer of the stomach. Thus, the one recorded by Williamson in 1841 was almost certainly an example of hypertrophic stenosis of the pylorus; while in those related by Kaulich and Widerhofer both the nature of the growth and its primary location are very doubtful. Kuhn's case would appear to have been an example of lymphosarcoma.

In 1877 Cullingworth published the case of an infant who commenced to vomit on the tenth day after birth, and died of exhaustion in the sixth week. The necropsy showed a soft pedunculated tumour, which had almost occluded the pyloric orifice and given rise to great dilatation of the stomach. On microscopical examination the growth was found to possess the structure of a cylinder-celled carcinoma, and as an example of such it has always been quoted, although it seems to us that the possibility of a congenital adenoma has never been duly considered. Ashby and Wright's case was a child eight years of age who presented an abdominal tumour during life. After death considerable dilatation of the stomach was found, along with an ulcerated cylinder-celled carcinomatous ulcer of the duodenum and two growths the size of peas on the cardiac side of the pylorus.

Between ten and twenty years of age the complaint is less infrequent, although, of the sixteen cases which are said to have occurred at this period, at least four appear to have been examples of lymphosarcoma. In this category we would include the case of a boy who came under our notice for pain in the abdomen, vomiting, and emaciation. The epigastrium and umbilical region were occupied by a hard tender swelling, which moved slightly with respiration and rapidly increased in size. After death, which occurred in three months, the stomach and duodenum were found to be infiltrated with a soft growth, which had produced metastases in the mesenteric glands and peritoneum. Although the disease was supposed to be encephaloid cancer, microscopical examination left little doubt that it was really lymphosarcoma.

After the age of twenty the disease increases in frequency, and the cases that occur during the third decade comprise 2.4 per cent, of those at all ages.

Sex appears to exert little influence upon the period of life at which the disease commences, since between thirty-five and seventy-five years of age its proportional incidence in males and females is practically identical in each decade. It is interesting to notice, however, that the precocious development of gastric cancer is far more common in males than in females, for out of twenty-five cases occurring before the age of thirty which we have collected from different sources, no fewer than twentyone were of the former sex, a ratio of more than 5 to 1. As far as our researches go, spheroidal-celled carcinoma appears to be more frequent in the young than the cylinder-celled, variety, and colloid changes are relatively more common.