In the following table we have arranged 154 cases of carcinoma of the stomach according to the month in which the disease terminated and that in which it was supposed to have commenced.

We have given preference to the lethal event because the accuracy of the figures upon this point is beyond dispute, whereas those that refer to the commencement of the complaint are always open to doubt. It will be observed that the deathrate varies at different times of the year, being lowest during December, January, and February, and highest in September, October, and November ; or if the months be grouped according to seasons, that over 60 per cent, of the total deaths take place between June and November. These figures are, of course, too small to warrant any absolute conclusions being drawn from them, but at the same time they are too striking to be ignored, more especially as both Virchow and d'Espine noticed, fifty years ago, that persons affected with malignant disease die more frequently in the summer than in the winter. This fact is of itself somewhat curious, since one would have supposed that a disease which reduces the vitality to such a low ebb, and so often terminates by pneumonia, would be more likely to prove fatal to the poor inhabitants of London during the cold and wet winter months than in the more genial weather of the summer and autumn. At first sight it would seem as if an increase of the temperature of the air might stimulate the growth of a malignant tumour, in the same way that it does that of plants, and by accelerating its progress curtail the period of existence. That an increased rapidity of growth really does take place during the spring and summer is more than probable, for not only do cases succumb more quickly between April and August, but, as will be seen from the table, only about one-seventh of the total number were admitted into hospital during December, January, and February. Personally, however, we are strongly of opinion that carcinoma obeys certain laws of epidemiology, like enteric fever, acute rheumatism, and other infective diseases, and is consequently more rife at certain seasons than at others. If it be allowed that the usual duration of the gastric complaint varies from nine to twelve months, we should expect from the facts already noted that in the majority of the cases it would commence between September and February; and in this connection the last part of the table becomes instructive, for it will be observed that in 60 per cent, of the entire number the first symptoms of illness showed themselves within that period. We venture to believe that further investigation upon these lines will throw an important light upon the etiology of carcinoma.