If we compare the death rate from cancer in the last 40 or 50 years we notice a universal increase in all civilized countries. The increase has taken place rather steadily and not by leaps. According to W. J. Whitney there was in Massachusetts from 1850-1910 during each 5 years an average but not quite uniform increase of 1.20 in the death rate for 10,000 persons over 30 years of age every five years. According to Dr. S. C. Dixon in an area comprising one half of the United States in population the death rate increased from 4.79 (1890)-7.31 (1907). In New York City there has been an increase from 5.4, which was the death rate during the 4 years ending 1882 to 8.1 during 1908-1912. A similar increase we notice in Germany, Norway, Amsterdam and everywhere else, where statistics are available.
While there can therefore be no doubt as to the actual increase in the death rate from cancer, the interpretation of this phenomenon is not quite clear. While some authors believe in a real increase, others believe it to be only apparent, due to improvements in diagnosis and to better registration. It is pointed out by some authors (especially by Bashford) that the increase in the death rate is found mainly in internal cancers which are difficult to diagnose. The increase concerns to a great extent cancer in the alimentary tract. This increase is found especially in persons over 60 years of age. Now in children cancer not rarely attacks internal organs, and still no noteworthy increase is reported in those cases. Furthermore errors in diagnosis would equally prevent physicians from diagnosing as cancer cases which are not cancer as well as from attributing to other causes deaths really due to cancer.
Furthermore the same increase has been noted in the well-conducted autopsies in the Charité Hospital in Berlin, where the diagnostic methods have not to any considerable extent changed within the last thirty-five years. According to Professor Orth, in autopsies on persons over 20 years of age cancer was found from 1875-1885 in 4 per cent., in 1904 in 10.7, 1909 in 19, 1911 in 20.8 of all the cases. It is, therefore, probable that a certain factor which is potent in the production of cancer has been gradually changed within the last fifty years causing a general increase in the cancer rate, while at the same time there has been a marked decrease in the death rate from tuberculosis in the corresponding period.
We will now consider the relative frequency of the different varieties of human cancer. Here we find again a similar distribution of the various kinds of cancer in all civilized countries in Europe, America and as far as we can judge at the present time also in Japan. The most frequent seat of cancer is the gastro-intestinal tract, especially the stomach, and in the second place the female generative organs, especially the cervix of the uterus and the female breast. Cancer of the stomach is approximately equally frequent in man and woman; in
