1911.] Ancestry and the Incidence of Cancer in Mice. 43 



They have been arranged in age-periods of three months' duration, this 

 being the shortest interval which gives reasonably large figures in each 

 group. 



From the pathological standpoint the data are practically perfect. All 

 the animals which did not present a tumour during life were carefully 

 examined for tumours after death, and it is scarcely possible that any 

 growth of considerable size has escaped being examined microscopically and 

 recorded. The mice in which tumours were discovered during life have 

 been kept under observation till death. The tumours have been examined 

 microscopically in every case, and only those which were undoubtedly 

 malignant are reckoned in the tables and ancestries. 



The figures refer to females only. The number of cases of malignant new 

 growths in males bred in the laboratory is so small that a statistical study of 

 their frequency could not give useful results, nor do the males which 

 developed new growths appear in the ancestry of the females at present 

 under consideration. 



The tables show the ratio which deaths from cancer, and more especially 

 cancer of the mamma, bear to deaths from all causes at each of seven three- 

 monthly age-periods for female mice over a number of years. The age- 

 period in which mice dying of diseases other than cancer are entered is 

 given by the age at death. The mice which have died of cancer are entered 

 in the age-period embracing their ages at the time the existence of a tumour 

 was discovered, and not in that embracing their age at death.* This basis 

 for the age groups has been chosen instead of the actual age at death, 

 because the latter varies with the exigencies of other experiments having no 

 direct connection with the statistical studies. In order to include 

 cancerous mice still living and so increase the volume of the data it was 

 necessary to increase the non-cancerous totals also, by including living 

 non- cancerous mice. These have been included in the age-periods embracing 

 their ages on a selected day (in the case of the present tables, October 24, 

 1910). 



Table I, giving this distribution, and discriminating between cancer of 

 the mamma and cancer of other organs, shows a rapidly increasing- 

 proportion of deaths from cancer commencing after six months had passed, 

 attaining a maximum in the three-monthly period ending at 18 months, 

 and then diminishing till, in mice over 24 months old, the frequency is 



* In the earlier paper, one mouse was recorded as exactly six months old when the 

 tumour developed. Actually, the tumour was discovered at the age of six months and 

 seven days, and the mouse is therefore entered in the six to nine months age-period in 

 the present tables. 



