432 
Miscellanea 
For 
?i = 377 
m = 2000, wc find : 
Pi 
P2 
P3 
Pi 
= 312-4 
giving a total number of cases = 377'06, 
= -01 + 17-67 +84-65 + 155+2 
= 257-75. 
P lies therefore outside any table, i.e. the probability that such a distribution could arise 
from random sampling is only one in many many millions. It seems therefore clear that the 
reduction of our number of houses by 800 makes no substantial difference in the improbability 
of our result. The houses with three cases are quite sufficient in themselves — even if we 
neglected the four-case house — to make the distribution indefinitely improbable. 
We have next to consider the sources of this improbability. 
The possibilities are : 
(a) that these "cancer bouses" are larger and contain more inhabitants than the others, thus 
they would be more likely to have multiple cases. 
(b) that constitutions liable to cancer are hereditary and so the "cancer house" marks merely 
the presence of a " cancer family." 
(e) that certain houses have been inhabited by jJersons following the same occupation, and 
that "cancer houses" are those inhabited by persons with a bad occupational mortality for 
cancer. 
(d) that some houses by their environment, or by the presence of some organism render their 
occupiers more liable to cancer. 
I propose to examine the seven instances of three or more cases from the problem of these 
possibilities. 
(A.) No. 1. G V . The first case that occurred in this house was that of a station- 
master, aged 28, cancer of rectum. He died in 1855. The second case was that of a school- 
mistress, who died, aged 64, in 1889 from cancer of the uterus. The third case was that of the 
servant to a nonconformist minister and his wife who were living next door. This servant 
attended the second case, that of the schoolmistress, in her last illness, and died in 1890, aged 
57, of malignant ulceration of the intestines with perforation- The fourth case lived to her 
78th year in this house in association with the third, but removed before her death elsewhere. 
She died at 88 years in 1899 of cancer of the liver. 
(B.) No. 2. G V . This is the next house to A above. The first recorded case in 
this house was that of the nonconformist minister referred to in A. He died about 1870, aged 
about 60, of cancer of the stomach. The second case was that of his wife who died in 1880, 
aged 71, of cancer of the rectum. The third case was the wife of the man who followed the 
widow of the nonconformist minister. She died of cancer of the breast in 1881, barely a year 
after the minister's wife. 
We have thus in 26 years seven cases associated with two small houses, the servant to the 
occupants of one living in the other. It is clear that neither the size of these houses, the blood 
relationship of their occupants, nor any similarity of their occupations will account for these 
multiple cancer cases. 
