340 
Francis B. Sumner 
the latter, 21 positive cases, 14 negative ones, and 1 case of 
equality. 
The diagrams (Figs. 8 and 9) show these relations graphically, 
and demand no discussion. As regards the ear, the case can hardly 
be regarded as convincing. 
In the next table (C) the averages for the weight- groups have been 
presented. In the 12 groups which admit of a comparison between 
Fig. 10. 
Tail length of the same generation of offspring, likewise at the age of 31/2 months; here grouped 
according to weight. 
the averages for all individuals (sexes combined), the mean length of 
tail, foot and ear is larger in the warm-room descendants in 8, 10 
and 9 cases respectively. The probabilities for these majorities are 
1 in 5, 1 in 52 and 1 in 14 respectively. The chances that in 27 out 
of 36 cases (considering all three characters) the »warm« figure would 
be larger are only 1 in 508. 
It must be granted, however, that when the sexes are considered 
separately, the figures are somewhat equivocal. There are 15 groups 
in which we may compare the >warm« and >^cold« figure for mice 
