328 
Francis B. Sumner 
of the warm-room parents, belongiug to 33 litters, and 145 offspring 
of cold-room parents, belonging to 30 litters, survived to the age of 
six weeks, when the first measurements were made. There was thus 
a somewhat greater mortality among the warm-room descendants, a 
tendency which was manifested even more strongly during the sub- 
sequent history of the animals. 
Since the atmosphere in which these mice were reared underwent 
considerable changes of temperature from day to day and from week 
to week, depending upon climatic conditions, it was of course important 
that the animals should commence their lives at nearly or quite the 
same time, in order that they might all be subjected to conditions 
as nearly identical as possible throughout their history. 
It is fortunate, therefore, that the mean date of birth differed 
in the two contrasted lots by only half a day. The extreme dates 
were May 21 and June 6 for the cold-room descendants, and May 22 
and June 7 for the warm-room descendants. Furthermore, nearly 80 
per cent of the former lot were born within a period of 5 days (May 21 
to 25, inclusive); while 85 per cent of the latter lot were born within 
an equal interval (May 22 to 26, inclusive). Thus it will be impossible 
to ascribe any constant differences which may be discovered between 
these two lots of mice to differences in the external conditions to 
which they had been subjected during their own lifetime. 
The young were measured first at the age of 42 days^). In order 
that the ear and foot length might be measured accurately at this 
time, the mice were rendered insensible by ether. The linear meas- 
urements were taken with a graduated sliding caliper, indicating 
tenths of a millimeter. In the case of the foot and ear, two meas- 
urements were made of each, the mean figure being employed in the 
computations 2). The caliper scale was at all times invisible to me 
until the instrument was finally adjusted. This practice of course 
diminished the chances of unconscious bias in making the measure- 
ments. Further precautions were taken at the time of making the 
second series of determinations at the age of 31/2 nionths (see below). 
1. In a small proportion of each lot the age was 43 days: in a yet smaller 
number it was 41 days. 
2) The average difference between the first and second reading of the 
caliper was 0.19 mm for the foot and 0.12 mm for the ear. Assuming that the 
value sought for lay between the two readings of the instrument, it will be 
seen that the average error for foot and ear equals in each case about half 
the mean difference which was found to obtain between the two contrasted 
lots of mice. 
