104 Francis B. Sumner 
for a few seconds, after which a second pin was passed through the 
basal part of the tail, which was thus likewise fastened securely 
in place. The distance from the tip of the snout to the anus and 
that from the latter to the tip of the straightened tail (exclusive 
of terminal hairs) were found by means of a graduated sliding 
caliper provided with a vernier. When the ear length was taken 
the mouse was laid freely upon the right side. In measuring the 
length of the right hind foot, the animal’s body was held in the 
left hand, the sole of its foot pressed lightly against the board, 
and a pin passed through the instep into the wood. The tarsal 
joint was bent at about right angles, and the heel allowed to rise 
above the surface of support as in life. ‘The caliper spanned the 
distance from the tip of the heel, which was touched rather lightly, 
to the tip of the nail of the middle (third) toe. It was found 
important to make all of these measurements as soon after death 
as possible. In the later series, two determinations were made of 
each part measured, and the reading was taken in tenths of a 
millimeter. : 
Throughout the whole of this work, the practice was followed, 
so far as possible, of measuring warm-room and cold-room indi- 
viduals alternately, or at least of alternating small groups of indi- 
viduals. Only thus would such gradual changes in one’s stand- 
ards of judgment as would inevitably result from growing expe- 
rience, affect equally the two groups to be compared. 
The characters selected are few in number, we must allow, and 
are, for the most part, such as would not be expected to respond 
adaptively to the treatment accorded to the animals. One must, 
however, accept the limits imposed by brevity of time and the 
nature of the material at hand. When subjected to statistical 
treatment a very few characters are found to involve a very great 
amount of labor.. Moreover, aside from thickness of fur, such 
adaptations as might be conceived of as resulting from tempera- 
ture conditions would probably be histological or chemical in their 
nature and therefore not accessible to ready methods of quantita- 
tive treatment. 
The measurements which I have chosen as being applicable 
to these mice are, with the exception of those relating to the hair, 
