342 FRANCIS B. SUMNER 



observations upon eight normal animals and the range was 36.1 

 to 38.6°." Pembrey tells us neither the age nor sex of these ani- 

 mals, nor the atmospheric temperatures in which they were kept. 

 His mean figure is about 0.7° higher than mine, but it is scarcely 

 higher than my figure for the warm-room females. 



It might be claimed that certain faulty conditions of my own 

 experiments partly account for these differences between my 

 results and Congdon's. For example, I had no constant tempera- 

 ture rooms at my disposal, and both lots of mice were accustomed 

 to wide ranges of temperature. Therefore, it might be claimed 

 their powers of temperature regulation were better developed. 

 It is possible that this is true in some degree, but surely Cogndon's 

 mice should have rallied after 19 days exposure to changed tem- 

 perature conditions. 



Again Congdon states that he put all of his mice through a 

 preliminary period of practice or habituation w'th the tempera- 

 ture-taking operations before making his final observations. This 

 was not attempted in my own experiments, although a few prelimi- 

 nary records were rejected as untrustworthy. It might be claimed 

 that the generally higher temperatures of my mice were due to 

 excitement, resulting from their unfamiliarity with the operation. 

 That this is not true I believe for several reasons : ( 1 ) In the thirty- 

 second readings of my experiments, it is not likely that the effects 

 of excitement are commonly manifest to any considerable extent. 

 Yet these readings differ from the maximum ones by only about 

 0.25°. (2) There was no invariable association between body 

 temperature and degree of struggle, although the two were often 

 seen to be related. (3) So far as may be judged from tables 7 

 and 10, in which it seen that certain mice were tested on several 

 different days, there seems to be no general tendency toward a 

 reduction in the later figures. (4) Finally, the habituation of 

 Congdon's own mice to the temperature tests does not appear 

 to have greatly reduced the variability of his figures. Unfortu- 

 nately, the individual thermometer readings have not been pub- 



»9 Journal of Physiology, vol. 45, 1912, p. 83. 



