338 FRANCIS B. SUMNER 



of the former room, at the time of these tests, was 21.7°C., that 

 of the cold-room being —3.3°. The difference of atmospheric 

 temperature to which these mice were subjected was thus 25. 



If we base our computations upon the 'maximum' figures 

 (rejecting those designated as unreliable), our difference is 0.25°, 

 while if we employ all of the 'maximum' readings, including the 

 less reliable ones, the mean difference becomes 0.31°. 



It thus appears that if any reduction of temperature occurred 

 among these mice, it was a slight one, probably not exceeding a 

 fourth or a third of a degree. It may be, however, that a more 

 extended series of observations, under more satisfactory experi- 

 mental conditions, would have revealed a greater mean difference 

 than is here shown. It is quite possible, also, that the difference 

 in reaction between the sexes, implied by the greater change 

 shown among the females of my experiments, would be found to 

 be constant. In that case, we could not fairly combine all of 

 the figures into a single average. 



Cons' dering these experiments in relation to the main purpose 

 of the investigation — that of determining the effect of the actual 

 temperature conditions of my two rooms upon the body tempera- 

 tures of the respective sets of mice — it will be seen that the course 

 adopted was calculated to reveal the maximum effect of these 

 conditions, rather than the average effect for the entire period. 

 In view of the considerably smaller differences of atmospheric 

 temperature which obtained for the period as a whole, it may 

 well be doubted whether any appreciable differences of mean 

 body temperature would have been manifested at most times. 



5. Discussion 



As stated earlier (p. 316), Finkler found a difference of several 

 tenths of a degree between the temperature of guinea-pigs in an 

 atmosphere of 25° to 27°, on the one hand and in an atmosphere 

 of 0° to 5°, on the other. It is not stated how many individuals 

 were used. The exposure lasted several hours. 



The results of certain other investigators are metioned in an 

 earlier section of this paper. So far as I am aware, Congdon has 

 conducted the most extended previous series of investigations 



