322 FRANCIS B. SUMNER 



jumps, these being commonly associated with the struggles of 

 the animal. On the other hand, the temperature sometimes re- 

 mained stationary in individuals which struggled a good deal, or 

 underwent a slow, steady rise in the case of some mice which 

 remained very quiet. Finally, in some cases, particularly in the 

 cold room, the mercury fell somewhat after the maximum reading 

 was reached, this fall amounting sometimes to a considerable frac- 

 tion of a degree. And the records likewise show that a fall, instead 

 of a rise resulted at times from the struggles of the animal. 



Considering all the circumstances, it may appear to the reader 

 as extremely doubtful whether any figure could be recorded which 

 fairly represented the 'normal' temperature of a given mouse. 

 And indeed it must be conceded that, amid so much variability, 

 individual figures are of little value, but that recourse must 

 be had to averages based upon as many determinations as 

 possible. 



It was my original intention during these tests to leave the 

 thermometer inserted until the mercury reached a point at which 

 it remained stationary for at least half a minute. This, in many 

 cases, proved to be impracticable, since the temperature often- 

 times refused to remain stationary for so long a period. More- 

 over, the point at which the mercury came to rest was sometimes 

 so high that it could not have represented the normal temperature 

 of the animal prior to the excitement and struggle of the test. 



I also recorded in all cases the maximum temperature attained, 

 and in some of the experiments the temperature which was indi- 

 cated thirty seconds after the insertion of the thermometer. This 

 last I regard as representing much more truly the temperature 

 of the undisturbed animal, since, on the one hand, the interval 

 was more than sufficient to allow the mercur}^ to become warmed 

 to the body temperature/^ while on the other hand it was not long 

 enough to permit of any considerable rise due to excitement or 

 struggle. In the tables (7-12) I have given in separate columns 

 the maximum temperature recorded, as well as the thirty-second 

 figure when this was availal:>le. Unfortunately, this last was not 

 systematically recorded during the earlier tests, though it is 



'1 In water, the final reading was obtained within 10 or 15 seconds at the utmost. 



