340 CHARLES G. ROGERS AND KENNETH S. COLE. 



1. Mechanical jarring of the galvanometer used in making the 



measurements. 



2. The temperature of the room must be maintained as nearly 



constant as possible during the course of an experiment. 

 It was found to be especially necessary to avoid the 

 possibility of drafts of air striking the apparatus. 



3. It was found necessary to shield the electrical apparatus from 



stray electrical currents. This was found to be of the 

 greatest importance following a severe electrical storm. 



4. It was found wise to avoid all stress and strain in the wires of 



the thermopiles, such as might be caused by too much 

 bending of the wires, or placing tension upon them. 

 It was found possible to obviate much of the possibility of the 

 errors of the first group by using a water cap, designed so as to 

 provide a current of water coursing continuously over the 

 experimental flask as well as around it. 



APPARATUS AND METHODS OF WORK. 



The general method employed was that of the micro-calorim- 

 eter, developed by Hill in his work upon muscle. Fig. i 

 shows the experimental set-up used. This method has one 

 serious disadvantage. Since corrections for heat loss are de- 

 pendent upon the temperature difference and the time, long 

 runs can not be made, since these corrections soon become a very 

 large part of the result. In this work the corrections could be 

 kept less than 15 per cent, of the total temperature change 

 during a period of two and a half or three hours. Two straight 

 sided commercial vacuum flasks of about 75 cc. capacity were 

 used, (I.) containing 50 cc. of the egg suspension, and (II.) an 

 equal quantity of water. These flasks had been especially 

 exhausted through the kindness of Dr. W. R. Whitney of the 

 Research Laboratory of the General Electric Company. When 

 flask (I.) contained 50 cc. of water it had a heat loss of 16 calories 

 per hour per degree difference in temperature between the 

 interior and the exterior. The flasks used by Hill (2) had a loss 

 of about 12 calories per hour for 250 cc. and those used by 

 Shearer (7) about 19 calories per hour for 800 cc. This small loss 

 for a flask of such small capacity is evidence of the great value 



