346 CHARLES G. ROGERS AND KENNETH S. COLE. 



So we may write 



- : = K a d a 4* K b 6 b , (2) 



since both theoretically and experimentally, Wi w% = o in (i). 

 If the quantity of heat H is liberated in (i) and the heat capacity 

 (water plus the water equivalent of the flask) is c, then 



dO*_^T_dH 



j ~ , -fc-aVa J^b"b 



dt c dt 

 and 



H= c\ f"dd + K a r Badt + K b ^ e b dt \ 



I '0o Je Je / N 



f r' r* \ (3) 



= c(e a - - 6 b ) -f c \K a I O a dt + K b I 6 b dt , 



Jo Jo j 



where is the value of 6 a when / = o. 



Independent runs were made to determine the values of K a 

 and Kb. With 6 b small as compared to d a , d9 a /dt was determined 

 over the range for the values of a used and found to be linear 

 in d a . The same was done for 6 with a small. With 50 cc. 

 in both flasks the values obtained were 



K a = .0046 microvolt per minute per microvolt difference, 

 Kb = .0014 microvolt per minute per microvolt difference. 



The water equivalent of the flasks was found to be 8 cc. so that 

 c = 58 cc. 



A TYPICAL EXPERIMENT. 



The preparations for an experimental determination of the 

 heat production of the eggs of Arbacia involved a variety of 

 considerations not usual in ordinary experimental work in 

 zoology. By careful tests the running sea water of the laboratory 

 had been found to be the most satisfactory form of available 

 thermostat, the temperature of the water changing only very 

 slightly during any experimental period usually only in thou- 

 sandths of a degree. Care was, therefore, taken to have all 

 glassware and implements used at the temperature of the sea 

 water. Flasks, pipettes, graduates, beakers, finger-bowls, dis- 

 secting instruments, wash bottles of the stirring apparatus, 

 thermopiles, water-cap and the animals to be used were all left 

 in running sea water of uniform temperature for some time before 



