40 Mr. Haldane Gee and Mr. Terry on 



was enclosed in a wide test-tube plugged with cotton wool, 

 and surrounded with a cloth. Later we used a steam- 

 jacketed copper vessel, into which the test-tube fitted. 

 The loss of material during the two hours' heating, start- 

 ing with rubber exposed to a saturated atmosphere as 

 tested by means of a hygrometer, amounted to 0"64 per 

 cent. 



Thermometers. 

 For the estimation of the temperature of the calorimeter 

 two thermometers by Hicks were used : — 



(i) No. 430296 had a range from 13° to 23° C. and was 

 divided into IO*^^ each degree being about 25*"*". 

 Hence : — 



1 o __ f) Krjim 



•or = •25'"'" 



a quantity which was readily estimated with the naked eye. 



(2) No. 430298. 



10 = 35'""* 



•1 = 3^5'"'" 



each -1° - 5 parts = -05 = •?'""* 



by estimation = -001 = •14"^"^ 



For ascertaining the temperature of the hot rubber, a 

 thermometer divided into degrees whose boiling point had 

 been determined, was used. 



The temperature of the enclosure was ascertained by 

 means of a thermometer by Heintz, estimating to "01°. 



Method of Experiment. 

 It was thought desirable to inter-roll the sheet of rubber 

 with metallic foil, so as to hasten the time of cooling. 

 This was accordingly done in most of the experiments, 

 though we found afterwards that it lessened the time of 

 mixture by no large amount. The foil employed through- 



