APPENDIX II. 147 



This result was confirmed by the following ex- 

 periments : 



1. Starch was burnt, firstly, in a current of oxygen 

 gas, and secondly, by admixture with chlorate of 

 potash and peroxide of manganese. 



Heat units furnished by one gram of starch burnt 

 with 9'75 grams chlorate of potash . . 4290 



Heat units furnished by the same weight of starch 



burnt in a stream of oxygen gas .... 3964 



Difference ..... 326 



2nd. Phenylic alcohol was burnt with chlorate of 

 potash, and the result compared with the calorific 

 value of this substance as determined by Favre and 

 Silbermann. 



Heat units furnished by one gram of phenylic al- 

 cohol burnt with 9'75 grams chlorate of potash 8183 



Heat units furnished by one gram of phenylic 

 alcohol when burnt with gaseous oxygen (Favre 

 and Silbermann) 7842 



Difference ...... 341 



These three determinations of the heat evolved by 

 the decomposition of 9 '75 grams of chlorate of pot- 

 ash, furnishing the numbers 378, 326, and 341, 

 agree as closely as could be expected, when it is 

 considered that all experimental errors are neces- 

 sarily thrown upon the calorific value of the chlorate 

 of potash. 



The mean of the above five experimental numbers 

 was, in all cases, deducted from the actual values 

 read off in the following determinations. 



It was ascertained by numerous trials that all 



