204 DR. A. VERNON HARCOURT ON VARIATION WITH TEMPERATURE, ETC. 
When ni has the constant value 20, the values of n in the fourth column of the 
table diminish regularly. 
The coiiditions of these experiments are such that no definite measure of the effect 
of heat energy upon chemical energy can ])e deduced from them. 
In the next paper, vol. 66, pp. 496-510, Trautz obtains a formula for n of a 
complicated character based on thermodynamical principles and containing expres¬ 
sions involving atomic heat, vapour pressure, and molecular heat of combination. 
The fundamental theorem of the investigation is open to criticism, and has been 
shown to be erroneous by Sackur in a paper in ‘ Zeit. flir Electrochemie,’ Nov. 15, 
1909, 15, 22, p. 865. Trautz uses his expression to calculate the values of Jc in 
several experiments, but the values found and calculated do not agree sufficiently to 
render the expression probable. 
In any experiments upon the effect of temperature on chemical change which are 
affected l^y the variation of the side effects of temperature enumerated above, the 
main effect is inevitably masked, because more than one condition is varied at the 
same time. In the experiments by Harcourt and the author, recorded in the 
Bakerian Lecture, 1895, and in the experiments by Harcourt set forth in the 
present paper, 1912, these complications have been avoided and the true relation 
between temperature and chemical change has been established.] 
