Diurnal Variation of Temperature. 117 
minima too early. Indeed he attempts to demonstrate these 
supposed facts by numerical illustrations derived from obser- 
vations made at Katharinenburg in April for 18 years, and at 
Tiflis in May for 10 years. The deserved influence of Dr. Wild 
in all that relates to meteorological investigation gives to his 
Opinion and counsel an importance which calls for a careful 
disproof of his mistakes in this respect, lest the progress of re- 
search be seriously impeded by the proposed disregard of alge- 
braic generalization. 
That a general mathematical formula, which absolutely rep- 
computation. In this particular case it would appear that 
been modified by an empirical correction, for the purpose of 
eliminating the so-called influence of annual variation. The 
* These, although insignificant in themselves, are in fact sufficient to change the 
Praha paras epoch of maximum by thirteen minutes, or from 14> 15™ to 14h 
