32 



ROYAL SOCIETY OF CANADA 



Annual Mean Temperature at Various Depths. 

 The annual means for the different thermometers are as follows 



The means are taken from May 8th to May 8th in each case. The 

 differences between the thermometers, so far as we have been able to 

 analyse them, appear to be due, partly to the time of reading in the case 

 of the upper thermometers, and partly to the peculiar conditions of the 

 winter. The mean in the case of the 20 " thermometer appears to be 

 unduly depressed, owing to the continuance of the frost at that de^ith 

 till May the first. 



If we take the average of the six lower thermometers, which, making^ 

 allowance for time of reading, should agree fairly with the annual mean 

 temperature of the soil, we find the value 45-45° F., which is 2-92° F. 

 higher than that of the aii*. If we adopt the suggestion made last year^ 

 and take this excess as being a measure of the protective influence of the 

 snow covering during the winter, we find that the protection during 1895- 

 1896 was more efficient than that of 1896-1897 in the ratio of 4-33° ta 

 2-92°. 



Yalues of the Thermal Diffusivity of the Soil. 



The average value of the thermal diffusivity of the soil at various 

 depths, as deduced from the annual range of temperature and retai'dation 

 in the case of the different thermometers below the 20 " level, is found by 

 analysis of the curves to be 000361 in C. G. S. units Centigrade. This 

 value is in exact agreement with that obtained last year with the same 

 thermometers. 



In our communication last year, we were able to show, by the appli- 

 cation of the graphic method previously explained, that there appeared 



