214 Love and iSmeal : 



that the differences are as often positive as negative — viz., out 

 of 63 observations. '11 differences are positive, 23 are nega- 

 tive, and 13 are zero. They are distributed quite arbitrarily, 

 without any reference to the vahies of / or /'. The largest 

 individual difference is 0.15°, and this is twice positive and 

 twice negative. The mean difference is 0.007°, which is 

 certainly negligible, and the mean without regard to sign is 

 only 0.05°, which is the observable limit. 



For linen and silk there are only 11 observations, but the 

 tendency to similar results is easily seen. Tlie mean diff'erence 

 in this case is 0.04*^, and the mean without regard to sign 

 0.07^ By comparison with the values to be expected from 

 Ekholm's formula, it is evident that these are of the nature 

 of accidental variations, and have no connection with any 

 physical fact. 



In Table II. are given the observations as a whole. The 

 values under the heading /' are the means, where necessary, 

 of the several wet-bulb readings ; the pressure is given reduced 

 to 0= C. 



Taule II. 



