458 



Lord Rayleigh. On the 



[Feb. 18, 



0*4778 gram stood on the left of the balance with globe (11) on the 

 right. The position of equilibrium of the pointer, as determined 

 after four different releasements, each observed in the usual manner, 

 was 19*02 scale divisions. In like manner, after the evacuation of 

 October 27, with the same weights in use, the equilibrium position of 

 the pointer was 18*46. After the filling of October 26, the weights 

 associated with (14) were 0*3220 gram, instead of 0*4778 ; and the 

 pointer reading was 20*08. So far as the weights are concerned, the 

 value of the hydrogen would be 0*47780*3220, or 0'1558 gram ; but 

 to this we must add a correction corresponding to 1*34 scale divisions, 

 being the difference between 20'08 and (19-02 + 18-46). At the time 

 in question, the value of a scale division was 0*00020 gram, so that 

 we obtain 



0-1558 + 0-00027 = 015607. 



The particular weights in use on this occasion were such that no 

 correction is necessary in order to allow for their errors. 



The mean barometer reading at the time of filling was 29" 742, so 

 that the factor required on this account is 30 : 29"742. The correc- 

 tion for temperature of gas is from 13*3 to 12. 



Log 0-15607 = 1-19332 



For barometer 0'00375 



For temperature 0'00198 



Log 0-15814 = T19905 



To this a correction for the temperature of the barometer has still 

 to be applied. For 1 F. the correcting factor is (1 0*000089), or 

 for 5 F. (10-000445). From 0*15814 we are thus to subtract 

 Q-00007, giving the tabular number 0*15807. 



A further minute correction to the mean of each set may be made 

 for the temperature of the glass. A warm globe is larger than a cold 

 one, and consequently holds more gas. If we suppose that the 

 volume expansion of the glass per degree C. is 0'000025, we find, 

 corrected to 12 C. 



Hydrogen. 



