150 Royal Society, 



Three such experiments gave, aside from the weight of the flask 

 itself, 



gi-ms. 



A. Mercury + alcohol at -39° C. =576-2029 



B. „ „ „ -41°-5 =576-2522 



C. „ „ „ -42° =576-2639 

 Deductiug the mercury, the quantity of which remained constant 



throughout, it appears that the flask contained of alcohol : — 



grms. 



A. at -39° 17-2676 



B. „ -41°-5 17-3169 



C. „ -42° 17-3286 



From the data in (10) these weights represented at the rsr 

 spective temperatures the following volumes : — 



cub. ceutims. 



A 20-2865 



B 20-3061 



C 20-3122 



From the data in (4) and (5) we get the capacity of the flask 

 up to the mark at the same temperatures : — 



cub. centims. 



At -39° 59-6609 



„ -41°-5 59-6568 



„ -42° 59-6560 



Subtracting the volume of alcohol in each case, that of mercury 

 was, 



cub. centims. 



In A 39-3744 



„ B 39-3507 



„ C 39-3438 



Hence the specific gravity as obtained 



In A 14-1954 at - 39° 



„ B 14-2034 „ - 4l°-5 



„ C 14-2064 „ - 42° 



(12) By comparing these numbers in pairs, we have as the 

 difference apparently due to a diiference of temperature of 1° C* : — 



From A and B -0032 



„ B and C -0060 



„ A and C -0037 ^ 



of which the mean is -0043. iw 



lieducing, by using this coeflicient, the abo\'e results to a single 

 temperature, and adopting that of the f using-point of the metal as 

 determined by Balfour Stewart, we get 



From A 14-1948 at -38^-85 C. 



„ B 14-1920 „ 



„ C 14-1929 „ 



* Of course really including errors in clLteruiiuatio.i of weights and tempe- 

 ratures. 



