﻿Prof. R. Bun sen's Calorimetric Researches. 163 



specific gravity of water at the same temperature, p the weight 

 -(in grammes) of melted ice which corresponds to the volume v 

 (that is, the indication of 1 division of the scale), then 



1 ,g, 



1 1 v 



or 



^^l=p . (i) 



S w Se 



There are numerous observations on the specific gravity of ice ; 

 but how little they agree with each other is shown by the fol- 

 lowing numbers found for s e : — 



Thomson 0*920 



Heinrich 0*905 



Osan 0*927 



Royer and Dumas . . 0*950 



Brunner 0*918 



Plucker and Geisler . . 0*920 



Kopp ...... 0-908 



Dufour 0*922 (maximum) 



Dufour 0*914 (minimum) 



Owing to the disagreement of these different observers, it 

 seemed to me essential to determine with greater exactness than 

 was possible hitherto the value of the constant p, and therefore 

 also of s e . I have for this purpose adopted the following plan, 

 by which the sources of errors which have made previous estima- 

 tions uncertain are completely removed. 



Fig. 5 represents a strong U-tube of hard glass which is drawn 

 out at a to a thick point. It is filled with mercury to b x b, which is 

 well boiled in both limbs as in constructing a barometer. The 

 point a is provided with an india-rubber tube, through which 

 (while the air in the limb a b is slightly warmed and again cooled) 

 distilled water free from air is allowed to pass over the mercury 

 at b. This water is then boiled for half an hour, and the india- 

 rubber tube (c) is kept in a beaker-glass filled also with boiling 

 water; as soon, then, as the boiling is interrupted at b, the 

 space a b becomes filled completely with water quite free from 

 air. The india-rubber tube (c) is then closed under water with a 

 piece of glass rod, and the point at a is sealed up. This is better 

 done with a Bunsen lamp than a blowpipe, by heating the part 

 where the tube covers the drawn-out point so strongly that it 

 is filled with vapour instead of with water. If the apparatus 

 has been weighed before filling with water, and is weighed again 



M2 



