R. Bunsen — Cahrimetric Investigations. 177 



, . O. ^u> G^u, 



then IS S^ "^ S^ "^ S^ 

 and therefore : ^"^" = S, 



On account of the great accuracy which the method ju^t 

 described permits, it appeared to me superfluous to iimke moi^' 

 than three experiments. In the first of these the wattM- IVoz.' 

 between — 3°C. and — 5°C., in the second between — TC. aiil 

 -3° a, in the third between 0° C. and -2° C. The lolh.wiiig 

 weights, reduced to vacuo, were obtained : 



G«, = 141580 grams. 

 1st expt. G„ = 17-4400 " 



2d " G, = 17-4624 " 



3d " G^ = 17-4757 " 



If we make further 



S^ = 0-99988 Sj = 13-59600, 

 then the resulting specific gravity of ice S« will be : 

 1st experiment, 0-91682 



2d " 0-91673 



3d " 0-91667 



Mean of three expts., 0-91674 



We have therefore for the terms occurring in equation (2), 



the values : 



V --= 0-00007733 S, = 0-91674 S„ = 0-99988 

 and thence 



p = 0-00085257. 

 The weight of melted ice e which corresponds to T corrected 

 divisions on the scale is therefore : 



e - 0-00085257 T (3). 



Let the latent heat of melting, for water, be I, then will one 

 scale division correspond to « Z of the previously defined units 

 of heat For the amount of heat w, expressed m umts of heat, 

 ^nich is indicated by T scale divisions, we have therelore, 



or, if the value 80-025, hereafter to be determined, be substi- 

 tuted for Z, 



w = 0-068227 T (4). 



