1885.] 



On the Atomic Weight of Glucinum. 



189 



Gl , .. 99-20 



G10 0-70 



Fe 0-10 



100 -00 



and has determined its specific heat at varying temperatures up to 

 450° with the following results (for pure glucinum) : — 



eft, 0-4286 



c\l 0-4515 



c& 0*4696 



cJi 0-4885 



c 



\t 0-5105 



312 



eft, 0-5199 



c£ 7 0-5403 



These results correspond to the following empirical formula for the 

 true specific heat of the metal at varying temperatures — 



or with numerical values — 



^=0-3756 + 0-00106^-0-00000114^, 

 whence the following values for kt are calculated : — 



h -3756 



A; 100 '4702 



&o 00 -5420 



hoo • • • -5910 



& 400 0-6172 



&500 0-6206 



In the following curve these values are graphically represented. 

 The curve reaches a maximum at about 470°. and then falls ; 

 but whether it represents the specific heat at higher temperatures 

 than 500° is doubtful. The specific heat of glucinum thus rises 

 rapidly up to about 400°, and remains approximately constant 

 between 400° and 500° at - 62. If this number is multiplied by 9*1 

 it gives the atomic heat 5'64. Glucinum, therefore, belongs to the 

 same class as carbon, boron, and silicon, which agree with Dulong 

 and Petit' s rule at high temperatures only. And the true atomic 

 weight is that required by the periodic law, viz., 9*1 and not 13'6, as 

 was previously deduced from the specific heat between 10° and 100°. 



This conclusion is confirmed by the author's determinations of the 

 vapour-densities of glucinum chloride and bromide in a platinum 



