68 PHYSICAL PROPERTIES AND COMPOSITION 



Also Behn l found that the atomic heats of other elements 

 fall much below 6 at low temperatures, changing from 6-3 

 to 4 in the case of iron if the range of temperature, instead 

 of 1 8 to 100, be taken as 79 to 186. 



(/3) Application of Specific Heat to determine Atomic 



Weight. 



Since for atomic weights above 39 the atomic heat is, not 

 exactly it is true, but approximately 6*5, it is possible to 

 determine atomic weights from the relation 



4-.^ 



" S ' 



which indeed does not give an exact value, but may serve 

 to decide if the choice lies between two considerably diverse 

 values ; that is frequently the case in atomic weight deter- 

 minations, and may be illustrated by an historical example. 

 When Mendelejeff proposed the periodic system, indium 

 presented a difficulty : the atomic weight 76 was attributed 

 to it, and such a value would not suit the system, coming 

 between the elements arsenic and selenium, 



As 75, Se 79. 



But Mendelejeff remarked that the analysis of the 

 chlorides, 



38 indium to 35*4 chlorine, 



may be interpreted by means of other atomic weights for 

 indium, since 76 is based on the assumed formula 



InCl 2 , 

 whilst e.g. InCl 3 leads to the value 



76x| = 114, 



which just fills up a gap in the periodic system between 

 cadmium and tin 



Cd 112, Sn 119. 



In such a case the specific heat is an assistance ; it was 



1 Wied. Ann. 1898, p. 242. 



