146 PROFESSOR W. A. TILDEN ON THE SPECIFIC HEATS OF METALS AND 



TABLE V. Molecular Heats of Compounds. 



A = calculated from the atomic heats of the elements. 



B = calculated from the observed specific heats of the compounds. 



The figures contained in this table show that in the cases of tin-telluride, nickel- 

 telluride and the silver-aluminium alloy containing 92'28 per cent, of silver, there is a 

 remarkahlv close approximation of the values under B to those under A, the 

 differences between the two columns being throughout well within the limits of 

 variation due to experimental error. 



With regard to the aluminium-silver alloy containing only 24 '9 per cent, of silver, 

 however, there are differences which are somewhat greater. The values for silver- 

 telluride must be regarded as open to suspicion, for reasons which have already been 

 indicated, and they are not included in the table. If the mean atomic heats of silver- 

 telluride are compared, it is found that the difference between the sum of the atomic 

 heats and the molecular heat of the compound increases considerably with the 

 temperature, as seen below : 



This is perhaps due to some change taking place in the constitution of the solid. 

 This, however, does not seem to be the explanation in the case of the aluminium- 

 silver alloy, in which the differences between the two columns of figures, though not 

 constant, do not increase appreciably 



