254 



PROFESSOR W. A. TILDKX ON THE SPECIFIC HEATS OF METALS, 



We know so little of the molecular constitution of solids that any empirical 

 formula which we may use for the calculation of the actual values of K, the specific 

 heat, must be regarded as having no application higher than 100 C. and lower than 

 182 '5 C. ; nor indeed can it have much correctness near these limits. 



We have used K = a + bt + cl z , and found the best values of a, b, and c for 

 calculating K from the given mean values ; but we find that according to this law K 

 both for cobalt and nickel reaches a maximum at 167 C. and rapidly diminishes for 

 higher temperatures, and this is certainly wrong. 



As it is possible that K is always greater at higher temperatures we were tempted 

 to use 



. i u 

 ~ 1 + cf 



But a came out negative. Now as a is the value of the specific heat at 273 C. or 

 t = we cannot imagine it ever negative. Indeed it is probably never less than k 

 the atomic specific heat. We therefore tried 



K = k + bt + ct* + et s , 



but the values of the constants are such that this causes K to reach a maximum at 

 270 C., and rapidly diminish afterwards. 



To satisfy all our notions we know of better formulae to use, but the labour of 

 calculating the constants of these more promising formulae coming after our other 

 failures seemed too great. 



We have, however, found a formula, easy to use, which fits Dr. TILDEN'S results 

 with some accuracy, which gives K = k when t = 0, and which causes K to con- 

 tinually increase with temperature. It is of the form 



