70 PHYSICAL PROPERTIES AND COMPOSITION 



mately the specific heat of elements which cannot be directly 

 measured in the solid state. Oxygen may be taken as 

 example, and its specific heat determined from that (0-156) 

 of Fe,0 4 . We have then 



M w = 232 x 0-156 = 36-2 

 M w = 2^^=3x6.5 + 4^ = 36-2 a; = 4-1 7, 

 so that the atomic heat of oxygen should be 4-17. 



(6) Theoretical Indications. 



Whilst experimental research has not succeeded in putting 

 Dulong and Petit's law into stricter form, and on the other 

 hand the solid state has not yet been satisfactorily brought 

 within the region of molecular-mechanical concepts, one is 

 driven on the theoretical side to possibly accidental relations 

 towards explained facts. 



In the first place, it must be observed that the phe- 

 nomena are simpler in the case of solids than liquids because 

 the work of expansion hardly comes into consideration, and 

 the specific heats at constant volume hardly differ from the 

 values forming the basis of Dulong and Petit's law. In 

 the second place, the complication due to formation and 

 decomposition of double molecules occurring in liquids is 

 possibly absent in solids. 



It is then striking how the atomic heats of solids are 

 quantitatively associated with that for gaseous mercury, 

 which is found both by theory and experiment to be 3. 

 It is noticeable that the probably monatomic metallic 

 elements have in no case so far given an atomic heat 

 lower than this, and it appears from Behn's experiments 

 that 3 is the lower limit for some of them. 



Element Atomic heat 



100 to 18 18 to - 79 - 79 to -186 

 Copper. . . 6 5-6 4-5 



Nickel ... 6-4 5.8 4.3 



Iron ... 6-3 5-6 4 



Aluminium . 6 5.3 4.3 



The values lying below 3 are for elements such as 



