Manchester Memoirs, Vol. Iviii. (1914), A''^ 10. 9 



bustfon, we find that the amount of heat evolved is different 

 for each form of carbon, although the number of atoms 

 taken and the number of CO2 molecules formed is identical. 



Looking for the cause of this dissimilarity, we are 

 driven to attribute it to the varying stability of the 

 molecules in the three cases, which must depend upon 

 the mode of linkage of the units inside the molecule. 



Returning to our table of classification, we expect 

 that the least stabilit}' will be shown by molecules whose 

 units have the power of free rotation ; the maximum 

 stability will be found in the molecule all the constituent 

 units of which are in a state of rigidity, the intermediate 

 case being a molecule having some units which are rigid 

 and some free. 



Now, considering the fact that the greater the stability 

 the smaller will be the evolution of heat on complete 

 combustion (compare the case of phosphorus), and corre- 

 lating this with the calorimetric measurements quoted 

 above, we find that 



Amorphous carbon is represented b)- Class I., where 



none of the units are rigid \ 

 Graphite, Class II., some units are rigid ; and 

 Diamond, Class III., (^r// units are rzWc/. 



Although these deductions are in good agreement 

 with Barlow and Pope's, Aschan's and Bragg's views, still 

 they cannot be taken as quite conclusive, as the calori- 

 metric measurements, which serve as a basis for these 

 conclusions, not only differ in magnitude with every 

 experimenter, but unfortunately contradict one another. 

 (Favre and Silberman, A. ch. [3] 35, 357, 1852. 



Berthelot and Petit, A. ch. [6] 18, 80, 1889. 



Mixter, Amer. J. of Science [4] 19, 440, 1905. 



Roth and Wallasch, BcricJite 46, 896, 1913.) 



