DAY AND SOSMAN I EXPANSION COEFFICIENT OF GRAPHITE 289 



ing products, reaches this tentative conclusion: "Graphite in the 

 most restricted sense of the term is an allotropic form of carbon 

 having a definite and perhaps not very complex molecular con- 

 figuration When an organic compound is decom- 

 posed, there results a mixture of substances constantly increasing 

 in complexity until finally carbon is obtained. This carbon need 

 not be regarded as a simple substance, but may be considered to 

 be a mixture of many varieties of carbon each with a different 



number and arrangement of atoms in the molecule 



In a given sample of amorphous carbon some of the molecules 

 will be capable of easily undergoing rearrangement under the 

 influence of heat to form graphite molecules, while others will not, 

 and the proportion of molecules capable of such change will 

 determine the character of the final product." 



Arsem's definition is as follows: "Graphite is that allotropic 

 form of carbon having a specific gravity of 2.25 to 2.26." This is 

 the specific gravity of Acheson graphite, and all the graphites 

 made by Arsem were equal to or less than this. "Those varieties 

 of carbon which have some of the physical properties of graphite, 

 such as color, softness, and streak, but a lower specific gravity, 

 may perhaps be regarded as impure graphites; that is to say, 

 mixtures of graphite with other forms of carbon." 



Against the view set down in preceding paragraphs, that so- 

 called graphite is not a simple and reproducible substance, we 

 have the recent work of LeChatelier and Wologdine. They 

 found that the density of Acheson and of five natural graphites 

 after purification to remove ash and compression to drive out 

 air, was 2.255, and conclude that this property defines graphite 

 as a simple and reproducible substance. But the wide divergence 

 in expansion coefficient which we have noted above seems quite 

 beyond the range of possible experimental error, and indicates 

 some fundamental difference between these graphites. 



8 Compt. rend. 146 : 49-53. 1908. 



