152 Henry Clifton Sorby, Esq., on the 



viz., as diamond and coke, though their primary forms are not the 

 same, making therefore in all four crystalline species, the mutual 

 relationship of which I will now endeavour to prove to.be very simple. 

 These primary forms are shewn in figs. 1, 2, 3, and 4, which repre- 

 sent the relative volumes of equal weights of the four species. 



To ascertain the specific gravity of graphite, hard coke, anthracite, 

 or charcoal with sufficient accuracy, I have found it absolutely neces- 

 sary to have them in the state of extremely fine powder ; and, in the 

 case of charcoal, to boil it first in acid to dissolve the glaze of ashes. 

 I then heat it to dull redness, without access of air, to dispel any 

 volatile matter which it may contain ; and afterwards well boil it in 

 water in a glass bulb, and fill it full of boiled water and cork it. 

 By this means, the water, having been deprived of air, dissolves when 

 cold any that may remain amongst the powder. It is then weighed 

 in water, and the weight in air is ascertained by subsequently drying 

 it in the bulb. The powder is then burned to ashes, and their 

 weight in air and water ascertained, which being deducted respec- 

 tively from those of the substance, we can calculate its specific gravity 

 free from the eff'ects of the ashes. The following tables exhibit in 

 column A the specific gravity of the substance, no allowance being 

 made for the ashes, in B the amount per cent, of ashes, in C their 

 specific gravity, and in D the specific gravity of the substance cor- 

 rected from the effects of the ashes : — 



Graphite, 



The mean of all these is 2-177. It will be observed that the 

 first and last experiments differ considerably from the others, which 

 I have some reason to believe is partly due to errors of experiment. 

 The mean of the others is 2*174. If, however, we take 2177 for 

 the specific gravity of pure graphite, I think we shall not be far 

 wrong. 



The mean of these, which agree very closely with one another, is 

 1-8915. 



