646 



Professor Dewar 



[June 11, 



external lustre of the graphites coming from this district. The 

 Siberian example was from the Alexandref Mine ; its structure was 

 columnar and striated, with little external lustre ; it was rather more 

 easily broken up than the Borrodale, but formed the same dull black 

 powder. The specimen from Ceylon was of -^ the type usual to that 

 island : highly lustrous and flaky, breaking up very easily, and 

 forming small shining plates when ground up. The last sample, 

 which was from the same cabinet as the others, but whose origin 

 was unfortunately unknown, had a dull external surface, was exceed- 

 ingly porous, and much more brittle than any of the previous ones, 

 grinding up very easily into a dull black powder. It had more the 

 appearance of the celestial graphites, which was heightened by having 

 slight incrustations of sulphide of iron on its surface. Its low specific 

 gravity also shows it to be some exceptional variety. 



It seemed most important in connection with this subject to 

 examine some matrix with which the graphites are usually found 

 associated. These rocks are very variable, but consist principally of a 

 kind of decomposed trap or gneiss. A good specimen was obtained 

 of semi-decomposed gneiss from Canada with a considerable quantity 

 of graphite disseminated throughout the mass, and also several 

 samples of Ceylon graphite imbedded in its matrix, which in this 

 case consisted of felspar and quartz. 



The results, as tabulated below, confirm Wright's analyses of 

 several trap rocks, in which he found principally carbonic acid and 

 hydrogen. The small quantity of marsh-gas no doubt comes from 

 the disseminated graphite, but the presence of the hydrogen is more 

 difficult to explain and requires further investigation. 



On comparing these samples of graphite, it will be seen that the 

 Borrodale and the Siberian give off about the same total volume of 

 gas, that the celestial and the unknown graphites closely approximate 

 each other in this respect, yielding more than double the volume of 

 the others, and that the Ceylon sample stands alone in yielding a very 

 minute quantity. All the terrestrial samples, except that from Ceylon, 

 are alike in giving off a very considerable quantity of marsh-gas, 

 though they differ somewhat in the actual quantity, and it is evident 



