WRITINGS OF JAMES SMITHSON. 93 



But I have found with much surprise that flint can be 

 melted without difficulty; and even of a considerable bulk. 

 "Where the heat is most intense, a degree of frothing takes 

 place; where it is less, there is a swelling of parts of the 

 surface. The effects wore the same with French and Eng- 

 lish flint, with black and with horn-coloured. Does flint, 

 like pitchstone, contain bitumen, which, at a certain heat, 

 tends to tumefy it ? This might explain the smell from its- 

 collision, and the oil which Neumann obtained by its distil- 

 lation, and to which no credit has been ever given. No 

 doubt can, I conceive, be entertained of flint being a vol- 

 canic production. On this point I may speak again at a 

 future opportunity. 



In using mere water, diamond, anthracite, plumbago, 

 were particularly difficult of trial, as any adhesion they had 

 contracted with the sappare was quickly destroyed by the 

 combustion of their surface, while, as the intention in their 

 case is not to subject to great heat, they may be so secured 

 in the clay as at least very much to retard their escape. 

 Here acting on very minute particles is essential, as when 

 large pieces are employed, the effect is too slow to be per- 

 ceptible. 



A pleasing way of demonstrating the combustion of 

 plumbago, and of even exhibiting the iron in it, is to rub a 

 little from the wetted point of a pencil on one of the clay 

 plates mentioned in a former paper.* 



In trying diamond it was imagined that its glow contin- 

 ued an unusual time after removal from the fire. The pres- 

 ent method afforded the means of making a comparison. 

 A fragment of diamond, and another of quartz, chosen pur- 

 posely of rather a larger size, were fixed near each other in 

 the clay; and it was observed that the diamond was most 

 luminous while under the action of the flame, and longer 

 so after removal from it. Its being a very slow conductor 

 of heat may occasion in part the latter quality. 



* Annals for May. 



