made with a Vieiv to its Improvement. 29 



the glass proved to be oxide of titanium, with a little oxide of 

 iron. These experiments have led us to doubt whether titanium 

 has ever been reduced to the metallic state. From the effects 

 of the heat upon the crucibles, which became soft, and almost 

 fluid, sometimes, in fifteen minutes, we had in fact no reason to 

 suppose the degree of heat inferior to any before obtained by a 

 furnace: — ^that used in these last experiments, was a blast fur- 

 nace, supplied by a constant and powerful stream of air ; the fuel 

 good Staffordshire coke, with a little charcoal ; both Hessian 

 and Cornish crucibles were used, one being carefully luted into 

 another, — and even three have been united, but they could not 

 be made to stand the intense heat. 



Meteoric iron .is, by analysis, always found to contain nickel. 

 The proportions are various, in the specimens that have been 

 chemically examined. The iron from the Arctic regions was 

 "found to contain three per cent, only of nickel, while that from 

 Siberia gave nearly 10 per cent. With the analysis of this last 

 we are favoured by J. G. Children, Esq., and, having permission 

 from that gentleman, we most willingly Insert the account of his 

 very accurate process. 



Thirty-seven grains of Siberian meteoric iron gave 48*27 grains 

 of peroxide of iron, and 4*52 grains of oxide of nickel. Sup- 

 ■posing the equivalent number for nickel to be 28, these quan- 

 tities are equal to Iron 33*69 

 Nickel 3*56 



100-00 

 A second experiment, on 47 grains, gave 61 grains of peroxide 

 of iron = 42"57 iron. The ammoniacal solution of nickel was 

 lost by an accident ; reckoning from the iron, the quantities per 

 cent, are, Iron 90-57 



Nickel 9-42 



99-99 

 A third experiment, on 56 grains, gave 73-OG grains peroxide 



uf 



