ON THE CHEMICAL NATURE OF ALLOYS. 183 



and RuSn 3 were obtained. When osmium was dissolved 

 in zinc or tin, and the alloy treated with hydrochloric acid, 

 a residue of pure osmium was left, so that osmium appears 

 to dissolve in these metals without combining with them. 

 When lead or copper were used as solvents, similar evidence 

 of chemical combination was obtained. 



Heycock and Neville have found that when gold is dis- 

 solved in sodium, and the apparently homogeneous alloy is 

 treated with alcohol, and hnally with water, a beautifully 

 crystalline mass is left consisting of lustrous microscopic 

 needles. These are of a pale brassy gold colour, but when 

 heated to about 200° C. they become incandescent and their 

 colour changes to that of pure red gold. After ignition, but 

 not before, the residual sodium can be extracted with 

 water and estimated. The percentage of sodium agrees 

 pretty well with the formula Au 3 Na, but there is never 

 quite enough sodium, probably because the solvent used to 

 remove the excess of that metal attacks the crystals super- 

 ficially. 



By dissolving titanium in aluminium and extracting with 

 hydrochloric acid, Levy 1 has obtained fine crystalline plates 

 which appear to be a mixture of TiAl 4 and TiSi 4 . 



Joannis, 2 by treating the metals lead, bismuth or anti- 

 mony with sodammonium or potassammonium, and removing 

 the excess of alkali metal by washing with liquid ammonia, 

 has obtained the bodies PbK, BiNa 3 , SbNa 3 , PbNa 2 , analysis 

 giving almost exactly the percentage corresponding to these 

 formulae. 



These examples by no means exhaust the list of definite 

 compounds that are known or suspected to exist. But 

 although the evidence in favour of chemical combination is 



o 



pretty conclusive in some of the above cases, it is well to 

 remember that the original ingot of alloy may consist of 

 crystals of the solvent metal imbedded in eutectic alloy, and 

 that this latter, though not a chemical compound, may be 

 such an intimate mixture of the two metals that reagents 

 can only attack it superficially. We should then get a 



1 Comptes Rendus, p. 66, 1888. 

 -Ibid., p. 585, 1892. 



