Effect of Copper- Tin Alloys. 553 



Hughes *. The figures given, can however, only be con- 

 sidered to represent the general relation among themselves of 

 the various alloys in the series, as the zinc scale does not give 

 absolute values. 



The induction-balance curve, as I have previously pointed 

 out, bears an evident relation to the curve representing the 

 density of the same alloys ; but its divergence from Matthies- 

 sen's curvef of conductivity is singular. This may perhaps 

 be explained by the fact that Matthiessen does not appear to 

 have examined any alloy between those which contain respec- 

 tively 16*4 and 85*1 volumes per cent, of copper, probably 

 because the alloys between these points are too brittle to 

 permit their being formed into wire. It may be noted, how- 

 ever, that the curve of the tin-gold alloys, which belong to 

 the same group as the copper-tin, is given in his well-known 

 paper % in 1860, and does not show an unbroken line in the 

 horizontal portion. Certain intermediate alloys, in the form 

 of rods, are now being examined by Dr. Lodge, whose results 

 will doubtless clear up the point. 



The alloys which occupy the critical points of the induc- 

 tion-balance curve are very interesting. They may be repre- 

 sented by the formulae SnCu 3 and SnCu 4 respectively ; and 

 although they only vary by a single equivalent, or by 6 '49 

 per cent, of copper, their structure and appearance differ 

 widely. The latter, SnCu 4 , is a speculum-metal. It has a 

 large conchoidal fracture and a yellow-grey tint. SnCu 3 , on 

 the other hand, has a blue-grey colour and a coarse surface 

 of interrupted crystalline planes. By successive additions of 

 copper, this alloy seems to pass into the other without any 

 sharp break. 



Possibly both alloys are chemical combinations; and the 

 difference of their ordinates probably marks a different allo- 

 typic state. For further information on such questions, how- 

 ever, we may look with confidence to Prof. Hughes's beautiful 

 and simple instrument, which will also help us to determine 

 whether the relation between conductivity for heat and elec- 

 tricity is really as exact as it has hitherto been supposed to be. 



* Phil. Mag. (5) vol. viii. p. 50. 



t Brit. Assoc. Report, 1863, p. 37, and Chem. Soc. Journ. 1867, p. 212. 



X Phil. Trans, vol. cl. p. 161. 



Phil. Mag. S. 5, No. 52. SappL Vol. 8, 2 P 



