14 REPORT 18G6. 



was the first to point out this fact "with regard to alloys, in a paper on the 

 alloys of antimony and zinc. He showed that those containing 43 to 64 per 

 cent, of zinc all crystallize in the same foim, but differently from the other 

 alloys. With the alloys of gold and tin it has been shown* that well-de- 

 fined crystals are not limited to definite proportions of the constituents, but 

 are common to all gold and tin alloys containing from 27 to 43 per cent, of 

 gold; and that crystals and mother-liquor never are of the same composition. 

 Storer t has also experimentally proved that aU the copper and zinc alloys 

 crystallize in the same form. 



These facts show that the ciystaUine alloys of carbon and iron do not prove 

 the existence of chemical combination between them, more especially when 

 we consider that several definite crystalline compounds of carbon and iron 

 have been obtained. In all probability, by altering the conditions of cooling, 

 &c., crystals of iron containing various amounts of carbon may be obtained 

 from the same sample of cast iron. For the chemical combination theory of 

 cast iron we have — 



(1) The evolution of carburetted hydi'ogens when white cast iron is treated 

 with dilute acids. 



(2 The existence of definite crystalline forms of carbon and iron. 



Against it — 



(1) The analogy of the alloys of carbon and iron with the other aUoys. 



(2) The fact of carbon in such a fine state of division that it may exist in 

 white iron (solidified solution) and may be able to unite with hydrogen at 

 the moment of being set free (as in the case of platinum-silver alloys, where 

 a portion of the platinum is dissolved by nitric acid). 



(3) We know of no other case where with two elements the different rates 

 of cooling determine the chemical combination or decomposition (conversion 

 of white into grey and grey into white cast iron). 



(4) That with alloys definite crystalline forms are not necessarily chemical 

 combinations. 



That iron in a molten state will not dissolve more than about 5 ]ier cent, of 

 carbon is analogous to the cases of lead and zinc, bismuth and zinc, ii:ercury and 

 zinc. For it has been shown t that pure lead will only dissolve 1-0 per cent, 

 a pui-e zinc, and pure zinc 1 -2 per cent, of pure lead ; that pure zinc will 

 only dissolve 2-4 per cent, pure bismuth, and pure bismuth from 8-6 to 14-3 

 per cent, pure zinc§. 



Now, although no actual determinations have been made, we may stxppose 

 that the solvent power in the above-mentioned cases of the one metal for the 

 other at higher temperatures will be greater than at lower ones ; hence, for 

 instance, zinc (containing a small percentage of lead) might possibly be 

 made to crystallize from a molten alloy of lead and zinc, containing, say, from 

 2 to 2-5 per cent. zinc. 



Supposing the metals lead and zinc be melted together in equal parts, 

 what takes place ? The zinc (according to the temperature) takes up a cer- 

 tain amount of lead, and floats upon the lead which has taken up a certain 

 amount (according to the temperature) of the zinc. If these two alloys were 

 intimately mixed together (bj" stirring or shaking) and cooled rapidly, we 

 might suppose that under certain conditions an almost homogeneous mixtui-e 

 might be obtained ; or supposing we were rapidly to cool a solution of zinc 



* Memoii'S of the American Academy (new series), vol. T. p. 337. 



t Proc. Eoy. Soc. vol. xi. p. 433. ' % Ihul. p. 430. § lUd. 



