406 W. P. Headden — Formation of the Alloys of 



That iron and tin readily unite with one another is well 

 known, but the character of the iron and the temperature are 

 important factors. When reduced iron was used the combina- 

 tion ensued quickly and the solution of the iron in the tin was 

 perfect, but when cast iron was used this was not the case ; 

 but the presence of the carbon did not prevent the formation 

 of these alloys. 



"Whether these alloys can endure remelting without decom- 

 position or not was not decisively proven, but the experiments 

 made tend to show that they can be melted by themselves and 

 kept at a temperature sufficient to melt cast iron for an hour 

 without perceptible decomposition, and if melted with tin 

 already saturated with the alloy they may be remelted several 

 times without material change, at least within the limits of 

 temperature under which m}*- experiments were made. While 

 it is evident from my experiments that the frequent remelt- 

 ings to which the reguluses were subjected had at most only a 

 very subordinate influence in determining the alloy formed, 

 it is equally evident that the influence of the ratio of tin to 

 iron is very great ; it appearing to be the determinative con- 

 dition in the formation of the series from FeSn to FeSn 3 , but 

 for the alloys Fe 3 Sn, Fe 4 Sn and Fe 9 Sn it would seem that we 

 will have to look for some other condition as the ratio of iron 

 to tin alone will evidently not suffice in explaining their for- 

 mation. 



It is probable that the alloy first formed during the reduc- 

 tion is not that with the highest ratio for the tin, as is indi- 

 cated by the results of the investigation of the buttons from 

 the cyanid assay which had not been remelted and some of 

 them had been in the fire but eight minutes and still yielded 

 the intermediate members of the series which according to 

 our observations are formed between wide limits in the ratio 

 of tin to iron, and this too when the atomic ratios of the iron 

 to tin in the buttons was such as to justify the expectation of 

 finding an alloy having a higher tin ratio. It is difficult to 

 say in which experiment we approached most nearly to the 

 conditions existing in the original mass of tin but probably in 

 the first in which the regulus resembled pure tin but yielded 

 an alloy having the ratio (Fe : Sn) 1:1; in another in which 

 case the fusion was effected at a higher temperature and con- 

 tinued longer the alloy had the ratio of 1 : 1*25 ; but in this 

 case we had a small amount of a very rich iron alloy with the 

 approximate ratio of 9 : 1 separated from the regulus which 

 might have had some influence upon the ratio of the alloy, 

 especially if this iron alloy is a product of the decomposition 

 of alloys richer in tin which already existed in the mass 

 which, however, I hold to be doubtful. In still another the 

 alloy obtained had the ratio of 1 : 1 and there was a large por- 



