412 E. T. Allen — Composition of Natural Bornite. 



tals which were " covered with a dark-colored coating which 

 was thoroughly removed before the material was crushed for 

 analysis. " " The material from specimen 1 was the lower por- 

 tion of the crystal represented by fig. 1." " The material was, 

 as far as could be determined by the ordinary methods, homo- 

 geneous in every respect. This was later substantiated by a 

 careful metallographic study of specimen 1." That is, No. 4 

 was not examined metallographically at all, and only the basal 

 portion of No. 1 was so examined. This was unfortunate, 

 since the basal portion of sulphide crystals has been found to 

 be especially liable to contain impurities.* The authors admit 

 that a part of the bornite from this same locality (Bristol, Conn.) 

 has the composition Cu 5 FeS 4 , as their analyses of a fragment 

 of specimen S05 of the Brush Collection at Yale University 

 showed. This was one of the specimens analyzed by Harrington. 



Through the kindness of Professors Dana and Ford another 

 fragment of specimen 805 was sent to us. It was examined 

 microscopically by Dr. H. E. Merwin, who reported that seven 

 out of eight pieces of this fragment were practically pure, 

 while the eighth contained about 30 volume per cent of 

 chalcocite ! 



On the basis of the two analyses quoted above and ancient 

 analyses found in the literature, Kraus and Goldsberry set up 

 a long series of copper-iron-sulphide compounds between 

 Fe„S 3 and Cu 2 S. More recently two other articlesf have 

 appeared on this subject, in both of which the evidence of 

 Kraus and Goldsberry is accepted, though other explanations 

 are advanced for it. But since these writers bring forward no 

 new evidence, it will hardly be necessary for us to consider 

 their hypotheses. 



For the sake of completeness, we may include here a recent 

 analysis of the bornite from Yirgilina, Virginia, by Chase 

 Palmer 4 Palmer's material was examined metallographically 



Found Cal. for Cu 5 FeS* 



Cu 62-50 63-33 



Fe 11-64 11-12 



S 25-40 25-55 



99-54 100-00 



and it is noteworthy that the principal impurity was chalcopy- 

 rite. Though the results are in a fair agreement with Cu 6 FeS 4 , 

 " the relatively high proportion of iron suggests the presence 



* Murdoch., loc. cit. , p. 36. 



\ A. F. Eogers. Science, xlii, 386, 1915 ; E. T. Wherry, ibid., p. 570. 



+.J. Wash. Acad. Sci., v, 351, 1915. 



