Bornite and its Relation to Other Sulp ho- Minerals. 543 



The above analysis leads to the formula Cu^Fe^. This 

 formula differs materially from the one, Cu^Fe.^Sg, derived 

 by Harrington from an analysis of crystallized bornite from 

 this same locality, namely Bristol, Connecticut. 



Inasmuch as we were confident that our material was well 

 crystallized and homogeneous, and the methods of analysis 

 could not be questioned, it was thought desirable to obtain and 

 analyze some of the material used by Harrington for his 

 analysis. Through the courtesy of Professor W. E. Ford a suf- 

 ficient amount of bornite from specimen No. 805 of the Brush 

 Collection of minerals of Yale University was placed at our 

 disposal. This material was from the same specimen from 

 which Harrington obtained his material. That the two 

 analyses agree very closely is clearly shown by the following 

 comparison : 



Harrington Kraus and Goldsberry 



Ratios Ratios 



Copper. .. 63-24$ 0-9948 10-000 63-18$ 0-9939 10*000 



Iron 11-20 0-2005 2-016 1T38 0-2038 2-051 



Sulphur.. 25-54 0'7964 8'005 25-43 , '7929 7*978 



99-98 99-99 



Since our analysis of this material agrees very closely with 

 that made by Harrington and both lead to the formula 

 Cu 10 Fe 2 S 8 , we are confronted by the fact that here is a mineral, 

 specimens of which from a single locality show a variable com- 

 position. It should be stated at this point that the material 

 used by Harrington, and also by us in checking his analysis, 

 was from a large specimen a portion of which is very well 

 crystallized. Harrington had determined the specific gravity 

 of this material to be 5-072. It will also be noted that this 

 value is somewhat lower than that obtained by us on our 

 material from Bristol, Connecticut, namely 5'086. 



Metattographic Investigation. 

 In order to be certain that the material used for these 

 analyses was homogeneous, surfaces were ground at random 

 on our specimen No. 1 and on one of the larger fragments used 

 for the check analysis of the Harrington material. This metal- 

 lographic investigation was carried out on specimen No. 1 sub- 

 sequent to the making of the analysis leading to Cu 12 Fe.,S 9 , but 

 on the Harrington material before the analysis was begun. 

 The surfaces were polished and examined under a metallo- 

 graphic microscope with magnifications of 85, 135, 200, and 

 about 400 diameters without revealing the presence of any in- 



