18S5.1 33 [Genth. 



been adopted by maDy chemists and mineralogists, although many facts 

 do not sustain it. 



Bismuth and tellurium are not strictly isomorphous. It is true that 

 both crystallize in rhombohedra of nearly the same angles ; bismuth, how- 

 ever, has an eminently basal and rhombohedral cleavage, while tellurium 

 shows a very imperfect basal and no rhombohedral at all, but a very per- 

 fect cleavage parallel to the planes of an hexagonal prism. 



There is also in all the tetradymites, excepting the two from Fluvanna 

 county, Va., and from Highland, Montana, a portion of the tellurium re- 

 placed by sulphur, and, if therefore tellurium replaces bismuth, sulphur 

 necessarily does it also.* 



That tetradymite is not a native bismuth, mixed with an indefinite quan- 

 tity of tellurium, becomes more than probable from the fact that all reliable 

 analyses agree very closely with the formulae of either of the two modifi- 

 cations, viz : Bi 2 Te 3 or Bi 2 S 3 -j- 2 Bi 2 Te, ; there are only the Cumberland 

 (England) tetradymite, which, according to Rammelsberg, contains : Bi 

 = 84.33, Te = 6.73, and S = 6.43, and the jos&te, for the expression 

 of a rational composition of which we must look for another explana- 

 tion. 



This seems to be very easy, if G. Rose's suggestion would be reversed, 

 and that, instead of making tellurium (and sulphur) to replace bismuth, 

 we make the latter substitute tellurium and sulphur. 



This view is supported by numerous examples, and, if we examine 

 the constitution of the natural sulphides, tellurides, arsenides, &c, &c, 

 we find such substitutions very frequently ; the hexagonal millerite, Ni S, 

 becomes niccolite, NiAs, or breithauptite, NiSb ; the isometric pyrite 

 FeS 2 , by substituting the greater portion of the iron by cobalt or nickel, 

 smaltite (CoNiFe) As 2 , or chloanthite (NiCoFe) As 2 or bismuth-chloan- 

 thite (NiCoFe) (AsBi) 2 ; the rhombic markasite, FeS 2 , in the same man- 

 ner gives : lollingite FeAs 2 , safilorite (CoFeNi) As 2 , and rammelsbergite 

 (NiCoFe) As 2 ; or, if only a portion of the sulphur is replaced, we get as 

 analogues for pyrite : cobaltite CoAsS, Ullmannite NiAsS or corynite Ni 

 (SbAs) S, and for the rhombic marcasite wolfachite (NiFe) (AsSSb) 2 , mis- 

 pickel Fe (AsS) 2 and alloclasite (CoFe) (BiAs) S. 



In the sulphosalts the substitution of bismuth for arsenic and antimony 

 is still more frequent, but it suffices that in the examples given it is shown 

 that sulphur is very often replaced by arsenic and antimony, and that bis- 

 muth, being analogous to these, can therefore replace sulphur and tellu- 

 rium as well. 



These views applied to tetradymite and allied minerals would lead to 



*I have already repeatedly called attention to the fact that the analysis 

 of the Virginia tetradymite, made by Coleman Fisher, in which he found 7.23 

 p. c. of selenium, was made with a part of the identical material which I have 

 analyzed, and which contains not more than a trace of selenium. Notwith- 

 standing these statements it seems to be impossible to eradicate this error, as I 

 find it continually repeated in our best books on Mineralogy. 



PROC. AMEB. FHILOS. SOC. XXIII. 121. E. PRINTED OCTOBER 12, 1885. 



