RE-EXAMINATION OF AMERICAN MINERALS. 115 



1 was from the cabinet of Professor Silliman, jr. ; No. 2 from 

 the mineralogical collection of Harvard University, received 

 from Professor Cook; No. 3 from the Lederer collection in 

 Yale College. 





Si 



SI 



Mg 



Fe 



Ca 



A 



1. 



4220 



trace 



42.50 



1.56 



trace 



13.28=99.54 



2. 



42.56 



trace 



43.15 



0.95 





12.84=99.50 



3. 



42.10 



trace 



41.23 



1.11 



1.90 



12.77=99.11 



These analyses give the oxygen ratio 4:3:2, and the for- 

 mula 2Mg 3 Si 2 +3MgH 2 , which calculated is 



'Si Mg S 



43.5 43.8 12.7 



This is the composition and formula of serpentine, and the 

 fact of its identity with that species is also borne out by its 

 physical characters. 



The large amount of lime obtained by Bowen was doubtless 

 due to the limestone and tremolite with which it is often very 

 intimately associated ; much care is required to separate these 

 substances entirely from the Bowenite, but the mineral so 

 purified contains no lime. 



7. WlLLIAMSITE, IDENTICAL WITH SERPENTINE. 



We notice that this species is considered distinct by Prof. 

 Shepard in the last edition of his mineralogy, notwithstanding 

 it has been shown to be serpentine by Hermann,* and previ- 

 ously from an analysis made by one of us, published in Dana's 

 Mineralogy, page 692. In this analysis referred to, 3.35 per 

 cent, of alumina and iron were obtained. We have since exam- 

 ined the relative proportions of these substances, and find that 

 the amount was due to iron with but a trace of alumina. Two 

 analyses made from very pure specimens gave 



Si Si Mg Fe Ni H 



1. 41.60 trace 41.11 3.24 0.50 12.70=99.15 



2. 42.60 trace 41.90 1.62 0.40 12.70=99.22 



It is evident from these analyses that the mineral is identical 

 with serpentine, and affords the same formula as the mineral 

 last mentioned. It may be well to remark that great care was 

 taken to see that no magnesia accompanied the oxide of iron 

 in its precipitation by ammonia; not satisfied with adding an 



*J. f. pr. Chem., liii, 31. 



