312 BISHOPVILLE METEORIC STONE. 



small fragment of caustic potash placed on the top of the mixed 

 powders in the crucible.* After fusion the analysis was pro- 

 ceeded with in the ordinary way. The results of two analyses 



were as follows : 



1 2 



Silica 60.12 59.83 



Magnesia 39.45 39.22 



Peroxide of iron .30 .50 



Soda, with feeble potash and strong lithia reaction, .74 .74 



100.61 100.29 

 The minute quantity of peroxide of iron came from exceed- 

 ingly fine particles of iron diffused through the minerals, and 

 could be seen by a magnifying-glass. One separate analysis 

 was made for the soda. The constitution of the mineral, as 

 made out from the numbers in analysis 1, is 



Oxygen. Oxy. ratio. 



Silica 31.22 2 



Magnesia 15 51) -. 



Soda 19/ 



corresponding to the formula Mg 3 Si 2 , equivalent to the general 

 formula of pyroxene, R 3 Si 2 . 



The excess of silica obtained by Professor Shepard in his 

 analysis is doubtless due to an imperfect fusion of the mineral 

 with the carbonate of soda, an error easily made if the precau- 

 tions I have already mentioned are not attended to. 



"Chladnite" approaches those forms of pyroxene known as 

 white augite, diopside, white coccolite, etc. ; these last-named 

 minerals having a part of the magnesia replaced by lime. It 

 is identical in composition with enstatite of Kenngott, a pyrox- 

 enic mineral from Aloysthal. in Moravia. 



From these observations it will be seen that the Bishopville 

 meteoric stone, however different in external characteristics 

 from other smaller bodies, is after all identical with the great 

 family of pyroxenic meteoric stones. 



*I would remark that I seldom or ever fuse a silicate with the alkaline 

 carbonates without the addition of a small piece of caustic potash or soda, 

 and never analyze a known or supposed pyroxene or hornblende without 

 this precaution. I have no doubt that there are many minerals classified 

 with hornblende which properly belong to pyroxene, the silica in the analyses 

 being rated too high, an error arising from an imperfect fusion. 



