NATURAL SCIENCES OF PHILADELPHIA. 349 



precipitated as before to insure a perfect separation from the lime 

 and magnesia. From tiie mixed liquor containing the alkaline 

 earths the lime was removed as oxalate; the magnesia finall}- 

 after expelling the ammoniacal salts as ammonio-magnesia phos- 

 phate. The iron and alumina were separated b}^ caustic potassa. 

 The iron also in this case was redissolved, and again thrown 

 down with an excess of caustic potassa to render the separation 

 of the alumina as complete as [)ossil)le. After carefully washing 

 the iron, etc., the alumina was separated from the alkali in the 

 usual way. 



The result of the analysis is as follows : 



Silica (Si) = 21.78%. 



Sesquioxide of chromium (^Fl) = 88.oi " 



" iron (Pe) = 13.29 " 



" aluminum (Al) = .81 " 



Lime (Ca) = 18.58 " 



Magnesia (Mgj = 7.88 " 



Loss on ignition = .11 " 



100.84 



By comparison between the silica and the other oxides, I found 

 the ratio to be as 1 : 2.4 nearly ;"that is, when the sesquioxides and 

 the monoxides are added together in their respective equivalent 

 quotients. But if the quotients of the added sesquioxides and 

 the quotients of the added monoxides are compared with the 

 equivalent quotient of the silica, then the ratio of the silica to 

 the sesquioxides and to the monoxides is as 1 : .9 : 1.4. 



These figures could be used for a complicated formula, which, 

 however, I omit, because it would not give any more correct idea 

 of the combination. 



Finally, I am aware that this mineral has been erroneously 

 called ouvarovite, which, as is well known, is a variety of garnet. 



To show the dilTerence in the composition of the two niinerals 

 I give the results of 0. L. Erdman's analysis of ouvarovite and 

 the composition of the new species Trautwinite : 



That these minerals are widely different from each other is 

 readily seen in the two lines of figures ; besides this we have also 

 the difference of form, garnet being isometric and Trautwinite 

 hexagonal. I endeavored repeatedly to project upon paper, 

 through the camera lucida, points for measuring angles of the 

 new mineral, but the results varied so much among themselves 

 that they were not fit for publication. 



