RE-EXAMINATION OF AMERICAN MINERALS. 129 



Having reason to suspect that the substance was not per- 

 fectly homogeneous— from our first analysis not agreeing with 

 Prof. Shepard's— various specimens were examined. The cor- 

 rectness of the supposition will be seen by comparing the 

 following results : 



12 3 4 



Silica 5^80 44.77 44.94 46.70 46.60 44.74 44.10 



Alumina 34.90 35.88 '25.05 31.0110.-,, 20.98 20.64 



Iron 3.01 2.52 3.33 3.69 j 60 " 10 4.27 4.03 



Manganese.. .30 .30 trace trace trace trace 



Lime 66 .52 8.44 trace trace 12.90 12.34 



Magnesia 42 .53 6.86 .50 .50 8.48 8.57 



Potassa 6.87 5.80 11.68 11.68 3.73 S.92 



Soda 3.60 trace trace trace 



Water 5.38 4.72 6.11 5.30 5.30 4.86 6.30 



99.94 100.53 98.88 99.13 99.96 99.90 



There is a remarkable agreement in the percentages of silica. 



The mineral was found to lose about two per cent, of water by 



desiccation. Some specimens showed the presence of a small 



amount of phosphoric acid. Nos. 1 and 4 were received from 



Mr. S. W. Johnson ; No. 2 from Mr. Silas E. Horton, and the 



exact locality of it is Diana, ~N. Y. ; No. 3 was received from 



Prof. Hume, of Charleston, who obtained it from Prof. Shepard. 



The original analysis by Prof. Shepard gave 



Si £1 Fe H CaMg 



47.68 41.50 5.48 4.83 traces 



This substance bears a close relation to agalmatolite, which 

 from the variable proportion of its constituents can not be 

 considered a mineral, but is a rock. Some of the specimens 

 of dysyntribite give the composition of pinite, but it is reason- 

 able to suppose that a mineral varying so much in its alumina, 

 magnesia, lime, and alkalies, may in different masses furnish a 

 resemblance to a vast number of minerals. 



25. GrIBBSITE. 



G-ibbsite was first described by Dr. Torrey* as hydrate 

 of alumina. This composition, confirmed by Dewey and 

 Thomson, was considered correct until Hermann f announced 

 the discovery of over 37 per cent, of phosphoric acid in it, and 

 that the mineral was a hydrous phosphate of alumina. To 



* N. Y. Med. and Phys. Journal, i, 68. 

 t J. f. pr. Chem., xl, 32. 



