J^i^ .H 



On the Houghite of Prof. Shepard. 363 



in the solution, alumina and magnesia were separated by bicar- 

 bonate of soda ; the magnesia weighed as pyrophosphate, the 

 alumina, as such, after solution and reprecipitation by carbonate 

 of ammonia. The insohible residue was treated with hot solu- 

 tion of carbonate of soda, to extract silica. 



Before subjecting the mineral to analysis, a portion in small 

 fragments, the most homogeneous that could be selected, was 

 used for determining its specific gravity. For this purpose after 

 Its weight had been taken^ it was boiled in the water in which it 

 was subsequently weighed^ as air bubbles adhered to it very per- 

 tinaciously. The number 2-175 was obtained; but the mineral 

 was afterward found to contain spinel. 



A quantity of the mineral after being pulverized, was placed 

 m a Liebig's drying tube, and exposed in a current of dry air to 

 a heat of 100^ C. ; it lost water for a long time. The heat was 

 afterward raised to 175^ C, and it continued to lose weight for 

 several days. It was finally submitted to the highest tempera- 

 ture admissible in an oil-bath, 280^ C. (536^ F.), and after more 

 than 100 hours of drying, it ceased to lose weight. As this re- 

 sult had not been foreseen, the original weight of the mineral 

 was not taken ; the loss was at least five per cent. 



Ill this dried portion were found 



*1 l&g £[ (by ilifF.) insol. (Spinel, &c.) Silica. 



l9-t4:j 36-292 24-223 8-458 8-264 8-020=100. 



^ A fresh portion of the mineral including that employed in tak- 

 ing the specific gravity^ was carefully intermingled, pulverized 

 and dried over SO 3 in vacuo. 



Two determniations of carbonic acid gave respectively 6-712 

 and S-094 per cent. ; mean, 7-3S0 per cent. Another portion, the 

 only remaining material, was ignited until it ceased to lose weight. 

 It then contained no carbonic acid. The loss was 40857 pr. ct, 

 It was analyzed with the following results. 



3tl Jig 



30-048 55-467 insol. (spinel and mica) 15-196, Silica, ^race =100-711 



Several water estimations were made, varying from 33 to 41 

 per cent., but I attach no value to them. If we subtract 7-380 

 the mean result for carbonic acid from 40857 the total loss on 

 ignition, we obtain 33-477 as the per-centage of water; rejecting 

 the insoluble minerals and calculating the remaining constituents 

 on lOOj the composition stands as follows : 



23-867 43-839 26-452 5-833 = 99*995 



Excluding, in both analyses, all the ingredients but alumina 

 and magnesia, and reducing these to per-centage relations, we find 

 the following numbers. 



»., 1st Analysis. 2d Analysis. ^I^^"; 



^ 86-238 35*137 85185 



^ 64'763 64-8e2 64'812 



