Foote and Bradley — Chemical Composition of Analcite. 435 



ing the uncertain quality of many of the analyses, however, 

 we have undertaken a new series. 



The material used in the analyses, which will be given below, 

 fulfilled two conditions : it was of ideal quality and it came 

 from widely different localities. All the specimens were well- 

 crystallized and all were colorless or translucent, with one ex- 

 ception, that from the Kerguelen Islands, which was milk-white. 

 The attempt was made to obtain samples from localities which 

 had furnished material for the analyses giving ratios less than 

 4 for silica, as quoted in Table I, but we were not successful. 

 In some cases, the crystals required no special treatment in 

 order to prepare a sample suitable for analysis. In others, 

 where small glassy crystals had been removed from the rock, a 

 thin layer of foreign matter would usually adhere. To entirely 

 free the crystals from this substance, it was found necessary to 

 crush them and after sifting the fragments to a uniform size, 

 to treat them with heavy solution. 



Microscopic examination of all the samples prepared for 

 analysis proved them to be free from inclusions and to have a 

 perfectly homogeneous glassy structure. In several cases, the 

 specific gravity of the mineral was obtained from that of the 

 heavy solution, while in others the ordinary method with a 

 balance was used. The results obtained were as follows : 



Table II. — Specific Gravity of Analcite. 



No. Locality Specific gravity 



1. Two Islands, N. S 2-254 



2. Cyclopean Is 2*260 



3. Kerguelen Is. 2-257 



4. Victoria, Australia 2-219 



5. Michigan, U. S. A. _. 2-223 



6. Montreal, Can 2-231 



It seems desirable to give at least a brief outline of the 

 methods of analysis, although the mineral is a simple one from 

 an analytical standpoint. At the very outset it may be stated 

 that great care was exercised in regard to the purity of the 

 reagents, and platinum vessels were used wherever possible. 

 The mineral being readily decomposed by hydrochloric acid 

 made it possible to omit the ordinary fusion with sodium car- 

 bonate and thus obviate the use of a solid reagent at this point. 

 Two evaporations were made to render silica insoluble, which 

 was treated in the ordinary manner. It may be stated that the 

 specimen from the Kerguelen Islands, which was milk-white in 

 color, decomposed with separation of silica and also with the 

 formation of gelatinous silica when treated with hydrochloric 

 acid. In all other cases the mineral decomposed only with 

 separation of silica. Alumina was precipitated by ammonia, 



