50 THOMAS STEERY HUNT ON A NATUEAL SYSTEM IN 



The sparry orthorhombic picrosmiue, with D = 266, which is sometimes librous and 

 asbestiform, is a hydrous bisilicate represented by (mg-,si2)o3+ }aq, and Terreil has very 

 recently described as chrysotile from Canada, with D := 2'56, an asbestiform silicate which 

 is at once more basic, more hydrous and denser than ordinary chrysotile, and approaches 

 matricite in composition. His analysis gives silica :1710, magnesia 39'94, ferrous oxyd, 

 b'^'B, ahimina traces, water 16'8o := 09"62. This corresponds very closely to (mg^.-/e„.;iSis)0|.-, 

 + 6aq.' These Avarions prismatic hydrous silicates of magnesia, including chrysotile and 

 picrolite, constitute an important group of what may be designated as magnesian pec- 

 tolitoids, Avhich have for the most part an atomic volume approaching to dioptase and to 

 datolite, and demand further study, but with the exception of chrysotile have not been 

 placed in our table. 



§ 60. In the accompanying table (No. II) of the principal Pectolitoids are given their 

 atomic formulas as deduced from chemical analysis, the unit-weight, P, calculated from 

 these, the density, D, (water ^ 1.00) and the atomic volume, V, = P-rD. In calculating 

 the value of P for these silicates, we have to consider that two or more protoxyd bases are 

 often present, and that the proportions of these miTst be estimated as nearly as possible. 

 As the specific gravity of species is in many cases inexactly determined, we have, where 

 more than one value of D is given by mineralogists, selected that which seemed most 

 probably correct, and, where determinations of density are wanting, have left a blank in 

 the table. In a case like friedelite, a manganesian silicate including some magnesia, 

 which has been considered in calculating the value of P, we have indicated in the right-hand 

 column of the table the relative proportions of the two bases. Where the species includes 

 several bases, these are represented in the atomic formula by the g'cneral symbol m, and 

 the specification of their nature and their proportions follows in parenthesis. 



§ 61. It has been thought well, for reasons which will be apparent when we compare 

 the pectolitoids with other tribes, to represent their contained water by the symbol aq, 

 preceded by the sign +. It will be noted that in the atomic formulas here employed, the 

 symbols of the metals, with those of silicon, boron and titanium, are placed within paren- 

 theses, and those of oxygen, sulphur, fluorine and chlorine, together with water, without. 

 From this it will be clear that the atomic weight deduced from these formulas must, in 

 order to arrive at P (the weight of the atomic unit), be divided by the number of these 

 units ; that is to say, by the sum of the coetficients of the elements outside of the parenthe- 

 sis. The present table is far from complete ; the determinations of density are in many 

 cases uncertain, those assigned to the same species by different observers often presenting 

 wide variations. Again, the value of P in cerite is calculated as if it were simply a 

 silicate of cerium, while it contains unknown proportions of lanthanum, didymium and 

 samarii^m. The general agreement in the value of V is noticeable, save in two cases, — that 

 of dioptase, for which another recorded determination of D = 3.28 gives V = 6.00, and 

 that of datolite, whose volume shoAvs a condensation approaching to that of the adaman- 

 toid protosilicates. 



§ 62. For the better understanding of the formulas given in the accompanying tables 

 of the various tribes of silicates, it may be well to recall the A^alues of the chemical symbols 

 here employed, which are atomic, — the small letters representing atoms of the elements. 



' Compte Rendu de I'Acad. des Sciences, Jan. 26, 188.5. 



