66 



THOMAS STEERY HUNT ON A NATUEAL SYSTEM IN 



him the discordant and apparently incommensurable numbers, 144.6, 16*7.3, 241.0, llt.O, 

 and 148.0, which fail to show any relations between the species compared. 



These, however, from their close resemblance in external characters, on the one hand, 

 and their chemical diflerences on the other ; from their varying relations between protoxyds, 

 sesquioxyds and silica ; from the partial replacement of silica by boric oxyd, and of alumina 

 by ferric and manganic oxyds, are peculiarly fitted to test the correctness of the new 

 method of study ; and this, the careful determinations of Rammelsberg enable us to apply 

 to the tourmalines with guarantees for accuracy not often to be met with. The results of 

 such a study of these five species, as set down in the table above, may here be stated at 

 greater length. In calculating the mean weight (P) of the oxyd-unit for each species of 

 tourmaline, care has been taken to get the nearest approximation to the results of Rammels- 

 berg's analyses of that species. The manganic oxyd in indicolite and rubellite is included 

 with ferric oxyd, and the various protobases always present are grouped under the heads 

 of ferrous oxyd, magnesia, soda and lithia. The formulas thus arrived at, with their 

 fractional coefficients, and the value of the oxyd-unit, got by dividing the calculated 

 equivalents by the number of units in each formula, are subjoined. But these formulas 

 do not take into account the fact that all of these tourmalines contain from 1.5 to 2.5 of 

 fluorine, replacing a portion of oxygen ; the mean of this, as calculated for coronite, being 

 about one sixtieth in atomic ratio. This proportion woirld giA'e for coronite 1.93, and for 

 rvibellite 1.90 of fluorine, an addition of 0.1 V and 0.18 to the value of P as calculated for 

 the above formulas without fluorine. These values are placed within parentheses, while the 

 corrected values for this mean proportion of fluorine are given in the subjoined table 

 under P, and are those employed for calculating the atomic A^olume of the tourmalines. 



The values of V show for these five species a remarkable agreement, which testifies at 

 once to the correctness of the principles upon which our system is based and to the accu- 

 racy of the determinations of Rammelsberg. 



Tribe 9. — Protopeephylloids. 



§ 8*7. The tribe of the Protoperphylloids presents illustrations of all the principles 

 which we have endeavored to set forth in the accounts of preceding tribes. It shows 

 great variations in the relations of protoxyds to sescjuioxyds, from highly protobasic species 

 like phlogopite and the chlorites, with the ratio 2 : 1, to mvrscovites, in which it is 1 : 12. 

 Ferric and chromic oxyds, in some species, replace more or less completely alumina, and 

 the protobases vary, from species in which alkalies only are present, to others in which 

 alkalies are wanting, their places being partially or wholly supplied by ferrous oxyd, 



