120 H. S. Washington & H. E. Merwin — Mineralogy. 



essentially a mix-crystal of the components Si0 2 , CaO, 

 (Mg, Fe)0, and (Al, Fe) 2 3 " whose saturation bounda- 

 ries are defined according to his tetrahedral projection. 

 According to this view the aluminous augites would differ 

 from the other pyroxenes in the lack of stoichiometric 

 relations or the presence of definite molecules, and would 

 be regarded as indefinite mixtures. Zambonini assumes 

 the presence in the aluminous augites of three general 

 molecules, diopside-hedenbergite (with their compo- 

 nents), acmite-jadeite, and spinel. While we agree with 

 him that the augites are best regarded as made up for the 

 most part or almost wholly of the first two mineral mole- 

 cule groups, yet there are serious difficulties in the way 

 of assuming the presence of a spinel molecule to account 

 for the presence of the sesquioxides. The most important 

 of these, and the only one to be mentioned here, is that 

 combination of the basic RO needed for the spinel would 

 subtract just that amount from the bases needed to sat- 

 isfy the silica in order to conform to the metasilicate 

 ratio and would leave unsatisfied the equivalent amount 

 of silica. This is especially true of the best analyses of 

 the augites, 7 and would seem to be an insuperable objec- 

 tion to Zambonini 's view of the composition of the 

 aluminous augites. 



For various reasons, which it is not necessary to dis- 

 cuss fully here, we assume that, in general, the alumina 

 and ferric oxide (above that needed for acmite-jadeite 

 molecules) are present as such in solid solution with 

 diopside-hedenbergite (with or without clinoenstatite) 

 and acmite-jadeite. As was pointed out many years ago 

 by Piccini, 8 the Tschermak molecule is equivalent to one 

 of (Mg, Fe)O.Si0 2 plus one of alumina or ferric oxide, 

 that is, RSi0 3 + K 2 3 . It should furthermore be noted 

 that the assumption of the Tschermak molecule binds an 

 amount of (Mg, Fe)0, equal to that of the (Al, Fe) 2 3 , 

 which would otherwise enter into the diopside molecule ; 

 this tends to the formation of molecules of wollastonite, 



7 It may be pointed out that a very considerable number of the analyses of 

 augite> used by Boeke (and also by Zambonini), are of very poor quality, 

 either because of imcompleteness (as regards titanium and soda especially), 

 or because of inaccuracy in the execution (such as Doelter's analyses of 

 Cape Verde augites). A much more critical and exacting selection of 

 analyses is necessary for study of the problem. 



8 A. Piccini, Trans. Acad. Lincei, (3), 4, 224, 1880. Cf. Zambonini, op. 

 cit., p. 10. 



