58 Washington and Wright — Feldspar from Linosa. 



those of the lime-soda feldspars. The specific gravity, the 

 crystallographic angles measured and the birefringences are 

 those of a labradorite of about the composition Ab 1 An 1 . 

 Although the optic axial angle and the extinctions are decid- 

 edly variable, they correspond to those of audesines somewhat 

 more sodic than A^An^ on the average about Ab 3 An 2 . On 

 the other hand, the chemical composition is not that of any 

 possible member of the normal plagioclase series, or mixtures 

 of albite and anorthite. The ratios of A1 2 3 , CaO and Na 2 

 are those of Ab 2 Au 1? but the amount of Si0 2 is lower than that 

 demanded for these by the known feldspar formulas, and the 

 mutual ratios of this with those of the other constituents indi- 

 cate a composition which corresponds to that of an anhydrous 

 mesolite, or a salt of the acid H 2 Al 2 Si 3 O 10 , with Na : Ca •= 1:1. 

 Our mineral is, therefore, physically closely allied to labrado- 

 rite and andesine, but chemical distinctly different in the ratios 

 of the constituents. 



The possibility that the material analyzed was a mechanical 

 mixture of particles of two minerals, such as labradorite and 

 nephelite, and that only fragments of the former were subjected 

 to optic investigation, is rendered untenable by the following 

 facts. Careful examination of the unbroken crystals, as well 

 as microscopic study of the crushed fragments and cleavage 

 flakes, by both of the authors independently, revealed the 

 presence of but one mineral, colorless, transparent, cleavable 

 and feldspar-like. Apart from the small, opaque inclusions, 

 each crystal appeared to be homogeneous, and all appeared to 

 be of identical material, except for the optic variations. The 

 action of hydrochloric acid showed that no readily decom- 

 posable mineral was present as separate individuals. The very 

 close agreement between separate portions in specific gravity 

 and in chemical composition renders the mathematical chances 

 against the hypothesis of a mechanical mixture of particles of 

 two minerals so ^reat that it may be safely eliminated from 

 consideration. Finally, the lavas of Linosa are all typical 

 feldspar basalts, and only a few very small amounts of nephe- 

 lite possibly existent as a glassy base. . Assuming, therefore, 

 that the material was homogeneous, two hypotheses present 

 themselves to account for the anomalies observed. 



One is that the Linosa mineral is to be regarded as a distinct 

 species, chemically, of the formula Na 2 Ca 2 Al 6 Si 9 3 „ but with 

 physical properties which correspond very closely to those of 

 a plagioclase of the composition Al^A^ to Ab 3 An 2 . The 

 uniformity of the material as shown by the specific gravity 

 and the chemical analyses and, above all, the very close approach 

 to exact rationality of all the ratios, are in favor of this view. 

 But the peculiarities of chemical composition are explicable in 



