Washington and Wright — Feldspar from Linosa. 59 



another way, and as it would be contrary to our notions of the 

 definitions of mineral species to consider two chemically similar 

 minerals as distinct which possess essentially identical crystallo- 

 graphic form and physical and optic properties in so many 

 respects, we may also disregard this hypothesis. 



The alternative view is that the mineral is a labradorite of 

 abnormal optic characters and chemical composition, due to the 

 presence of another mineral in intimate molecular mixture as a 

 mixed crystal or solid solution. The amount of this must be 

 small, and in its optic characters the subordinate mineral should 

 presumably resemble a plagioclase, since the optic constants of 

 the Linosa mineral are essentially those of a lime-soda feldspar, 

 but indicate a composition somewhat more sodic than that 

 indicated by the specific gravity. 



The range of possibilities as to the mineral molecule which 

 may be supposed to be admixed molecularly with the labrado- 

 rite is very limited, as it must be, chemically, an alumino- 

 silicate of either soda or lime, or of both of these, with alumina 

 equal molecularly to the basic oxide or oxides, and with the 

 ratio of silica to bases less than in labradorite. Furthermore, 

 it seems necessary to assume that the base is soda alone, because 

 if the subordinate mineral is purely calcic, all the soda entering 

 albite, the lime must be equally distributed between the mole- 

 cules CaO.Al 2 3 .2Si0 2 (anorthite) and CaO.Al 2 3 .Si0 2 , to 

 obtain the ratios and percentages shown by our mineral. This 

 calcic subsilicate is not known to occur either in nature or 

 artificially, its nearest analogue being kornerupine, MgO.Al 2 3 .- 

 Si0 2 , and its existence does not seem probable. Also no 

 mineral in which CaO + Na 2 = A1 2 3 is known which suits 

 the requirements of the case. 



The sodium-aluminum metasilicate, Na 2 0. Al 2 3 .4Si0 2 , occurs 

 in nature as jadeite, and might also presumably exist as an 

 isometric and isotropic soda-leucite. The presence of this 

 molecule would yield a composition identical with that of our 

 mineral, if soda is equally distributed between this and albite, the 

 wdiole forming the mixture: ]^"a 2 O.Al 2 3 .6Si0 2 -f Na 2 O.Al 2 3 .- 

 4Si0 2 + 4(CaO. Al 2 3 .2Si0 2 ). But the presence of either of these 

 mineral molecules may be considered as impossible here, on 

 crystallographic as well as on optic grounds. 



The most probable mineral molecule, therefore, is the ortho- 

 silicate Na 2 O.Al 2 3 .2Si0 2 . Assuming this to be present, the 

 composition of the Linosa mineral in terms of mineral mole- 

 cules can be calculated to be as follows : 



KAlSi 3 (V__. -016) 4-48 



XaAlSi 3 G 8 .._ -137 \ 36-16 



CaAl 2 Si 2 8 '192 9-84 53'78 



NaAlSiO -019 1*00 5'58 



