62 Washington and Wright — Feldspar from, Linosa. 



applied to feldspars composed of the three molecules men- 

 tioned in somewhat varying ratios. If more representatives of 

 the series are discovered, these limits might be more sharply 

 defined, centering round the composition of the present case. 



The very close approach to stoichiometric ratios of the 

 oxides may seem to be inconsistent with the view that the 

 mineral is an isomorphous mixture or solid solution. It is, 

 however, in harmony with the well known fact that solid solu- 

 tions tend to form with their elements in simple ratios, in 

 which case they possess greater stability, giving rise to the sub- 

 stances known as molecular compounds. This is exemplified 

 in many mineral groups, such as the plagioclases, pyroxenes, 

 olivines, nephelite, and the calcite group, the intermediate 

 members of which are very apt to show simple ratios of the 

 end components. 



The fact that soda anorthite is not known to occur inde- 

 pendently in nature cannot be brought up against the hypoth- 

 esis of its presence in this case, as the assumption of the 

 presence of a molecule unrepresented by itself in nature is not 

 uncommon in explaining the constitution of many complex 

 mineral groups. Its non-existence as a mineral may be best 

 explained on the ground that the conditions necessary to its 

 formation seldom obtain, the physical conditions being gen- 

 erally such that the more stable nephelite is formed in its 

 stead. 



Assuming the presence of the soda-anorthite molecule, it is 

 certainly remarkable that it does not appear to have been 

 detected as yet among the hundreds of chemical analyses 

 which have been made of the feldspars. It would seem to be 

 certain, at least, that it does not enter into the composition of 

 many of these, so far as known, and that, if present at all, it 

 forms only a very small percentage of the feldspar substance, 

 so that its effect in altering the silica ratios would either be 

 overlooked or attributed to impurities or analytical error. 



It is a plausible, and indeed reasonable, supposition that we 

 have to do here with a case of imperfect isodimorphism or 

 limited miscibility, as it is termed,* instances of which have 

 been investigated by Retgers and others. As stated by Groth : 

 " If the temperature intervals for the stability of the indi- 

 vidual modifications of two substances differ so widely that, 

 under the conditions ruling during the crystallization, the cor- 

 responding state of the one substance is metastable, then as a 

 rule this substance can crystallize along with the other in the 

 form which is stable for it [the other], but only to a limited 

 extent." This is illustrated by monoclinic FeSG 4 .7H 2 and 



* Cf . P. Groth, Introduction to Chemical Crystallography, New York, 

 1906, p. 92. 



