Washington and Wright — Feldspar from Linosa. 63 



orthorhombic MgS0 4 .7H 2 0. Monoclinic mixed crystals with 

 the form of the former can be obtained with up to 54 per cent 

 of magnesium sulphate, indicating the existence of a mono- 

 clinic form of this salt. Then a gap occurs, until we obtain 

 orthorhombic mixed crystals with 81 to 100 per cent of the 

 magnesium salt, showing the existence of an orthorhombic 

 ferrous sulphate. Yery unstable monoclinic crystals of the 

 magnesium salt have been prepared, but orthorhombic ferrous 

 sulphate is as yet unknown in the free state. An analogous 

 case is that of rhombohedral sodium nitrate and. orthorhombic 

 silver nitrate. Rhombohedral mixed crystals have been pre- 

 pared containing up to 52*5 per cent of silver nitrate, while 

 orthorhombic crystals containing only up to 4*5 per cent of 

 sodium nitrate have been obtained. In accordance with these 

 results, no orthorhombic modification of sodium nitrate is 

 known, but the pure rhombohedral silver salt is formed from 

 fusion on cooling. 



Analogously we can suppose that Na 2 Al 2 Si 2 8 and CaAl 2 Si 2 8 

 are isoclimorphous, each forming hexagonal and triclinic modi- 

 fications. Of these, however, only the hexagonal form of the 

 sodium salt and the triclinic form of the calcium salt are stable 

 under ordinary pyrogenetic conditions, while the converse 

 forms are metastable and capable of existence in mixed crystals 

 with the other only in small amount and within a very narrow 

 range of temperature or other physical conditions. The cal- 

 cium almost always present in nephelite, up to about two per 

 cent, may be thus regarded as existent as hexagonal calcium 

 nephelite,* which must possess a. very limited degree of stabil- 

 ity, while the stability of the triclinic modification of the 

 sodium salt is apparently somewhat greater, to judge from the 

 percentage shown by anemousite. 



It is obvious that the existence of soda anorthite and its 

 presence in the lime-soda feldspars, or the possibility of the 

 assumption by these of other molecules in solid solution, would 

 have a very important influence on determinative mineralogy 

 and petrography. The optic characters of such an abnormal 

 feldspar would not indicate its true chemical composition in 

 accordance with the tables and diagrams in use at present, as 

 the feldspar would be apparently more sodic than it is in reality. 

 The optic determination of the soda-lime feldspars in thin sec- 

 tion would thus not be the comparatively simple and unerring 

 matter that it is now supposed to be, as the possibility of the 

 presence of soda anorthite and its influence on the optic con- 

 stants would have to be taken into consideration. The recogni- 

 tion of celsian introduces a similar uncertainty. 



* Morozewicz (op. cit., p. 988) and others consider the calcium as replac- 

 ing the sodium in nephelite. 



