572 N. L. Bowen — The Binary System. 



lization of nephelite from most rock magmas would begin, 

 otherwise carnegieite would be common. 



From certain magmas very rich in 'nephelite,' the crystal- 

 lization of carnegieite might be expected, but even this state- 

 ment needs qualification, for, as we have seen, some nephelite, 

 viz. that from Magnet Cove, may crystallize from its own melt 

 directly as nephelite. 



E. Esch 22 describes an abnormal nephelite from the nephe- 

 linite of Etinde volcano in German Kamerun. This mineral 

 shows inclined extinction and a complicated twinning and is 

 biaxial. In these properties it agrees better with carnegieite 

 than with nephelite. On the other hand, the crystal outline is 

 like that of nephelite and the determined mean index, 1*5376, 

 very close to that of nephelite, but such an index is not im- 

 possible in carnegieite with certain compounds in solid solu- 

 tion. The mineral may possibly be carnegieite, but no definite 

 decision can be arrived at until the possibilities of solid solution 

 in carnegieite have been further investigated. This single in- 

 stance appears to be the only one in which a mineral whose 

 properties suggest carnegieite has been described. 



Thermometry. 



In the fact of its non-appearance in nature, carnegieite is like 

 a great many other allotropic forms of minerals that have been 

 made in the laboratory. These modifications are, in nearly all 

 cases, the high temperature forms and their absence constitutes 

 one of the proofs of the prevailingly low temperatures of 

 natural magmas during crystallization. 



It is evident from the discussion in the preceding paragraph 

 that care must be exercised in putting to use the inversion 

 of nephelite on the geologic thermometer scale. 23 



Minerals of simple, definite composition, Si0 2 , CaSi0 3 , and 

 others have definite inversion points and the limits of stability 

 of each form give definite points on the scale. On the other 

 hand, a mineral of variable composition has a temperature of 

 inversion which varies conjointly. With known composition, 

 however, it will be possible, when sufficient experimental data 

 has accumulated, to set equally definite limits to its stability as 

 is exemplified in the present case of nephelite with variable 

 content of lime. 



There is reason to believe, as has been pointed out, that 

 natural nephelites first crystallized as such. 



In the event of the discovery of carnegieite in a rock its 

 crystallization above 1200°* could probably be asserted with 

 but little fear of error. 



*It is necessary to leave some margin for a possible lowering due to solid 

 solution. 



