TEE BEHAVIOR OF INCLUSIONS IN IGNEOUS MAGMAS 565 



amount of feldspathoid in a magma already capable of precipitating 

 both of these. Such adjustment of the relative amounts of minerals 

 we have found to be a common effect of inclusions. 



It is probable that alkaline rocks are ordinarily produced by 

 crystallization differentiation from subalkaline magmas. A possible 

 method in the case of a leucite-bearing rock has been demonstrated 

 by Morey and Bowen, who show that a liquid of the composition of 

 orthoclase or even one with a moderate excess of silica over the 

 amount necessary to form orthoclase will precipitate leucite as the 

 iirst-formed crystals.^ If an orthoclase-rich Uquid came into being 

 by fractional crystallization of a more basic magma it might, under 

 the appropriate conditions, show the above effect. The fact that 

 the excess silica must be no more than a small amount should be 

 noted, for this fact renders it possible that limestone may, in spite of 

 the many objections that have been raised above, have some influ- 

 ence in promoting the formation of alkaline rocks. The influence is, 

 however, an emphasizing of a normal tendency rather than a funda- 

 mental necessity. This we have found to be a general rule in con- 

 nection with the effects of inclusions. If the differentiation of the 

 magma which gave rise to the orthoclase-rich liquid took place in 

 the presence of a supply of limestone inclusions this would tend to 

 reduce to a minimum any excess silica that might otherwise be asso- 

 ciated with the orthoclase. Thus the normal tendency of the ortho- 

 clase to break down into leucite under the proper conditions would 

 be free to assert itself. We may therefore accept the possibility that 

 reaction with limestone may emphasize the tendency toward the 

 formation of an alkaline differentiate, though it is not essential to it. 

 Other factors, such as the failure of olivine to form at an early stage 

 in the magma's history, or the free resorption of such olivine as does 

 form, may also assure a low excess of silica at a late stage with likeli- 

 hood of the separation of feldspathoids. The tendency of feld- 

 spathoid to separate under these conditions has been demonstrated 

 as yet only for leucite, but the frequent intimate association of 

 leucitic and nephelitic rocks renders it probable that the factors 

 • governing the formation of the nephehtic rocks are not unrelated. 



^ Amer. Jour. Sci., Vol. IV (1922), pp. 1-21. 



