82 JOURNAL OF GEOLOGY— SUPPLEMENT 



Instead of decreasing basicity or increasing acidity it would be more 

 correct to say increasing alkalinity, for in the later stages increasing 

 acidity of the liquid does not hold. Increasing alkalinity appears, 

 however, to be maintained throughout. 



The question of the fate of sinking crystals may be discussed 

 at this point because it bears directly on the order of intrusion. 

 Some have supposed that solution of crystals formed in the cooler 

 upper portion will take place freely in the warmer lower portion 

 into which they may settle. It is even imagined that complete 

 re-solution of all the crystals may take place with the formation of 

 a magma basin entirely Hquid, but, in virtue of the foregoing action, 

 much enriched in its lower parts in the products of early crystalli- 

 zation. A certain amount of the re-solution postulated must in- 

 deed take place, but there is a Hmit to the process. The re-solution 

 of crystals would usually require a relatively large amount of heat 

 and the heat available is Kmited on account of the relatively 

 small specific heats of silicate Hquids and on account of the fact 

 that the actual downward increase of temperature must be small, 

 for significant sinking of crystals takes place only in those parts of a 

 magma body that are slowly cooled and therefore exhibit small 

 temperature gradients. The re-solution of a small proportion of 

 crystals by a certain layer of the magma should lower its tempera- 

 ture very considerably and, at the same time, raise the temperature 

 at which the hquid becomes saturated with crystals. Shortly, 

 therefore, the liquid is cooled to the temperature at which, in its 

 slightly enriched condition, it is saturated with these crystals and 

 the process of re-solution ceases. The extreme condition of com- 

 plete re-solution of crystals approximately keeping pace with their 

 arrival from higher layers could scarcely be even approached. 

 Whatever crystals sink from upper layers remain, for the most part, 

 as such. 



This deduction is entirely confirmed, indeed it finds its geologic 

 expression in the "normal order of decreasing basicity." If it were 

 common to have magma basins still entirely hquid but much 

 enriched in their lower portions in basic material through re-solution 

 of crystals, the normal order would be from acid to basic, for all 

 portions of the magma would be freely eruptible and the upper, 



