DEVELOPMENT OP PHEXOCRYSTS. 267 



From this it is evident that magmas of similar chemical composition, 

 which were erupted at different times, reached positions of like elevation 

 within the volcano in different stages of crystallization. Some carried 

 large crystals of plagioclase, augite, and olivine, some only augite and 

 olivine, and others had no crystals developed in them. For, as shown in 

 the chapter on Electric Peak (Chapter III), the molten magmas must have 

 heen completely fluid when they reached those places in the conduit where 

 the character of crystallization referable to the surrounding conditions 

 affected all of the constituent minerals, including the apatite and zircon. 



From the fact that the magma which was forced into the outlying 

 dikes must have been the advanced portion of that which stopped in the 

 conduit in any particular eruption, and since the dike rocks are more 

 usually porphyritic, and the core or stock rocks show by their micro- 

 structure and mineral development that they were generally completely 

 fluid when they came to rest in the conduit, it may be inferred that the 

 phenocrysts of the dike rocks were formed in the advanced portion of the 

 magma of a particular eruption, and that the rear part of the magma was 

 in most cases free from them. 



The magma having been assumed to be chemically homogeneous, and 

 the influence of pressure not being recognized in the crystallization of 

 these rocks, the most variable condition which remains is the temperature of 

 the rocks through which the magma flowed, and the consequent difference 

 in the rate of cooling of the advanced portion of the magma and of the rear 

 portion. The former would cool more rapidly. While the magma advances 

 through hot rocks it may cool gradually, but when it enters less highly 

 heated rocks the cooling will be more rapid. 



There are many reasons for concluding that the phenocrysts of porphy- 

 ritic rocks are the result of crystallization which has taken place very shortly 

 before the final solidification or crystallization of the whole rock mass, and 

 that they are comparatively rapid growths, and are not minerals that have 

 existed within the molten magma for any considerable length of time prior 

 to its solidification. 



As indicating their rapid growth, we may cite the abundant inclusions 

 of mother liquor with gas cavities, which are of common occurrence in 

 phenocrysts in extrusive rocks; and that they could not have existed for 

 any great length of time within the molten magma is proved by the 



