24 H. 8. Washington and H. E. Merwin — Note on 



Comparison ivith pyroxenite. — It is not necessary here 

 to discuss my analysis of the Vesuvius augite, partly 

 because of its being based on impure material, and partly 

 because it will be discussed later, when the series of Ital- 

 ian augite s is more nearly complete. Attention must, 

 however, be called to the very remarkable correspondence 

 between it and an analysis by Pisani of a pyroxenite of 

 Monte Somma, published by Lacroix, which is given in 

 column 4 of Table I. Lacroix does not describe this rock 

 in detail, but it would appear to be holocrystalline and 

 composed almost entirely of pyroxene. It is a somewhat 

 notcAvorthy example of the possibility of diverse crystalli- 

 zation of a part of a magma, either as a monomineralic 

 or almost monomineralic, granular rock, or as well- 

 formed crystals of a definite mineral. 



Lacroix mentions and gives analyses of several types 

 of the pyroxenite, all of which are more or less closely re- 

 lated chemically. It occurs, according to him,'^ as homoe- 

 ogenic rocks, that is, "granular forms, not only of the 

 flows, but also of the differentiated portions of the same 

 magma which have not necessarily reached the surface." 

 Such homoeogenic rocks are assumedly of abyssal origin 

 and, if the general theories of gravitative adjustment of 

 Daly and Bowen are correct, they would be expected to 

 have come up, or been carried up, as broken-off blocks, 

 from very deep down in the volcanic mass. My augite 

 crystals, on the other hand, occur, as has been said, in the 

 light scoria s that form the uppermost scum or froth of the 

 ascending magma. This would indicate that, in this case 

 at least, separation by gravity has not been carried to 

 completeness. That this is also true elsewhere is indi- 

 cated by the occurrence of such augite crystals (almost 

 always of the same crystal form) at other volcanoes, 

 such as Stromboli, Etna, and the Alban volcano. Indeed, 

 they are rather common at many volcanoes, and similarly 

 crystals of cossyrite and kaersutite (sodic hornblende) 

 are met with as such loose crystals at Pantelleria and 

 Linosa, to speak only of Italian volcanoes. 



But the incompleteness of such a separative process is 

 not to be wondered at, considering the convection currents 

 in, and the presumably violent movements of, the upward- 

 welling mass of magma, as well as the presence of gas, 

 whose bubbles would act like those in a glass of cham- 



^A. Lacroix, C. R., 165, 205, 1917. 



