34 PHILOSOPHICAL SOCIETY OF WASHINGTON. 
to which they belong, is not merely determined by accidental cir- 
cumstances, but has its inception in the magma before the eruption 
takes place. By the process of crystallization magmas are fre- 
quently divided into a crystalline solid portion, and an amorphous 
more or less fluent portion. Basic minerals are the first to crys- 
tallize, so that as the process advances the amorphous remnant of 
the magma becomes more and more silicious. The crystals are 
generally thoroughly intermingled with the amorphous magma, and 
in the latter are accumulated nearly all of the absorbed gases 
under great tension, so that when the pressure is relieved it may 
be blown to fine silicious dust, which may be carried by the wind 
many miles from its source, while the solid crystalline portion will 
contribute chiefly to the formation of sand, and be precipitated 
comparatively near the crater from which it issued. 
In cases where no previous crystallization has taken place in the 
magma before it comes to violent eruption, the voleanic dust then 
formed will have about the same chemical composition as the lava 
to which it belongs. Myr. Russell has recently described an inter- 
esting case of this kind in the western part of the Great Basin. 
It appears to be generally true that if other conditions are favor- 
able the difference in chemical composition between volcanic sand 
and dust is directly proportional to the amount of crystallization 
in the magma before its ejection. 
The basic character of the Unalashka sand may be explained by 
supposing that the silicious portion of the magma was carried away 
in the form of dust. ‘ 
The source of this sand is supposed by the collector, Mr. Apple- 
gate, the Signal Service Observer at Unalashka, to have been the 
new crater formed last autumn, near the Island of Bogosloff, about 
sixty miles away. 
Mr. Durron spoke in commendation and amplification of Mr. 
Diller’s contribution to geologic philosophy. Mr. Datu described 
the geographic relations of the volcano from which the Unalashkan 
dust was presumably derived, showing the improbability of the 
eruption having been directly observed. He spoke also of the dis- 
tribution of the Aleutian volcanoes and the lithologic characters 
of their ejectamenta. - 
There ensued a general discussion of the nature and properties 
