CORUNDUM AND DUMORTIERITE IN PEGMATITE 483 
“by cementation.”’* Bearing in mind that ‘‘under proper conditions 
water and liquid rock are miscible in all proportions,’’? the criteria 
for judging between pegmatite dikes which originate by igneous 
injection, and those in which heated waters play a larger réle than 
molten rock, must be chiefly those of structure. The pegmatitic dike 
in question, by reason of its eugranitic structure in the field, appears 
to be among those which deal chiefly in molten rock, and are not 
greatly charged with heated waters. It lies very near the granite 
mass along Grape Creek. It tapers rapidly as it runs up the ridge. It 
has sent off occasional stringers into the schist six inches or more in 
width, sharply marked off, as are injections of igneous rock into 
fissures in the form of dikes. These observations support the view 
that the great dike in question is near the igneous end of the pegma- 
tite series. It would appear that the acid magma was in places 
uncommonly rich in alumina, and that the excess of Al,O, over that 
required to satisfy the alkalies and lime for the formation of 
feldspars crystallized as the oxide corundum. The mineral does 
not appear as a filling of cavities in the pegmatite. Under the micro- 
scope it bears the same indigenous aspect as do the other constituents 
of the rock. No evidence was obtained such as would point to a 
replacement of pre-existent minerals by the corundum. It is fresh 
and clear, and does not show itself as an alteration product of some 
other mineral. There is nothing to make it plain that the corundum 
has resulted from some other aluminous mineral by dehydration 
attendant on metamorphism. It is difficult to reconcile its distribu- 
tion in the rock with the assumption that the Al,O, needed for it was 
gained by the working-over in the magma of included portions of the 
wall rock. 
Chemical characteristics oj the pegmatite—The calculation of the 
analyses of persalanes for their norms frequently presents an excess 
of Al,O, to be referred to corundum. In the corundum-bearing 
subclasses of persalane are found quartz felsite, granite, quartz kera- 
tophyre, the Uralose rocks described by Morozewicz under the 
names corundum syenite and corundum pegmatite, and the two 
« C. R. Van Hise, ‘‘ Treatise on Metamorphism,”’ Monograph 47, U.S. Geological 
Survey, p. 723. 
2 [bid. 
