636 FRANK F. GROUT 
magma, nor any possible mixture. Mr. T. M. Broderick, of Min- 
nesota, has recently checked up the previous work on the chemical 
nature of the rocks by determining the alkali content of all the rocks 
at the contact of gabbro with an inclusion of quartzite. If red 
rock in this situation is derived from quartzite it is derived from 
the immediately adjacent inclusion. The gabbro has 3.49 per cent 
alkali and the quartzite 3.60 per cent, but the red rock has more 
than either, 8.33 per cent. Some differentiation must have 
occurred in this case, and probably in all similar contacts, many 
of which have been listed by Bowen." 
If it is granted that differentiation occurred, it is easily shown 
that local assimilation has been relatively slight. The composition 
of any known differentiate would become evidently hybrid with 
small additions of the sediments near by. 
Deep-seated assimilation and differentiation—syntexis—may 
have occurred, but there is little evidence that it played much part 
in forming red rock. There are several examples of assimilation 
in the gabbro. The quartz gabbro described by Winchell? is near 
an inclusion of quartzite. Similar endomorphic contact effects are 
exposed in other parts of the gabbro. None of the assimilation 
phases resembles red rock. 
It is therefore maintained that the granites associated with the 
Duluth gabbro were dependent more on differentiation than on 
assimilation of acid rocks. The general discussion of the processes 
by which granites separate from gabbro magma would be mostly 
theoretical and too long for this paper, but one theory has been 
forced into prominence by local observations on the gabbro. The 
several occurrences may all be explained by supposing that the 
original magma contained some vapors under pressure and that 
these tended to separate and escape from the main magma, bearing 
with them those acid, alkaline constituents for which they seent to 
have a special affinity. The accumulation of a definite upper zone 
of red rock would then be the result of a quiet rise of the lighter 
vaporous separate under an impervious roof. The aplitic areas 
near the top would be similar gravitative separates, disturbed by 
tN. L. Bowen, ‘“‘The Gowanda Lake District, Ontario,” Jour. Geol., XVIII, 658. 
2 A. N. Winchell, ‘The Gabbroid Rocks of Minnesota,”’ Am. Geol., XXVI, 348. 
