308 
are generally vertical, and the contact relations on the whole 
are analogous to those of the Ilimausak batholite, and thus in- 
dicate that the igneous masses have replaced corresponding 
portions of the invaded rocks. No indications of a pushing 
aside of the wall rocks have yet been found. 
6. Fragments of crystalline rocks, identical with those of 
the surroundings, are sometimes found enclosed within the 
batholites, but these, on the whole, are rare. Although the sur- 
rounding rocks are entirely crystalline the batholite of Cape 
Desolation at some places encloses numerous sandstone frag- 
ments. 
The salient feature is thus the contrast between the mode 
of occurrence of the dense gabbroid rocks and the spe- 
cifically lighter, more acid rocks. The same contrast, as is 
well known, may also be observed in many other regions of 
similar geological structure. It is true that large gabbro ba- 
tholites are found at several places in other countries, but 
there can be no doubt that the rule which is illustrated so 
conspicuously in southern Greenland, with its many and excel- 
lently exposed Plutonic rock bodies, is of more than local sig- 
nificance’. No theory of intrusion, therefore, can be regarded 
as satisfactory which fails to account for this difference in the 
typical mode of occurrence of the gabbroid and the other 
abyssal rocks. Perhaps both the assimilation theory and the 
theory of subsidence of roof blocks may to some extent meet 
this demand, but the insufficiency of the former theory is shown 
by the fact that the chemical composition of the batholitic rock 
is independent of the wall rock in South Greenland as else- 
where. In addition, direct support is given to the latter 
theory by the finding in the granite of Cape Desolation of 
sandstone fragments, which can only be derived from a sand- 
stone layer formerly overlying the crystalline rocks, and which 
! Compare Dary, Amer. Journ. of science (4), vol. XXVI, 1908, р. 40. 
