309 
must, accordingly, have sunk many hundred meters through 
the magma. 
All the batholitic intrusions of southern Greenland, thus, 
are therefore believed to have mainly come to place by the sub- 
sidence of roof blocks, that is by the same process of intrusion 
which brought the Ilimausak batholite. 
From the fact that dykes of ordinary granites and syenites 
are extremely rare, while batholites of these rocks are common, 
we may conclude that, as a rule, the batholitic magmas of this 
area did not possess the relatively high degree of liquidity 
which is necessary for dyke intrusions. This implies also 
that the process of magmatic stoping must have acted very 
slowly. 
The question of the final fate of the sunken roof blocks 
has been hitherto entirely neglected. The Paleozoic batholites of 
South Greenland supply no information on this matter, for the 
blocks, with very few exceptions, have sunk down beyond the range 
of observation. Some indications may perhaps be obtained in 
another way. Within the immense areas of pre-Cambrian granites 
in South Greenland fragments of crystalline schists and gneis- 
ses of widely different sizes occur abundantly in many places. 
This may perhaps suggest that at greater depths some of the 
rock fragments, without necessarily being assimilated, may cease 
to sink. 
In the general problem of the mechanics of igneous intru- 
sions the distinction drawn by Harker’, between intrusions 
which are connected with tangential pressures of the part of 
the earth crust which is invaded and those which are not, is no 
doubt of fundamental significance. It must, therefore, be borne 
in mind that the above discussion refers exclusively to intru- 
sions of the latter kind, which occur under conditions allowing 
of the formation of a large number of massive vertical dykes. 
1 Natural History of Igneous Rocks, 1909, р. 60. 
