AN i 
ÅA N:oM) Igneous Rocks of Sviatoy Noss in Transbaikalia. 93 
was lower than was required, under the prevailing pressure, 
to allow the reaction between silica and calcium carbonate 
under liberation of carbon dioxide to take place. 
In the inclusions, on the other hand, the conditions for 
metasomatic replacement were more favorable, probably 
because the temperature in the inner parts of the magma 
masses, to which the inclusions sank, was higher. As another 
possible assumption it may, however, be inferred that the 
formation of skarn might have taken place at the contacts 
at some earlier stages and the skarn-blocks afterwards sank 
down in the magma. I am not able to decide which of these 
two hypotheses is more plausible, and it is also of minor inte- 
rest. The cardinal point here is that the subsequent process 
of assimilation of the skarn presents itself as a consequence 
vEm agmatic stopping 
When the metasomatic alteration of limestone to skarn 
had proceeded to a certain extent, the rock had become 
much heavier than originally, the specific gravity of limestone, 
when altered into andradite, increasing from 2.7 to 3.8. 
The blocks therefore began to sink (or continued to sink) 
down in the magma. By the increasing heat they were now 
more or less resorbed depending upon the amount of super- 
heat disposable. Those blocks sinking down to deeper levels 
were assimilated completely. Currents resulting from the 
upward intrusion and movements of the earth-crust, aided 
by diffusion, restored the homogeneity of the magma. 
Doubtless as it is, that assimilation has mainly taken 
place in the andradite-skarn, no less certainly the calcite 
also has been assimilated, as it is present in a great abundance 
in the skarn blocks. 
The assumption of a large scale production of skarn 
postulates very considerable changes in the composition of 
the magma as a consequence of the deprivation of ferric 
oxide and silica needed to form andradite-skarn. As we 
have seen above, the sviatonossite is geologically connected 
with the surrounding granites, and it lies close at hand to 
assume only the granite magma to have been the parent 
magma of the sviatonossite. This deduction from the field 
