E. Greenly — Diffusion of Granite into Schists. 211 



of temperature diffusion is greatly accelerated, (3) that if the 

 temperature is kept permanently above the fusion-point of one of 

 the substances, diffusion is still further accelerated. 



It may be asked why there is any necessity to appeal to these 

 results, seeing that we have long known that igneous magmas are 

 intruded in a liquid state, and into remarkably narrow veins. But 

 if ordinary igneous intrusion can account for all the phenomena 

 under consideration, why do we not find permeation zones surrounding 

 all intrusive rocks, even ordinary basalt dykes, for tliere is reason 

 to believe that basic magmas have a high degree of liquidity ? 



On the other hand, it is clear that solid diffusion does not take 

 place between rocks by mere close adpression, even after long periods 

 of time, for junctions of normal igneous, even plutonic masses with 

 sedimentary rocks of all ages, as well as junctions of igneou? and 

 sedimentary rocks with one another, can be seen at which no 

 permeation whatever has taken place. Rocks do not, it is evident, 

 diffuse with the ease that gold and lead do. 



It is clear that another factor must be necessary, and this can be 

 found, I believe, in the existence of an already high temperature in the 

 surrounding rocJcs. 



Ordinary igneous intrusions, as is shown by their chilled edges, 

 found the rocks into which they were injected relatively cold, i.e., not 

 appreciably above the temperature proper to a zone of the earth- 

 crust far outside that from which the magma came. They cooled, 

 therefore, at the margin soon after injection, and did not remain in 

 contact at a high temperature for any length of time. 



But there is abundant evidence in permeation regions that the 

 granite at the time of injection found the surrounding rocks already 

 at a high temperature. The junctions in Eastern Sutherland are 

 clearly exposed in a great many places ; and yet no sign of a chilled 

 margin has been detected anywhere. The same is the case in other 

 regions. (Indeed, " lit par lit " injection itself would appear, 

 a priori, to be possible only among hot rocks, as seams so thin would 

 soon consolidate among cold rocks, and so fail to make their way for 

 any distance.) 



If, now, we suppose a granitic magma introduced among rocks with 

 a pre-existing temperature scarcely lower than its own (it might be 

 even higher if the rocks were less fusible), not only would much 

 more intimate intrusion be possible, but even when actual intrusion 

 ceased the granite sills and the adjacent rocks would retain a high 

 temperature at the junction for a very long time. The conditions, 

 in fact, would be related to those of ordinary intrusion somewhat as 

 those of a column of liquid lead poured into a hot cylinder on to hot 

 gold would be to those of lead poured into a cold cylinder on to cold 

 gold. Solidification would be long delayed, and all this time the 

 magma might reasonably be expected to diffuse into the surrounding 

 rocks, following their natural divisional planes, and giving rise to 

 all the phenomena of a permeation zone. 



The singular fact that the inclusions of gneiss in granite, even down 

 to the thinnest films, are so very seldom, in these zones, disturbed in 



