THE NATURAL HISTORY OF IGNEOUS ROCKS. 213 



mensions to an indefinite depth. Others regard an ordinary- 

 granitic intrusion as differing only in its greater size and less 

 regularity from other intrusive bodies which can be more 

 completely examined, and comparable in its general shape 

 and relations with a sill or laccolite, or in some cases perhaps 

 with a " plug ". The question is closely connected with an- 

 other one, viz., to what extent a rock-magma under plutonic 

 conditions "assimilates" or incorporates by fusion portions 

 of the neighbouring solid rocks. One extreme view is that 

 intrusive rocks in general have replaced rather than dis- 

 placed the rock-masses which they have invaded. Pushed 

 to its limit, this hypothesis leaves little or no room for any 

 invading magma at all, and therefore implies that the 

 igneous rock has been derived almost wholly from the 

 fusion in situ of pre-existing solid rocks. Such a supposi- 

 tion seems to be scarcely tenable in the case of an igneous 

 rock-mass occupying a definite space among surrounding- 

 rocks and encircled at most by a limited aureole of meta- 

 morphism. It is difficult to believe that a relatively small 

 amount of intruded magma could carry sufficient heat to 

 melt a large bulk of rocks, and equally difficult to account 

 for the narrow localisation of heat if supposed due to other 

 causes. That a molten rock-magma may to some extent 

 and under suitable conditions melt or dissolve its encasing 

 walls and fragments detached from them is well known, 

 and observations which may be referred to on another 

 occasion seem to indicate that in certain cases this action 

 is of considerable importance, but its application is probably 

 very limited. There is no doubt that a channel once formed 

 for an igneous intrusion may^be enlarged in this and other 

 ways, the magma penetrating fissures and detaching frag- 

 ments, which are partly dissolved, partly carried forward or 

 upward. Perhaps something of this kind has occurred, for 

 example, in the Ross of Mull granite as described by Good- 

 child (20), but such an action cannot be described as an 

 assimilation of the surrounding rocks by the granite 

 magma. 



The question may obviously be approached from the 

 chemical side, and here the evidence is decidedly against 



