162 Proceedings of Boyal Society of Edinhurgh. [fee. 20 , 
To sum up these results, we may say that all the rocks in question 
belong to a single type, which we may call olivine-didbases. 
The material at my disposal, having been selected in the field, 
enabled me to make a special study of the endogenous contact- 
phenomena. The rocks are comparatively fresh ; and they represent 
specimens from the peripheral portion (Salband) as well as from 
the central and from intermediate parts of the volcanic body. Thus 
I was able to point out that the volcanic mass has been modified, 
both in regard to substance and to form, and not only throughout 
the whole mass, but also in its different parts, and that this modifi- 
cation is to be ascribed to the influence of the associated rocks. 
I. Modification in Substance. 
We are often astonished at the great quantity of fragments 
enclosed in volcanic rocks, and it has been stated in several treatises 
that fragments of the associated rocks have been melted down by 
the igneous magma. Nobody, however, has succeeded in proving 
that a given modification of the volcanic magma has been produced 
by partial or complete assimilation of fragments of other rocks. 
The investigation of this question could be most favourably under- 
taken in the intrusive masses of the Salisbury Crags, Hound 
Point, and Stewartfield. These bodies, up to 100 feet in thickness, 
present in their central portions a rock which has been formed by 
a magma that cooled down with extreme slowness. At the immediate 
junction with the associated rocks, however, the magma consoli- 
dated rapidly. In this way the volcanic mass avoided all the 
influences that would have arisen if the magma had melted down 
extraneous material. We should, therefore, expect the portion of 
the rock that consolidated most rapidly to represent that rock- 
type to which the magma originally belonged. All these phenomena 
are very clearly illustrated by the above-mentioned specimens. At 
the immediate junction the diabase exhibits a great quantity of 
well-formed crystals of olivine. At a small distance from the con- 
tact these crystals of olivine are more or less corroded. Towards 
the heart of the mass the olivine is either absent, or it occurs 
sparingly in isolated rounded grains. Olivine, which, owing to 
its basic composition, is one of the first minerals to crystallise 
