120 Pirsson—Complementary Locks and Radial Dikes. 
takes place by the diffusion of the oxides of lime, iron and 
magnesia towards the outer cooling margins of the enclosed 
mass. Thus border zones of more basic material are formed 
and accordingly as one portion or the other is erupted at vary- 
ing stages of the process, rocks of different types will be made. 
This assumption has been reached by indirect proof, but recently 
a positive demonstration of it has been shown to occur at 
Square Butte in the Highwood Mts. of Montana.* At this 
locality the mountain has an inner, acid, feldspathic core sur- 
rounded by a broad marginal zone of basic rock rich in ferro- 
magnesian silicates, and a variety of facts and considerations 
prove that it was not formed by two separate intrusions but by 
the process of differentiation im an originally homogeneous 
molten magma. 
While examples are known which seem to be the reverse of 
the process by which the outer margins of the differentiating 
mass become more basic, it may yet well be called the normal 
one. What laws condition it are not yet well known. 
If we now imagine an intrusion of igneous magma to take 
place into the superficial crust of the earth from depths below, 
it can be seen that while the magma is cooling and crystallizing 
and the heated rocks still further cooling and cracking from 
contraction, that during this period processes of differentiation 
can go on in the still liquid material below. If further ejec- 
tions of this differentiated material take place into the cracks 
produced by contraction, dikes, sheets and laccolites of comple- 
mentary rocks will be formed. 
If the differentiation has followed the normal course, it would 
then be reasonable to expect that the acid forms (the oxy- 
phyres) would be most commonly found cutting the central 
stock or in its immediate neighborhood, while the basic forms (the 
lamprophyres) would chiefly occur in the outer zone of sedimen- 
_taries as dikes and intrusive sheets. Every geologist will 
recall how common it is to find granite areas cut by dikes of 
aplite, how much rarer these forms are at a distance from them, 
while with basic dikes the reverse holds true. 
Moreover, another factor may contribute to this result. 
While the liquid mass below is cooling it is also becoming more 
viscous and this viscidity would naturally be more pronounced 
in the case of the acid feldspathic portion than with the more 
easily fusible basic one. Hence we should expect to find the 
oxyphyres in greatest amount where the rocks are the most 
intensely heated, that is at the center, and that the lampro- 
phyres, being more liquid at a lower temperature, would be also 
more likely to be erupted and would therefore occur in greater 
* Weed and Pirsson, op. cit. 
