GENERAL PETROLOGY OF SQUARE BUTTE. 419 



very regular and syni metric arrangements of the parts, tlie absence of 

 all inclusions or " sclilieren,'' the cleanness of the zonal edge, together 

 with the common properties already })ointed out, uttei'ly preclude this 

 idea. There are, indeed, places in the Highwoods where intruded masses 

 show further movements after differentiation has taken i)lace, with the 

 result of remarkably banded and streaked rocks, whose very occurrence 

 shows that such was not the case at Square butte. 



We are, indeed, forced to conclude at every step that the mass was 

 originally homogeneous, and that differentiation took place by the dif- 

 fusion of the bivalent oxides toward the outer surfaces. 



It would add greatly to the value of the results here presented if we 

 could know or could obtain the composition of the original magma in 

 which the differentiation took place. This, however, cannot be done by 

 comparing the masses of the two rocks, because, although it is probable 

 that the amount of syenite now present represents pretty nearly the 

 original one — that is, that there has been only a small erosion of that 

 rock — the case is quite different with the shonkinite, a very large part of 

 which has been carried away ; hence, not knowing the relation of the 

 two masses involved, we cannot estimate the composition of the original 

 magma. It is evident, however, that it must have been between the 

 syenite and shonkinite. 



Shonkinite, however, occurs in large bodies in the neighborhood of 

 Square butte and elsewhere throughout the Highwood range, while rocks 

 closely related to it in chemical and mineral composition are found in 

 the form of dikes, extruded lavas and breccias. Throughout the district 

 what may l)e called acid or highly feldspathic rocks i)lay but a sul)ordi- 

 nate role. In view of these facts, we are inclined to believe that the com- 

 position of the original magma approximated more closely to shonkinite 

 tnan to the syenite. 



It will be seen, therefore, that Square butte presents in a demonstrative 

 way the same idea that Brogger inferentially deduced and presented as 

 the explanation of the processes of differentiation by which the varied 

 rocks of the region of south Norway have been formed.* 



Recently Harkerf has described an interesting occurrence of a gabbro 

 massif, which grows steadily more basic or richer in the ferro-magnesian 

 minerals as the outer boundary is ap[)roached. Harker explains this 

 occurrence by pointing out that the order of concentration of the min- 

 erals is the same as the order of their crystallization, and hence accounts 

 for the differentiation as a process of crystallization. S(piare l)Utte is also 

 more l)asic as we approach the outer boundary, but the transition occurs 

 abruptly, so to speak, or within such a narrow zone that it practically 



•Zeit. fQr Kryst., vol. zvi, 1890, p. 85. 

 tQimrt. Jour. Geol, Soc, vol. I, 1894, p. 311. 



