420 WEED AND PIRSSON — HIGHWOOD MOUNTAINS OF MONTANA. 



does. It is evident, however, that differentiation did not take place at 

 Square butte as a process of crystallization, but in a liquid magma before 

 any crystallization occurred. This is rendered quite evident, since none 

 of the ferro-magnesian minerals of the shonkinite are found in the 

 syenite. The only one, indeed, which is found in the syenite is the 

 barkevikite-like hornblende, while in the shonkinite are found iron ore, 

 biotite, olivine and pyroxene. Thus Square butte affords a striking con- 

 firmation of the ideas recently expressed by Brogger in his remarkable 

 work on the basic rocks of Gran.* 



It is a matter of some interest here to compare the composition of the 

 augite of the shonkinite, b}^ far its most prominent constituent, and the 

 hornblende of the syenite from Melville's analysis. 



Barkevikite. Augite. 



SiO., 38.41 49.42 



TiO, ] .26 .55 



AlA 16.39 4.28 



FeA 3.75 2.86 



FeO 21.75 5.56 



MnO 15 .10 



MgO 2.54 13.58 ' 



CaO 10.52 22.35 



. Na.,© 2.95 1.04 



K2O 1.95 .38 



H2O 24 .09 



99.91 100.21 



The result of the increase of magnesia and lime shows itself in the 

 change in composition of the dark mineral. The iron shows a move- 

 ment in the opposite direction ; in the syenite it is all found in the horn- 

 blende ; in the shonkinite large quantities had been used for the iron ore 

 and olivine, and to some extent for the biotite before the augite began 

 crystallizing ; hence it is not so prominent as in the barkevikite. 



In general, however, the difference is of like kind with that shown by 

 the mass analyses of the rocks and shows clearly how the composition 

 of the prominent dark mineral is a function of the magma in which it 

 is formed. That minerals indeed are so often conditioned by the magma 

 in which they are formed is without doubt the fact that has given to 

 some the idea that definite mineral molecules individualized as such can 

 exist in the molten magma. 



Recently Johnston-Lavis f has formulated a theor}^ for the different 

 composition of igneous rocks occurring at the same eruptive center by 

 supposing that the body of molten magma which gave them birth was 



* Quart. Jour. Geol. Soc, vol. 1, 1894, p. 15. 

 t Natural Science, vol. iv, February, 1894. 



