118 GEOLOGY OF THE YELLOWSTONE NATIONAL PARK. 



In Table X the percentages are all placed in the extreme right-hand 

 column, and also in separate columns corresponding to the groups described 

 in the first part of the paper. From this it is seen that the main body of 

 the diorite varies from 53.72 to 60.56 per cent of silica, and in certain con- 

 tact forms reaches 67.54 per cent. The dikes of later rocks related to the 

 diorite and cutting the main body of the stock range from 63.78 to 69.24 

 per cent. 



Of the various bodies of magma that have followed one another 

 through the conduit at Electric Peak, there is a variation in chemical com- 

 position in each, the different series of changes overlapping one another. 

 Thus the average chemical composition of each subgroup of varieties shifts 

 somewhat, and is more basic for one than for another. But the end varieties 

 of each subgroup overlap, so that the most basic modification of the more 

 acid group is more basic than the most acid end of the more basic group 

 which immediately preceded it. 



Since the rocks of Group I belong to outlying dikes of the main stock 

 and are contemporaneous with it, their silica percentages may be placed in 

 the proper subgroup of the stock rocks, making- Subgroups II a and II & 

 practically continuous. It appears from Table X that the succession of 

 magmas which came up through the vertical fissures was from a basic one 

 to more and more acid ones, and that the previous intrusions which formed 

 the sheet rocks were of a magma of medium chemical composition. 



The variations of the other chemical constituents of these rocks are 

 best comprehended by comparing their molecular proportions. This has 

 been done graphically in the accompanying diagram, fig. 2, in which the 

 molecular proportions of the principal oxides are plotted as ordinates, those 

 of the silica being taken as abscissas. The origin of abscissas is located 

 some distance to the left. 



The first impression derived from the diagram is that of the irregularity 

 of the variations in all the oxides besides silica, especially in the magnesia. 

 Moreover, these variations appear to be independent of one another. But 

 this apparent independence disappears on closer study. The most striking 

 evidence of connection between the molecular proportions exists in the case 

 of the two oxides of iron; the ferrous and ferric oxides are noticeably 

 inversely proportional to each other, an increase of ferrous oxide being 

 accompanied by a decrease of ferric oxide. The total amount of iron 



