ROCK CLASSIFICATION 107 



duration, embracing a whole geologic period, as, for example, 

 the Tertiary. It may include a number of lesser series of erup- 

 tions localized within the province, as at centers of eruptions such 

 as volcanoes ; and may be highly complex, having many branch- 

 ings. In some series the range of chemical variations is com- 

 paratively small, in others it is comparatively great. Examples 

 of these appear in the rocks of the Yellowstone Park and in 

 those of the Christiania region, whose analyses have been plotted 

 in Diagram 6 ; the red spots representing those of the Chris- 

 tiania region. 



If more limited genetic series are compared it is found that 

 in one case the chemical variation is chiefly in the line of silica, 

 from much to little, accompanied by abundance of feldspar 

 molecules for the higher silica, and abundance of ferromagnes- 

 ian molecules for lower silica. While in another case the chem- 

 ical variation affects the silica but slightly, and shows itself in 

 the relative abundance of alkalis and alumina on the one hand, 

 and of ferromagnesian molecules on the other. 1 The definition 

 of a genetic group or family of rocks, as expressed chemically, 

 must, therefore, be very flexible and indefinite. 



The natural consequence of the variability of composition 

 among rocks of one genetic series, and of the existence of 

 genetic series closely similar to one another chemically, is the 

 close resemblance of some rocks of one series to certain rocks 

 of other series. And since the differences in most cases consist 

 in the relative proportions of the same chemical elements, it 

 follows that some rocks of one genetic series are quite as much 

 like certain rocks of another series as these are like other 

 rocks of the same series. Hence, a chemical definition broad 

 enough to cover several closely similar rocks of one genetic 

 series may apply equally well to similar members of another 

 genetic series, and it cannot be framed so as to exclude them. 

 In other words, it follows from the very nature of a chemico- 

 physical differentiation of rock magmas that some rocks belong- 



^ddings, J. P., Absarokite-Shoshonite-Banakite Series, Jour. Geol., Vol. 

 Ill, Chicago, 1895, P- 935- 



