IGNEOUS ROCK-SERIES AND MIXED ROCKS 397 



actual instances, we have only to take a collection of trustworthy 

 rock-analyses, such as that published by Clarke and Hillebrand, 1 

 and plot diagrams upon a convenient scale. The lavas of the 

 Lassen Peak region in California afford a good example. Here 

 the curves of some of the bases approximate to straight lines 

 v throughout a considerable part of their extent. 



Towards the basic end, however, the alumina- 

 line becomes strongly convex and those for 

 magnesia and lime decidedly concave, while 

 the soda-line is slightly but distinctly convex 

 throughout. This diagram includes the nor- 

 mal basalts, andesites, dacites, and rhyolites ; 

 and the very slight amount of smoothing 

 required to obtain flowing 

 Y' curves stamps this group 



7*^ of rocks as a natural series. 



j£ The quartz-basalts, on the 



other hand, refuse to 



^ ^*~ ~x adapt themselves to this 



FlG> 4 scheme, and their abnor- 



mal composition is clearly 

 brought out by plotting their analyses on the same diagram. 

 In their content of lime and potash they do not differ notably 

 from normal rocks of like silica-percentage, but they show a 

 marked deficiency in alumina and ferric oxide, and to a less 

 degree in soda, and an excess of magnesia and ferrous oxide. 



Although natural series of igneous rocks differ considerably 

 from one another, they nevertheless possess certain broad char- 

 acteristics in common. This is recognized by the very general 

 practice of speaking roughly of acid, intermediate, and basic 

 rocks, etc., as having more or less distinctive characters ; which 

 tacitly assumes that, in the broadest view, they fall approxi- 

 mately into a single line. Given a large number of analyses of 

 normal (unmixed) igneous rocks, we might average the com- 

 position of those having like silica-percentages, and construct 

 'Analyses of Rocks, Bull. No. 148, U. S. Geol. Surv. 1897. 



