472 SCIENCE PROGRESS. 



igneous activity have been derived by processes of dif- 

 ferentiation from one oreat subterranean masfma-basin. we 

 must admit that these processes have completely masked 

 any original differences that may have subsisted between 

 one parent magma and another. 



There is another point of view from which the chrono- 

 logical sequence of igneous rocks may conceivably become 

 an element in a natural classification, viz., with regard to 

 the order of eruption of different rock-types belonging to 

 one period of activity, and within a given region. One of 

 the first geologists to address himself to this question was 

 Richthofen (6), whose "Natural System of Volcanic Rocks," 

 worked out chiefly in the Western States of America and 

 in Hungary, has been given a wide application. The rocks 

 considered by the author were those of the Tertiary epoch, 

 as developed in what he has styled " massive " eruptions. 

 He divided them into five orders — propylites, andesites, 

 trachytes, rhyolites, and basalts — each subdivided into 

 families, and he laid down as a universal law that these 

 several orders have been erupted in regular sequence as 

 named. More extended knowledge has only partially con- 

 firmed this generalisation of Richthofen as regards the 

 succession of the Tertiary lavas ; nor do the rocks of other 

 periods — e.g., those of the Hercynian eruptions in Central 

 Europe — appear to conform to any such simple law. The 

 law that the extravasation of material of intermediate com- 

 position has normally preceded the appearance of either 

 acid or basic lavas seems to be borne out in a large number 

 of instances, and such a law connects itself simply with 

 speculations on the differentiation of rock-magmas. But 

 any classification built upon such considerations as this would 

 evidently be founded upon genetic principles, the factor of 

 relative age not being really an essential one. 



We pass on to consider the classifications based upon 

 descriptive characters, the foremost being, of course, mineral- 

 ogical constitution. This being a necessary element in the 

 description of rocks has naturally come to figure prominently 

 in the definition and classification of rock-types, and has 

 often been assigned the first place. The earlier writers 



