60 BOOKS FOEMED THEOUGH IGNEOUS AGENCIES 



The researches of the past few years have made it evident 

 that eruptive rocks are to be satisfactorily studied only when 

 considered in their geographical as well as geological relation- 

 ships; that is to say, the eruptives of any particular region 

 must be considered with reference to their genetic relation to 

 others of the same region; such a relationship as is suggested 

 by regarding them all as but varying phases of a process of 

 differentiation from a common magma. 



That such a relationship in many cases exists has apparently 

 been conclusively demonstrated by the work of Iddmgs^ in the 

 Yellowstone Park, J. F. Williams^ in Arkansas, Pirsson^ in 

 Montana, and Brdgger* in Norway, and many more recent 

 workers. The attempt at correlation of local types with those 

 of a somewhat similar nature at a distance is interesting and in- 

 structive, as showing on the whole a remarkable unity in nature's 

 methods; but we must never lose sight of the fact that each 

 eruptive centre, throughout periods of activity interrupted it 

 may be by thousands of years, works out its own results accord- 

 ing to local conditions which may or may not harmonize with 

 those at distant points. It is possible to conceive that, could all 

 the rocks of any successive periods of eruption from a single 

 centre be once more relegated to a common magma, such might, 

 in its entirety, be an exact equivalent of others in remote portions 

 of the globe. The consolidated results from the cooling of ex- 

 truded portions of this magma may, however, show ever-varying 

 differences due to local conditions. In short, eruptive rocks must 

 be considered by geographic groups and with reference to 

 magmas. 



Attempts at a satisfactory classification on other grounds 

 must prove invariably futile and tend only to retard, rather 

 than to promote, the science. 



In the following pages the rocks are discussed in groups, 

 each group comprising all those having essentially the same 

 chemical composition, but differing (1) in degree of crystalliza- 

 tion, (2) in mode of occurrence, and (3) in geological age. In 

 all, there is, within certain limits, a considerable range in min- 



^ Bull. PHlos. Soc. of ■Washington, XII, 1892. 



^Ann. Eep. Geol. Survey of Arkansas, Yol. II, 1890. 



^ Bull. Geol. Soe. of America, Vol. YI, 1895. 



*Die Bruptivgesteine der Kristianiagebeite, Christiania, Norway, 1894. 



