334 The American Naturalist. [April, 



MINERALOGY AND PETROGRAPHY. 1 



Petrographical News. — Still another attempt to arrive at a just 

 view concerning the chemical relations of eruptive rocks has been 

 made, this time by Lang. 2 Calcium and the alkalies are regarded as 

 the best indicators as to the relationships of the rock masses, and in 

 this respect the new investigation departs widely from older ones, in 

 which the silica was alwa\ aps the most character- 



istic of a rock's chemical constituents. After citing a large number of 

 analyses of rocks chosen from carefully examined types of all classes, 

 the author divide rock magma> into four great groups, viz : Those in 

 which the proportion of K,0 present exceeds that of CaO and Na 2 

 combined, or K 2 > CaO + Na 2 0, and those in which the propor- 

 tions of the components correspond to the following formulae : Na 2 0>, 

 CaO + K 2 0, Na 2 + K 2 > CaO and CaO > K 2 + Na 2 0. Each 

 of these groups is then subdivided into types. In the first group, for 

 instance are two orders in one of which Na 2 > CaO, and in the other 

 CaO> Na 2 0. In the first order fall the Cornwall granites with 

 CaO : Na 2 : K 2 = 1 : 4 : 14 ; the Heidelberg porphyry with an 

 alkali ratio of 1 : 1.5 : 8 ; the dyke granite type with a' ratio of 

 1 : 3.7 : 6, the granite-rhyolite type with a ratio of 1 : 2 : 4, 

 and the orthophyre type with 1 : 3.8 : 7.3. The second order 

 includes the Hesse granite, syenite and bolsenite, with the respective 

 alkali ratios CaO : NaO : K 2 ==2:1:6, 2.5 : 1 : 4 and 1.9 : 1 : 4.8. 

 The other groups are likewise subdivided into orders, and in each of 

 these are ranged the types. Brief notes accompany the descriptions 

 of each type, and a table giving the percentages of the principal con- 

 stituents of 247 fresh rocks closes the paper. Some of the relation- 

 ships brought to light by the author's discussion are so unexpected 

 that it may safely be affirmed that the views put forth in his article 

 will meet with much opposition among petrographers. The granites, 

 for instance, are discovered to occur in different orders, under differ- 

 ent groups, the types being often further removed from each other 



than are normal granite and phonolite. The new rock iohte, 



described by Ramsay and BerghelP fills the place in Rosenbusch's 

 scheme that was h- ft. for the plutonic . univalent of the nepbelinites. 



tion of the Mo 



ihn-l. 



