SYSTEMATIC PETROGRAPHY 48 1 



compounds, they are not accidental mixtures; (3) one must 

 consider the relative amounts of all oxides of bases to each 

 other and to silica ; (4) as silica is the dominant constituent it 

 is proper to take it as the basis for the primary classification ; 



(5) the next factor to be applied must be the contents in the 

 three oxide groups, alkalies, alkaline earths, and the sesquioxides ; 



(6) various single oxides may be used for further subdivisions. 

 From this statement it might be supposed that a classification 

 created by the successive use of the chemical factors named was 

 to be set up by Loewinson-Lessing, and such an arrangement of 

 magmas would have claim to being a chemical classification. 

 But the author does not do that, as we shall see. 



The actual system proposed by Loewinson-Lessing is to use 

 the silica contents for the formation of four general groups : 

 (1) Acid rocks; (2) Neutral rocks ; (3) Basic rocks ; (4) Ultra- 

 basic rocks. 



The second division is obtained by taking a certain number 

 of analyses representing known rock families (established on 

 the unsatisfactory basis of mineral composition) and determin- 

 ing the mean of these analyses, which is then set up as the com- 

 position of a rock type, and its formula and coefficient of acidity 

 are ascertained. 



It is clear that the grist of this mill depends entirely upon 

 what is put into the hopper. It is not a chemical classification 

 but a chemical cliaracterization of mineralogical rock groups 

 arbitrarily selected by the author. It will, of course, be possible 

 to secure means corresponding to any formula desired as a type, 

 and the rocks thus having typical position could be adopted as 

 centerpoints of groups or families. For the ordinary range of 

 rocks these types would often coincide with recognized rocks, 

 assigned certain names in existing systems, and these names 

 might then be given by redefinition to the families thus indi- 

 cated. But what would be the purpose of such a scheme? It 

 could not express the existing relations between the mineral 

 composition of the rock and the chemical constitution of the 

 maerma. 



