A REVIEW OF IGNEOUS ROCK CLASSIFICATION 79 



Division VI., consisting of rocks with dominant mafic con- 

 stituents, is really a survival of the ultrabasic class of the 

 qualitative classifications. The ultrabasic rocks, however, occur 

 in very small quantity as compared with those of most of the 

 other divisions. They rarely form independent masses, large 

 or small, and are usually encountered as differentiation-facies 

 of rocks belonging to the other divisions, especially III., IV., 

 and V. Moreover, the erection of a division based on mafic 

 minerals contravenes the principle that the main factor in 

 igneous rock classification should be based on the predominant 

 felsic minerals. The proportions of the mafic constituents 

 are sufficiently recognised by the erection of groups based 

 on the felsic mafic ratio (see subdivision of the syenites, 

 ante p. 78). It is in accordance with the results obtained 

 in the foregoing discussion that the rocks of Iddings' Division 

 VI. should be distributed amongst the other five divisions, 

 where they would at once find their place as the domafic 

 and permafic members of these divisions. 1 



A second criticism is that the second factor in the majorit}^ 

 of the classes is the ratio of alkalic to lime-soda felspars. 

 The latter may range from oligoclase to anorthite. Albite, 

 which from a mineralogical point of view, forms part of the 

 plagioclase series, is, however, and rightly, treated as an 

 alkali felspar. But as the lime-soda felspars contain the albite 

 molecule in varying proportions, the latter appears in both 

 the quantities that form the ratio. This is neither necessary 

 nor desirable. What is required is to contrast all the alkali- 

 felspar with all the lime-felspar ; and to that end the albite 

 molecule included within the lime-soda felspar should be 

 counted with orthoclase and pure albite in the numerator of 

 the fraction. As pure albite rarely occurs independently in 

 igneous rocks, and only to a small extent in solid solution 

 or isomorphously mixed with orthoclase (as soda-orthoclase, 

 anorthoclase, etc.), the ratio used by Iddings essentially con- 

 trasts the orthoclase molecule on the one hand with the 

 albite plus the anorthite molecules on the other. This, 

 however, is a meaningless ratio as compared with that of 

 orthoclase plus albite to anorthite. The latter expresses 



1 Holomafic rocks could be treated as suggested for the holofemic rocks ; 

 from their affinities and paragenesis it should be possible in the great 

 majority of cases to determine in which Divisions they should be placed. 



