BY 11. I. JENSEN. 913 



basaltic, uneven-grained rock, containing olivine and titaniferous 

 augite. Truly both these rocks are alkaline, but their facies, 

 mode of occurrence, constitution, and origin are all so different 

 that the system of classification which brings them together is 

 most unnatural. The one type, N.15, is a hypabyssal rock 

 derived by magmatic differentiation from an alkaline magma; 

 the other is a volcanic rock which originated by magmatic mixing. 



A system of rock-classification cannot be both chemical and 

 mineralogicftl, and an attempt to create such a system must be 

 futile. To separate the lime combined with alumina from that 

 of diopside, calcite, and apatite in estimating the rang is also 

 unnatural. 



Personally, I favour adherence to the old nomenclature, but 

 believe as well in having as many analyses as time will permit. 

 To determine the norm is also highly desirable. A chemical 

 analysis is almost as quickly made as Rosiwal measurements, and 

 is much more reliable than calculating the norm from the mode. 



Further, Rosiwal measurements are sheer waste of time unless 

 the rock to be studied is medium,- or coarse-grained, and of fairly 

 even grain-size; in addition, its minerals should be of definite 

 and known composition (as in granite and gabbro). 



Volcanic Sequence. — As already stated the sequence observed 

 in the Nandewar Mountains was — 



(1) Injection of basic laccolites and sills as N.17; after which 

 prolonged denudation. 



(2) Eruption of arfvedsonite trachytes (N.30, N.59, N.55, etc.) 

 and their tuffs. 



(3) Earth-movements and intrusion of sills of pulaskite and 

 nordmarkite porphyry, accompanied by eruptions of phonolite 

 (N.ll, N.IO, N.49). 



(4) Eruptions of andesite followed, and the sills of labradorite 

 porphyry being of andesitic composition probably belong to this 

 phase of activity (N.12). 



(5) Flows of alkaline basalt, followed by normal basalt. 



