164 Alfred Bell — Molluscan Deposits of Wexford, etc. 



biotite. Here we have a case where the volcanic representative of 

 a saturated plutonic rock is a highly unsaturated rock. 



A continuous sequence, with but little variation in composition, is 

 often observed, from pyroxenite through wehrlite to dunite; neverthe- 

 less, olivine-free pyroxenite would have to be sharply discriminated 

 from both wehrlite and dunite. The same remark applies to the 

 allivalite-dunite series. Such examples, where the local, and 

 sometimes apparently accidental, occurrence of an unsaturated 

 mineral might widely separate closely related rocks, could be 

 multiplied indefinitely. Enough has been said to show that the 

 Principle of Saturation, applied in such an empirical fashion, is not 

 sufficiently elastic, and also that it tends to obscure petrological 

 affinities as well as the relative petrological values of minerals. 



With regard to other classifications, there are several alternatives to 

 Professor Shand's other than the silica-percentage one and WincheLTs. 

 Thus Tyrrell 1 has recently indicated the possibilities of a quantitative 

 modal system. Many of the difficulties of such a system would 

 disappear were more analyses, of the groundmass of aphanites 

 available, while in addition the ' typical ' composition of constituent 

 minerals could also be utilized. This author has also pointed out 

 that one reason of Iddings' failure to correlate chemical with 

 mineralogical composition is the fact that very few modes have been 

 determined. At the present day any biological classification which 

 was not primarily based on the phylogeny and ontogeny of species 

 would not be considered. In view of the recent great developments 

 in petrology, there seems no good reason why these two factors should 

 not also be fully utilized in rock-classification. As Harker has said, 

 " There are those who would have us abandon in despair all endeavour 

 to place petrography on a genetic basis and fall back on a rigid 

 arbitrary system as a final solution of the difficulty. This would be 

 to renounce for ever the claim of this branch of geology to rank 

 as a rational science." 2 



VII. — The Fossilifebous Molluscan Deposits of Wexfokd and 

 ISToeth Manxland. 



By Alfred Bell. 



THE notes by Professor Grenville A. J. Cole & Mr. T. Hallissy 

 on the Wexford Gravels, 3 and those by Dr. F. Cowper Reed on 

 the Manxland shells in the Strickland Collection at the Sedgwick 

 Museum, Cambridge, 4 published in last year's Geological Magazine, 

 will be welcomed by geologists as drawing attention to an interesting 

 but too long neglected subject. The principal object of this paper is 

 to point out the faunal relations of the deposits in question to each 

 other, and wherein they differ from those of the neighbourhood 

 which are usually grouped with them, including the high-level shelly 

 drifts of Moel Tryfaen, Macclesfield, Gloppa, and the Wicklow 



1 Science Progress, ix, No. 33, pp. 60-84, 1914. 



2 Bep. Brit. Assoc. Portsmouth, 1911, pp. 380-1. 



3 Geol. Mag., pp. 498-509, 1914. 



4 Op. cit., pp. 544-5, 1914. 



