J. G. Goodckild—On Drift. 497 



theory of the origin of drift deposits that seems to throw some light 

 upon this and some other imperfectly understood points in Pleisto- 

 cene Geology. 



Much of the difficulty which has hitherto been felt in accounting 

 for the transportal of marine shells to high levels by means of ice in 

 the way suggested by Messrs. Croll, Tiddeman and Belt, has arisen 

 through what is probably a misconception respecting the method of 

 formation of the beds in which the fossils occur. If it can be proved 

 that the greater part of the till has been accumulated, in the first 

 instance, between the ice and the rocky bed over which it was 

 passing, then it would follow that the force resulting from the 

 advance of the Irish Sea ice-sheet must have been sufficient, not 

 only to crush and completely destroy all traces of any organisms 

 which may have found their way beneath it, but also to knead up 

 the beds in which they occurred in such a way as completely to 

 obliterate all traces of stratification. To deny this, as Mr. Bonney 

 has pointed out, is to deny to ice that efficacy as an abrading agent 

 which those accustomed to the practical investigation of glacial 

 effects see every day additional reason to ascribe to it. 



Eeferring to descriptions of the drifts of the Basin of the Irish Sea, 

 we find that most writers concur in describing those of the low 

 ground as divisible into three groups — a lower till group, with traces 

 of stratification throughout, and occasional seams of sand and gravel ; 

 — a sand and gravel series, which includes occasional thin beds of 

 till, and deposits of loam and laminated clays ; — and an upper till, 

 resembling that beneath in every essential respect, except that it is 

 occasionally noted that it contains a larger per-centage of far-derived 

 boulders than does the older deposit. Between the lower till and 

 the sand and gravel series, and also occasionally between this and 

 the upper till, the junction is uneven, from the denudation effected 

 in the interim between tiie formation of the overlying and the under- 

 lying beds. 



Eeferences to any contortion of the beds on the large scale are 

 rarely to be met with, and the few instances that have been described 

 show that the amount of derangement is wholly incommensurate 

 with that which must have resulted from the long- continued onward 

 movement, over beds which are yet unconsolidated, of a mass of ice 

 that was certainly not less than 2000 feet in thickness. 



Following the drifts inland towards the heads of the principal 

 river-basins, the medial sand and gravel group is locally found to 

 pass horizontally, through frequent alternations of sand and gravel 

 and till, into a deposit which cannot be distinguished from ordinary 

 till, so that the three subdivisions of the drift are no longer traceable. 



It is usually the case that the proportion of washed and water- 

 worn material included in the whole of the drift of any given spot is 

 in exact proportion to the area of the catchment basin lying above it. 

 In other words, the quantity of water-worn drift, and the volume 

 of water flowing along the principal river, at any given part of a 

 district, almost always bear an exact proportion to each other ; so 

 that, near the sources, as a rule little else than till is to be met with ; 



DECADE II. — VOL. I. NO. XI. 32 



