132 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. 



times this seems to have been effected by direct vertical transmission. 

 At other times many species appear to have been destroyed in some 

 particular part of the area at the period of these changes, or rather, the 

 change brought about unsuitable conditions whereby they were tem- 

 porarily displaced. Yet nevertheless, when during a recurrence of 

 somewhat similar conditions^, subsequent immigrations introduced at 

 a later period a comparatively distinct newer fauna, many of these 

 old forms which had been preserved in the less disturbed districts, 

 and had there remained, as it were, in exile during the whole, or 

 during one or more divisions, of an intervening period, returned. 



At other times some species seem to have travelled more rapidly 

 or more readily in a horizontal than in a vertical direction ; and thus 

 many forms, which are common in some particular zone in one area 

 and absent in the adjacent ones, seem, in process of time, to have 

 migrated into these latter, and flourished there in the next subse- 

 quent geological period ; whereas in the area in which they first ap- 

 peared they have not been transmitted upwards, and are not found 

 in these newer beds. Thus, though extinct in one area, they may 

 continue to mark a higher series in an adjacent area. 



It is also a point of very considerable interest to note how constant 

 and steady is the progression of change in the several subdivisions of 

 each formation. The disappearance of some species and the appearance 

 of new species seem continued quite apart from those great breaks 

 which from time to time give abruptness and prominence to a law 

 equally operative throughout all time. 



How far the Upper Bagshot Sands are related to the Bracklesham 

 series it is difficult to say. The few fossils I have found in those 

 sands are not sufficiently distinctive to enable me to pronounce a de- 

 cided opinion. As, however, the fossiliferous Middle Bagshot Sands 

 are very thin, and represent apparently only the lower or middle part 

 of the Bracklesham series, I think it probable that it is the upper 

 beds of sand and clay of the latter which pass northward into the 

 thick sands of the Upper Bagshot Sands. Still it is possible that 

 part of them may represent the Barton series, for we see at Barton 

 how shifting the upper part of that series is — how clay predominates 

 at one place and sands at another *. If we had to limit the Calcaire 

 grossier series at Cassel to the beds beneath the Nmnmulites vario- 

 larius zone, it would render the latter view the more probable. I 

 now, however, there give that series greater dimensions than I did in 

 my last paper (see PI. VIII. vol. xi. p. 241). 



We are now at the conclusion of an important section of the 

 Eocene period. The changes we have traced thus far have been 

 essentially under marine conditions. We already begin to perceive 



* The occurrence in the middle beds of the Sables moyens of blocks of sac- 

 charoid sandstones traversed with rootlet-shaped casts and impressions has its 

 counterpart in the blocks of sandstone lying on the top of the Upper Bagshot 

 Sands in the Bagshot district. 



