388 M. L. FULLER — GEOLOGY OF FISHERS ISLAND 



type locality of Pleistocene Coastal Plain formations in the northeastern 

 United States. If the interpretations outlined in this paper are correct, 

 four glacial and three interglacial stages have been represented in the 

 upbuilding of both islands. 



The question of whether the succession of ice invasions in this region 

 was the same as in the central portion of the country and in Canada 

 is difficult to determine. The lobate character of the margins of the 

 earlier ice-sheets in the Mississippi valley is such as to suggest that 

 they might be, to a certain extent, confined to that region. On the other 

 hand, the Wisconsin invasion is known to have reached from the Rocky 

 mountains to the Atlantic coast, reaching nearly as far south in this 

 region as at points farther west. 



The most important evidences pointing to the similarity of succession 

 in the Atlantic and Mississippi regions are the relative weathering of the 

 drift, the erosion features of the interglacial stages, and the evidence of 

 elevation or depression of the land in the two regions. For instance, the 

 old extra morainic drift of Pennsylvania (the oldest drift on the main- 

 land east of the Mississippi valley) was deposited when the land was low 

 and the streams relatively sluggish, and on Long island the earliest inva- 

 sion was associated with a marked depression, the land standing 300 feet 

 below its present level. In each case the period of low level was suc- 

 ceeded by a pronounced uplift, which on Long island resulted in the 

 nearly complete removal of the Mannetto gravels and in Penns}dvania 

 in the very considerable deepening of the Allegheny, Monongahela, and 

 other river valleys. This strongly suggests that the sequence of these 

 early events was not only the same in each case, but that they were 

 contemporaneous. 



As to the absolute age of the Mannetto gravel of Long island and the 

 older drift of Pennsylvania, there is little direct evidence except the state 

 of weathering of the pebbles, which is materially greater than in the 

 Kansan deposits of the Mississippi valley: so much so, in fact, that it 

 seems as if an interval of time, perhaps as great as all subsequent Pleis- 

 tocene time, must have intervened between the deposition of the Man- 

 netto and the Jameco. It is believed that these two drifts, together with 

 the intervening interglacial stage, can be correlated with a fair degree of 

 certainty with the Albertan, Aftonian, and Kansan stages of Canada or 

 the Mississippi valley. 



The Jameco gravel is followed conformably by the Gardiner clay and 

 the Jacob sand, both of which are interglacial deposits, and would nat- 

 urally be referred to the Yarmouth interglacial stage. The Jacob sand 

 is normally followed, in turn, on Fishers island by the Herod gravel and 

 Montauk drift, and on Long island by still other gravels. Between each 



