PLEISTOCENE GEOLOGY OF NASSAU CO. AND BOROUGH OF QUEENS 633 



ing them. If then, we reconstruct the cross-section of the forma- 

 tion, it would appear as a wide belt of gravels and sands declining 

 sontliward to the sea and rising with a cuestalike bluff from the 

 sound on the north. 



In ordinary nonglacial, coastwise beds, such a bluff would indicate 

 the retreat of the beds, extending originally north to an overlap on 

 the mainland, to their present northern limit. But in the case of 

 glacial deposits laid down about the border of an ice sheet, it is highly 

 probable that the beds never thinned out landward to an overlapping 

 series on the ancient gneiss of the region beyond the sound. Some- 

 what similar glacial gravels and sands of the last ice epoch, on Xan- 

 tucket, end abruptly on their northern or iceward margins in bluffs 

 overlooking lower ground once occupied by the basal portion of tlie 

 ice sheet against whose mural front they were laid down.^ Upham ^ 

 has expressed his belief in the origin of these gravels and sands in this 

 manner, differing only from the view here set forth in that he sup- 

 poses tlie beds to be essentially contemporaneous with the moraines 

 which rise above their level. 



It is evident that these Columbia beds, exposed in the bluffs and 

 rude terraces along the north coast of the island, may once have 

 extended much farther to the northward, but how much farther 

 into the area of the sound is not now definitely determinable. 

 Their occurrence on the Connecticut mainland has not as yet been 

 reported, and till that area is carefully studied with this problem in 

 mind, it can hardly be satisfactorily settled. The same indefinite 

 answer is elicited from a study of the equivalent beds on Block 

 island and Marthas Yineyard. In other words, the precise position 

 of the ice front and terminal moraine of this earlier ice advance is 

 unknown, though it could not have been many miles north of the 

 inner limit of these gravels and sands with their intercalated bed of 

 true till. 



Aside from the disturbances above noted, two classes of changes 

 have affected these beds since their deposition : 1) the discoloration 

 of the beds by local and secular chemical changes in the iron-bear- 



> Curtis, G. C. & Woodworth, J. B. Jour. geol. Chicago, 1899. 7 : 226-36. 

 '^ Upham, Warren. Glacial history of the New England islands. Am. geol. 

 1899. 24 : 79-89, with bibliography, p. 89-92. 



