26 NEW YORK STATE MUSEUM 



Most of the higher channels of this series are shallow scounvays 

 in the limestone or only benches with low south banks. The lime- 

 stone was too resistant to yield deeply to the lowering waters, but 

 the gorge at Split Rock quarry (Solvay Process Company) and the 

 channels east and north are down in the Salina shales and very 

 pronounced. 



The next chapter will make clear that we must postulate a 

 readvance of the ice sheet that largely buried these channels, and 

 subsequently with the final recession of the ice a reexcavation by 

 the last drainage. In consequence of this repetition of stream work 

 we can not discriminate the lower limits of the first drainage work 

 from the second; but it must have been sufficiently low to carry 

 the waters which cut the Victor-Phelps channels, which were under 

 500 feet. This seems to require that the ice should have uncovered 

 all the land surface as far north as Jordan. Apparently all the 

 east-leading channels depicted on the map might have been pri- 

 marily excavated by the first or sub-Vanuxem drainage. The map 

 shows the general features and relations of these channels and 

 little further description is necessary. 



The delta at Hartlot seems to have been built in glacial waters 

 by land drainage through Skaneateles creek; which also contrib- 

 uted to the Elbridge delta. 



The swampy, broad, indefinite tracts followed by the railroads 

 and Erie canal represent the lowest paths of the latest glacial drain- 

 age, probably the course also of the sub-Vanuxem drainage, and 

 possibly a heavy flow past the on-coming ice during its invasion, 

 or invasions, of the region. It has not seemed necessary to cover 

 the broad low spaces with the designation or convention represent- 

 ing channels, as it is impossible to closely determine the limits of 

 stream work or to separate it from lake action, 



Marcellus channels: Lake Warren escape 



These two great channels have been briefly described in a former 

 writing [title 25. p. 53-55]. In that writing they were attributed 

 to the flow of the falling Warren waters; and such is still believed 

 to be their origin, though the present conception places them later 

 in time than the channels, with higher altitude, on the north. 

 The description of the channels will be in order, after which we 

 may discuss their relations and history. 



The western one of the two canyons is called, in lack of some 

 geographic name, after the local appellation, the " Gulf." It 



