EVOLUTION OF IRONDEQUOIT VALLEY. 155 



lowering master lake, was checked at this point by encountering 

 an obstruction of some more resistant nature than the sih, perhaps 

 firm till or moraine, perhaps hard rock even. Such an obstruction 

 is indeed demanded by the subsequent events in this vicinity, there- 

 fore we would not lay too much stress upon the possible relation 

 subsisting between the Vermont level and that of this resubmerged 

 plain, but would look upon the latter as marking the natural end 

 of Lake Irondequoit, upsilted in its shallow senescence and converted 

 into a swampy meadow, persisting henceforth as such until drowned 

 long afterward by the returning flood of Lake Ontario. This 

 meadow, then, composed of the materials swept out from the up- 

 clogged area on the south by the rejuvenated Irondequoit River, 

 established automatically the permanent minimum base-level of the 

 latter's activities, a base to which the river speedily adjusted itself 

 from Penfield northward and by its subsequent meanderings evolved 

 the characteristic scenery of the "dugway" section. Meantime, 

 however, the master lake without, now naught but an arnr of the 

 ocean, was still "falling" as the land was lifted out of it, and thus 

 the channel continued to be deepened north of the meadows, through 

 what is now the deepest, narrowest part of the Bay. 



Our map E (Plate VIII) shows all these details at a tin-ie when 

 the waters of the Ontario basin had withdrawn beyond the northern 

 limits of our map area. The depth of 137 feet (23 fathoms) cor- 

 responding to our local estimate of 110 feet above present sea level 

 for the minimum stage or transition from Gilbert Gulf to inchoative 

 Lake Ontario by emergence of the Thousand Islands above the sea, 

 is shown by the soundings three miles north of the present shore 

 line, but allowance must be made for subsequent fillings and the 

 actual shore was probably somewhat nearer than that. Allowing 

 four miles from the 14 fathom sounding in the Bay to this assumed 

 outlet at 23 fathoms, and with no correction for land tilting since, 

 a gradient of 13.5 feet per mile is indicated for the bottom of the 

 channel then cut. This is not heavy for a stream in drift, as may 

 be seen by comparison with the Irondequoit above Penfield (on the 

 map) where the present gradient, even following the meanders, is 

 over 20 feet per mile. On the other hand, the jump from 84 feet 

 up to 45 in one mile, from this deepest sounding to the submerged 



