PHENOMENA OF FLINT CREEK VALLEY 39 



swamp, which may have been produced by the filling of a shallow mo- 

 rainal lake. At Potter the valle}^ bifurcates, the larger and more direct 

 branch leading nearly south and joining the Branchport arm of the 

 Keuka valle3^ The col between these valleys is one mile south of Friend 

 postoffice, with an altitude not closel}^ determined, but of about 1,000 

 feet. A scattering kame-moraine stretches for three miles north of the col- 

 The col is a swampy flat, but no good channel leads from it. 



The southwest arm of the valley, known as Italy hollow, is a long, 

 narrow, winding valley, which holds the headwaters of Flint creek. The 

 divide at the head of the valley is over 1,500 feet elevation and only 

 some two or three miles from Naples. 



There seems no escape from the conclusion that the Italy branch of 

 the valley must have held for a time a deep local glacial lake, with its 

 outlet into the Naples lake at the highest level of the latter. We will 

 call this the Italy lake. A well defined channel leads from the col, but 

 soon descends rapidly as a ravine and bisects the ancient delta, situated 

 southeast of Naples village. The latter was named in the former paper 

 the Tannery Glen delta. The broad summit level of this delta has an 

 aneroid altitude of about 1,370 feet, which correlates with the Naples 

 lake at the highest level. 



AVhen the ice-front receded so as to uncover the junction of the two 

 valle5^s at Potter, the Italy lake was lowered into some other water. 

 There are reasons for thinking that its successor was the extended Ham- 

 mondsport lake. The Flint Creek valley received for a time the over- 

 flow of the Naples-Middlesex lake, and the channel cut by the draining 

 stream, described above (see page 38), piled the debris derived from the 

 gorge excavation in a large and conspicuous delta close to Potter vil- 

 lage.* The summit j^lateau of the delta has an elevation of about 1,150 

 feet. This is 150 feet over the col at Friend and correlates with the 

 level of the higher waters in the Keuka valley (see next page). 



The only objection to supposing that the Hammondsport lake ex- 

 tended north into the Potter section of the Flint Creek valley is the fact 

 that the present outlet of Keuka lake, at Penn Yan, into Seneca lake, is 

 farther south than Potter. But it should be understood that the north- 

 ern end of the Keuka valley is contiguous to, and indeed blends with, 

 the larger Seneca valley, and that heavy lobes of the ice-sheet lingered 

 in the larger, deeper valleys of Seneca and Cayuga. It seems likely that 

 the ice blocked the present Keuka outlet long after the ice was removed 

 from the latitude of Potter. In no other way does it seem possible to 



*This delta, with an interesting kettle, which it includes, is made the subject of a separate 

 paper, Kettles in Glacial Lake Deltas. Journal of Geology, vol. vi, p. 589. 



