THE CONNECTICUT RIVER LAKES. 613 



the latter l»y a si-arp rising- up from the terrace (the bank of tlie river 

 when the terrace was part of its bottom), and Hmited towai-d the river by a 

 descending- scarp which was the river bank at a later time, generally when 

 the terrace made part of its flood plain. Corresponding- to this, we shall 

 find the plain covered with the meadow loam laid down in the floods of the 

 river, and under this loam the strong river-bottom sands of an earlier date, 

 the last underlain unconforraably by older deposits which the river had not 

 reached and eroded while it flowed above them. 



On tlie other hand, we shall tind the highest terrace or bench bounded 

 outwardly by a slope which, as to its material and structure, has no relation 

 to the river. At most, the river has undermined this more or less exten- 

 sively at the water level, and, by caving, an escarpment of till, sandstone, 

 or gneiss has resulted. The ten-ace itself, widening into extensive sand 

 or g-ra^'el plains where the allu\aal cones or deltas of the side streams were 

 thrust out into the lake, narrows in places remote from these, and its level 

 is often represented by shelves in the sandstone scarcely covered by sands, 

 or in the till deeply ccmcealed by gravels concentrated from the till itself 

 Pursuing the same level, we soon come upon the continuation of the normal 

 sand beds which make the bulk of the bench. 



Inwardly, h()\^"ever — that is, toward the center of the lake — esjjeciallv 

 around all the Hadley basin and its prolongation in the Deei-field and East- 

 hampton valleys, the terrace is for the most ^lart bounded, not by an 

 escarpment of steep and constant pitch — an abandoned river bank — but by 

 the slope of passage from shallow to deep water. This is sharpest and most 

 constant on the face of the large deltas (but here of less angle than in the 

 former case, as the highest angle at which sands come to rest under water is 

 less than that assumed in air), less and less marked in other places, until at 

 last the case occurs where from the rocky bank the sands pass with gentle 

 and continuous slope to the deepest central line, where was the thread of 

 the current, and rise in the same way to the opposite bank. 



This slope of passage I have called a scarp of deposition, or, as locallv 

 synonymous therewith, the delta front, in contradistinction from the ordi- 

 nary scarp of erosion. On the map the normal high terrace or bench 

 (1 s h, PI. XXXV) and its widening- into great delta flats are not separately 

 indicated. One passes by a scarp of deposition to the broad area of the 

 old lake bottom (1 b t), which was synchronous with the bench itself. 



