284 J. W. GOLDTHWAIT GLACIATION IN THE WHITE MOUNTAINS 



(a) The southeastward advance of the Canadian ice-sheet over the Car- 

 roll district during the late stage of glaciation, registered by southeast- 

 ward stria? on the hill near the Twin Mountain station, picked up blocks 

 of "Mount Deception" granite from ledges there and carried them to the 

 fields at the foot of the hill, east of the Twin Mountain House, (b) As 

 the ice departed from the district, it withdrew first from the flanks of 

 Cherry Mountain and the Twin Mountain Eange, still covering the low- 

 lands of the Ammonoosuc and Carroll Stream. Along one of the main 

 lines of glacial drainage, in a canyon or tunnel draining southward from 

 the Beech Hill-Cherry Mountain gap, the high esker at Twin Mountain 

 House and related kames were built, (c) As the ice melted out of the 

 pass, and as the Ammonoosuc Valley around Twin Mountain station, 

 still dammed by the ice-sheet between Bethlehem Junction and Wing 

 Eoad, came to be occupied by a lake, the continued discharge of glacio- 

 fluvial torrents southward through the gap built the two pitted plains 

 and the connecting kame terraces, enveloping and extending beyond such 

 stagnant ice-masses as still remained. Subsequently, as it receded from 

 the northern edge of the Carroll plain, the melting ice-sheet dumped its 

 debris, still in ponded waters, in a confused mass of hillocks, forming the 

 kame field north of Carroll. 



RELATION OF THE CARROLL MORAINE FIELD TO THE BETHLEHEM MORAINE 



The exact relation of the ice contacts and morainic ground of Carrol] 

 to those of the Bethlehem district is not yet known, but judging; from 

 the persistent northeast trend of the main lines of the Bethlehem moraine, 

 as shown on the map (plate 13), it appears that the ice departed from 

 the Carroll district at a somewhat earlier stage. The small and frag- 

 mentary moraine lines east of Maplewood and north of Bethlehem Junc- 

 tion are nearly in line with the Carroll moraine field ; but time did 

 not allow us to trace these northeastward around Beech Hill. The coun- 

 try is thickly wooded, swampy, and not traversed by convenient logging 

 roads or trails. If there is any actual connection between the deposits 

 west of Beech Hill and those east of it, a good deal of field-work will be 

 required to determine it. 



Ammonoosuc Glacier 



discussion of evidence presented by earlier observers 



Inseparably connected with the problems already discussed, yet pre- 

 senting elements of its own, is the question of the former existence of a 

 local Ammonoosuc glacier. Of the several local glaciers which Professor 



