226 SURFACE GEOLOGY. 



towards Umbagog lake. There is a pretty lateral moraine of granite 

 blocks at the uppermost landing of the steamboat in Wentworth's Loca- 

 tion, resting upon Huronian schists. The rocks for several miles up the 

 Magalloway are finely embossed. The numerous striae in Errol, S. 73° 

 E., have no connection with the Androscoggin, and we have no further 

 observations from this valley till we reach Gorham. The rocks here are 

 much weathered, but have the aspect of ledges struck from above, with a 

 few south-east striae. According to Vose, the strias upon the top of Mt. 

 Hayes, directly opposite the opening of the Peabody valley, run S. 40*^ E., 

 or nearly at right angles to the direction up the river towards the Glen 

 house. The Androscoggin valley, with its low, broad terraces, is very 

 different from the high sand-banks near the mouth of its Peabody tribu- 

 tary, suggesting the origin of the former under very unlike circumstances. 

 The utter absence of all material transported down the Peabody at Gor- 

 ham would indicate its removal by the powerful current of the Andros- 

 coggin, for certainly such deposits could not fail to have been brought, 

 and seemingly at a later date than the existence of the Androscoggin 

 glacial river. These facts suggest either extensive freshets in post-gla- 

 cial times, or the longer continuance of the Androscoggin glacier. The 

 valley below Gorham village, for two miles or more, consists chiefly of 

 rubbish from the sides, moved by the rains and frost, not in the terrace 

 form. No particular signs of glacial action are seen before reaching the 

 middle part of Shelburne, near the most northerly course of the river. 

 Here, at the height of 150 feet above the railroad, upon a projecting 

 ridge, are grooves pointing down the valley, S. 88°, N. 88° E. The east 

 side is a rough cliff, never touched by ice action. At the extreme north 

 point of the hill, the course is S. 63° E. On this lee side of the dome 

 are many rough granite blocks, each from five to twenty feet in length. 

 They have been transported only a short distance, and are properly a 

 moraine caught behind the projecting ledge. These markings are those 

 of a well defined local glacier. 



On reaching Clement's brook are other fine embossments, with strias 

 S. 73° E. The former have been described upon page 195. Local mo- 

 raines appear upon the projecting side of these ledges. Similar emboss- 

 ments occur on the north side of the river, opposite Clement's brook. 

 About a mile below Clement's brook, opposite the first house with a con- 



