G. H. Stone — Glacial Sediments of Maine. 133 



ordinal*} 7 osars or two-sided ridges. Going south from five to 

 twenty-live miles we find in many cases the ridges having rather 

 steep lateral slopes expand into level-topped plains up to a 

 half mile in breadth. These plains are usually, but not always, 

 finer in composition than the narrow ridges of which they are 

 the extension. Measured transversely the tops of these plains 

 are horizontal, but measured length-wise, the broad osars are 

 found to go up and over hills just like the osars. In numer- 

 ous instances the osar-plains have been deeply eroded by 

 streams and boiling springs. A common type of erosion is 

 where a central ridge has been left uneroded, also a terrace on 

 each side of the original plain. Thus two valleys have been 

 cut down into the plain, one on each side of the axial ridge. 

 This central ridge has the same height as the lateral terraces, 

 or sometimes it is higher. The material of the ridge is coarser 

 than that of the rest of the plain. A good example of an 

 eroded osar-plain is found in the towns of Woodstock, Milton, 

 and Rumford, on the line of the great gravel system that ex- 

 tends from the upper Androscoggin Lakes to Portland. Here 

 an osar-plain extended across the whole valley of a stream 

 flowing northward into the Androscoggin River. Its average 

 breath is from one-fourth to one-third of a mile. The plain is 

 now deeply eroded so as to leave an alluvial terrace on each 

 side of the valley and a prominent central ridge widely known 

 as the Whales-back. A great glacial river here flowed from the 

 Androscoggin Yalley at Rumford Point, southward over a 

 divide at North Woodstock more than a hundred feet higher 

 than Rumford. The osar-plain is composed of sand near 

 Rumford, and becomes coarser as we approach the divide at 

 North Woodstock, from whence the glacial stream flowed 

 down the valley of the Little Androscoggin River. 



The mode of formation of the broad osars was approxi- 

 mately as follows. Firstly a ridge of gravel was deposited in 

 a narrow channel between ice walls. We need not now inquire 

 whether this was a sub-glacial tunnel or a superficial channel 

 cut down through the ice to the ground. By degrees this 

 channel enlarged laterally by the melting and erosion of the 

 ice until at last it became several, sometimes many times as 

 broad as the original channel. It is very common to find the 

 gravel at the sides of the osar-plain much less water-worn than 

 that found in the central parts. This points to much less 

 violent water action when the stream was widest. The broad 

 osars seldom if ever are overlaid with bowlders having till- 

 shapes. • It appears incredible there should be a sub-glacial 

 channel a half mile wide, overarched by ice, yet the sediment 

 carrying no till bowlders fallen from the roof. Whatever we 

 think of the narrow glacial rivers as to their being sub-glacial 



