G. H. Stone — Wind Action in Maine. 137 



like the Bethel bowlders under consideration. On the average, 

 the depressions are perhaps shallower in proportion to their 

 breadth when made by water, than when the abrading agent 

 is driven by the wind. This kind of sculpturing I have seen 

 in a great many places both in Colorado and in Maine, almost 

 everywhere, in fact, where the rock does not weather faster than 

 the streams can polish the surface. Fine examples can be seen 

 at the Rumford Falls of the Androscoggin, also at the High 

 Falls of the Saco, near Hiram. Crystalline rocks are more 

 likely to show this kind of eroded surface than sedimentary 

 rocks, unless the latter are very firmly cemented. 



So far as my observation goes, sand sculpturing by flowing 

 water only takes place when the water strikes^ against a surface 

 that is stationary for a considerable time. This is not likely to 

 happen except in case of the solid rock or of bowlders of con- 

 siderable size. The swiftness of current necessary to produce 

 sand-carving will from time to time roll the smaller stones and 

 bowlders into new positions. The attrition and concussion of 

 the stones against each other so modifies the forms due to the 

 sand and gravel driven by the water that the stones are rounded 

 into pebbles, and do not receive the peculiar conchoidal depres- 

 sions and tremulous groovings due to pure sand-carving. 

 Thus, for instance, in the beds of the swift streams of the White 

 Mountain region, the fall is often 100 or more feet per mile, 

 and the bowlderets and smaller stones are well rounded like 

 beach or Kame pebbles and cobbles. Some of the Bethel speci- 

 mens are only two or three inches in diameter, toe small to 

 have been sand-carved by water under ordinary conditions. 

 But Bethel was not far from the end of the local Androscoggin 

 glacier, and it may be said that it is possible that the peculiar 

 carvings w T ere due to sediment-bearing sub-glacial streams acting 

 on stones and bowlders which were for a time held fast in the 

 ice. At several points in the Androscoggin valley, between 

 Gorham, N. H., and W. Bethel, Me., I have found surfaces of 

 the solid rock grooved lengthwise of the valley, i. e., nearly par- 

 allel with the flow of the local glacier, which it is well known 

 lingered here for a time after the great ice-sheet had melted. 

 These were not the ordinary glacial scratches, but had the 

 gouged appearance and conchoidal depressions made by sand- 

 laden water. Some of these places were above the highest level 

 of the river during the Champlain floods. A fair inference is 

 that the unusual grooves were probably due to the muddy sub- 

 glacial streams of the local glacier. But even if we admit sand- 

 carving of the underlying rock by sub-glacial streams, can we 

 do it in the case of moraine stuff held in the ice? I have failed 

 to find any recorded observation on this point, though the Al- 

 pine glaciers ought to afford facilities for such observations. If 



