GLACIAL PHENOMENA IN MAINE. 
113 
moraines formed completely under the glacier, 
and resting immediately upon the rock or soil 
beneath. Of course, masses of drift below a 
great sheet of ice, moving steadily in the same 
direction over uneven, rocky surfaces, cannot 
preserve the same thickness throughout. Here 
and there the incumbent weight will press 
more heavily in one direction than in another, 
thus crowding the loose materials together, 
rolling them into ridges following mainly the 
direction of the movement. Occasionally such 
uneven pressure may drive these materials up, 
from either side, along the summit of a rocky 
ledge, or heap them at any height upon its 
slope. We have seen that the horsebacks, 
though uneven and winding, usually run from 
north to south; but occasionally also they 
trend from east to west. This is the case 
where a morainic accumulation of loose mate¬ 
rials may have been pushed forward, along the 
margin, in front of an extensive sheet of ice 
moving southward, and then left unchanged 
by the subsequent retreat northward of the 
whole mass. I conceive that such horsebacks, 
running east and west, may be compared to 
terminal moraines, which, as is well known, 
