GLACIAL ORIGIN OF CHANNELS ON DRUMLINS. 11 



feet, is at the \ioint where it cuts the highest portion of the hill, and south of 

 that the slope of the bed is gentle — a little less than that of the hill itself— until 

 it readies the base. 



A few rods south of the deepest point the depth is 2() feet, while at the same dis- 

 tance nortli it is but 17^ feet. In each case these depths are taken from a line con- 

 necting the crests on each side, that on the west being always slightly lower — two 

 or three feet— than that on the east. 



At one point the bed is o'n feet higher than at a point some hundred feet farther 

 up stream, but this is a common feature in the beds of streams with considerable 

 sloi)e. The greater depth at the latter point coincides with a bend of the stream 

 that would produce greater velocity and a deeper erosion. 



The bed is nearly free from bowlders, as is also the surface of the hill. This 

 absence can be accounted for by the numerous stone walls in the construction of 

 which all superficial stones have been used. A section of the interior of the hill 

 has been exposed on the east side by sinkintr a well 12 feet in diameter and 50 feet 

 in depth. The material excavated was a compact blue till, containing only a smal 

 number of stones, not exceeding one or two feet in diameter, most of which were 

 well glaciated. The small number here present would also indic-ate that but few 

 would be left by erosion in the bed of the stream. 



The outline of the hill, exclusive of the channel, indicates that it had been 

 formed as a symmetrical, well shaped drumlin, and that in this completely formed 

 hill erosion has taken place by some means directly along its axis. 



Origin of the Channels 



Now, by what means and at what time was the action of a stream of water in- 

 augurated at this elevation above the general level? It is necessary .to suppose that 

 at the time the stream came in contact with the north end of the hill it must have 

 been flowing at an altitude as high or higher than that point. As this end of the 

 hill is symmetrical below the channel and shows but slight, irregular indications of 

 erosion, there must have been some material occupying this region to form a bed 

 for the stream whose disappearance has left no visible trace. The only conceivable 

 material that could so disappear is ice. Hence, at the time the channel was eroded, 

 the ice-sheet nuist have lain to the northward, with a thickness exceeding the 

 height of the hill and so impacted u])on its northern slope that no current could 

 l)ass downward in that direction. The stream, flowing over an ice-bed, came 

 directly in contact with thesummitof the hill, and thence ouAvard eroded its chan- 

 nel in till to the base of the hill, where it lost its cutting power, and deposition 

 ensued. Before reaching the hill the stream may have been superglacial or engla- 

 cial. It could not have been subglacial ; but from this point it may have been 

 subglacial, as the ice-sheet may not have then entirely disappeared to the south- 

 wanl, or it may have entirely disajipeared in that direction and the stream have 

 thence l)een open to the sky. 



Just at the bottom of the hill, in the broad mouth of the channel, is an esker- 

 like ridge of a few rods in length, while the valley between this and the hill to the 

 south is filled with kanie material, some of which may Jiavctome from this source 



Over the irregular mass of till to the south and in line with this channel win<ls 

 an esker as a broad, flat ridge ; but on the south sloi)e the esker is replaced by a 



