134 - NINTH REPORT. 



broad erosion valley, more than a mile wide and with a forty foot bank on 

 the north side, in the middle of which the present insignificant stream now 

 runs in its own rather deep, narrow valley. There are at least two small, 

 partly buried moraines across this sand-]3lain, and the plain ends abruptly 

 on the south side in a steep descent to the south, which is boulder-covered 

 and has all the characteristics of the ice-side of a moraine, while the north 

 side is so level that it is difficult to see where the sand-plain ends until the 

 edge of the southward slope is reached. 



The conclusions reached from a study of these facts are: 



(1) That there exists in northern Marquette county an area of several 

 townships' extent which is almost without glaciation. This land rises to 

 nearly or quite 2000 feet above tide in the higher parts, and is 100 or 200 

 feet lower in the valleys. 



(2) From the fact that the drainage was across this highland and fol- 

 lowed pre-glacial rock valleys, while the ice was piled up around its outer 

 margin, it is evident that this area must have been early abandoned by the 

 local ice-cap which covered it. 



(3) From the position of the moraines, and the drainage lines, it seems 

 evident that the ice lying south of the uncovered area was moving in from 

 the west and not from the east, and while this is not yet established, it seems 

 probable that the direct movement of the ice from the northeast was prac- 

 tically checked by the Marquette highland and by the Copper Range. 



(4) The region to the south of this was covered by ice which pushed in 

 on the west side of the Keweenaw Range, with a generally southeasterly 

 movement, which spread out on the slope lying south of the highland, up 

 to, but not over, the cliffs bounding this, and formed weak moraines as far 

 east as Clarksburg, and possibly as far as the complicated region about Ish- 

 peming and Negaunee, which has some characteristics of an interlobate 

 area. This southeastward movement is indicated especially by the present 

 stream valleys which follow lines of ice drainage diagonally across the gen- 

 eral slope of the land, and l)y the fact that the clayey red till character- 

 istic of the high moraine assumed to limit the movement of ice from the 

 northeast, ends abruptly with this moraine, and is replaced in the adjacent 

 lower moraines to the south of it, assumed to be formed by western ice, 

 by gray till of a much more sandy structure. 



(5) The presence of strong moraines running east and west to the south 

 of the highland, which have drainage lines along their northern sides, indi- 

 cate that the axis of movement lay to the south, and adds to the probability 

 that the ice forming them was moving from the westward down the land- 

 slope, rather than from the east up the slope. In the latter case it would 

 seem as if the chief moraines would have been formed about northeast and 

 southwest, since the thicker ice and more rapid movement would have been 

 in the low lands to the south. In case the ice had pushed through the rela- 

 tively narrow passage at the head of the Keweenaw embayment and spread 

 out to the southeast, the main axis of movement would have been south- 

 ward, and it would seem that the moraines would have been more or less 

 concentric at right angles to this line, and the thin, more remote ones would 

 have run northeast and southwest on the eastern side of the axis. As there 

 is no evidence that the ice pushed in from the northeast over the Laurentian 

 highland and much that it did not, but a word need be said regarding this pos- 

 sibility. In case a movement is assumed from this direction, it is impossible 

 to explain the presence of the morainal bank against the foot of the bounding 



