206 W. H. SHERZER 
much steeper on the concave side, with a well defined crest, 
while that upon the convex side of the curve is less steep and 
flatter. This is the reverse of what we find in the case of stream 
erosion. In places these troughs have the appearance of branch- 
ing and anastomosing, but this is probably due to the overlap- 
Fic. 5.— Mold and cast of a patch of miniature rock-basins, Sibley quarry. The 
ice movement was from above,downward. The continuous parallel grooves which pass 
diagonally across the cast were made by the preceding (late Wisconsin) ice move- 
ment, but were slightly emphasized by the movement which produced the basins. 
Careful examination of both the cast and mold will show that the miniature ice cur- 
rents were deflected by these grooves for a short distance, but that the general direc- 
tion was then resumed. 
ping of those which have the curved form. Indeed, we have 
here in miniature glacial canyons, fjords, and a great variety of 
rock-basins, showing all their essential characteristics. If these 
have been difficult to account for satisfactorily how much more 
