Branner.] <34*j [Feb. 1<J, 



more sets of strife in one place, they are represented by the arrows cross- 

 ing each other at the proper angle. In a few instances the change in the 

 direction of the striae, either in the same place, or in places not far removed 

 from each other, has been so great that it might well be asked upon what 

 grounds I conclude that the flow of the ice was in the direction repre- 

 sented, and not exactly in the opposite direction. In such cases I have 

 depended upon the topographical features and the nature of the scratches 

 to settle the question — the lines frequently deepening in the direction of 

 the movement of the ice — and, as far as I am able to see, the results have 

 been satisfactory, though sometimes striking. 



The stria? themselves are of the usual character, modified by the rocks 

 upon which they occur. Those on the harder sandstones and conglomer- 

 ates are shallow, and frequently so fine as to produce a high polish, while 

 those upon the shales and softer sandstones are well defined and deep. 

 They are approximately parallel to each other, though often crossed by 

 other parallel sets of stria? pointing in different directions. Individual 

 marks frequently deepen toward the south, and end in a deep gouge. In- 

 stances occur of what were at first considered to be glacial grooves, but, 

 upon further and more careful study, these grooves were always found to 

 be channels in the rocks, polished and more or less modified by ice. In 

 one instance the impression of the trunk of a large Lepidodendron, lying 

 in the direction of striation, had been worn out so smoothly that it was for 

 some time mistaken for a glacial groove. The best defined stria? are found 

 where the glacier moved along upward, horizontal, or gently downward 

 gradients, and least prominent upon the steeper faces of hills that slope in 

 the direction in which the ice moved. Evidences of "upward flow " are 

 quite abundant, and where two or more sets of stria? occur in such a place, 

 those pointing upward are frequently, though not always, the deepest. 

 Furthermore, where the striae indicate an upward movement the glacier 

 appears to have moved forward with little or no regard to the smaller 

 details of topographical features. I would emphasize these smaller details 

 in this connection, for they had their share of influence later, as I hope to 

 show. 



The explanation of this upward movement and of these variations in the 

 direction of the ice stream must be sought in the topography of the region 

 and the varying thickness of the ice ; indeed, unless these matters be 

 taken into consideration, such phenomena are utterly meaningless. In the 

 conglomerate ledge east of Carbondale are a great many depressions, or 

 shallow holes, across which the ice has moved, to all appearances, without 

 being impeded or deflected perceptibly from its general course. These 

 depressions are of various sizes and depths, many of them being from a 

 few inches to two or three feet wide and one or two inches deep. Such 

 inequalities in the surface of the rock are not uncommon all through the 

 region under consideration, and they are doubtless to be found in all gla- 

 ciated countries. No one appears to be surprised that the ice should move 

 down one side of these shallow depressions and up the other, and when 



