Great Ice- Dams. — Taylor. ii 
causes. Where the lake basin was wide and had gentle side 
slopes the emerging places of the ice-dam would be far back 
on the sides of the lobe, so that the lake would run back on the 
two sides for many miles as long narrow arms between the land 
and the ice-front. Such an arm might be fifty or seventy-five 
miles long and average as little as ten or even five miles wide. 
In such a situation wave action would be less effective than in 
the open lake. And further, the fact that the shore of such an 
arm would be newly uncovered ground where the waves had 
had only a comparatively short time to act would also contrib- 
ute to the weakness of the beach produced. Whether a beach 
shall fade out gradually or end in stronger form and more 
abruptly, depends also in large measure upon the relation of 
the new outlet to the oscillations of the glacial retreat. While 
the ice-front stood at one of its halts the beach of the lake at 
that stage would have time to be formed in some strength, 
close up to the moraine, and if a new outlet were opened im- 
mediately on withdrawal from that moraine faint extension of 
the beach would never take place. Faint extension is favored 
most when the new outlet is not opened until just before or at 
the next halt and at the close of an active movement of retreat. 
This relation generally obtains where the outlet lies close along 
the face of a moraine.* 
The foregoing theoretical considerations are not imaginary, 
but are based on known facts. We know from the moraines of 
recession that the glacier retreated across the Erie basin in a 
certain manner. We reason, fvirther, that if the glacier acted 
as a great dam and the moraines mark its successive halting 
places, then certain changes and relations affecting the waters 
held in front of the ice must have followed substantially as 
pointed out above. It remains now to describe the things 
which have actually been observed and see how far the facts 
agree with theory. 
*An important factor of the lake history must also have arisen from 
the readvances of the ice-front during the oscillations of the general 
retreat. Outlets may sometimes have been opened for a short time 
and closed again and the level of the lake temporarily raised by a re- 
advance of the ice. Or, when no outlet was involved, a beach formed 
during extreme retreat might be encroached upon later by read- 
vancing ice, and moraines might thus be left in positions which would 
sTiow conclusively that this had happened. Actual evidences of such 
readvances have been found at several places, but for the purpose of 
the present paper a discussion of this interesting phase of the glacial 
retreat is not essential and is therefore omitted. 
