Great Ice- Dams. — Taylor. g 
But the series from Cincinnati to Toledo and thence by two 
branches to Mackinac straits and Rochester, N. Y., are be- 
Heved to be complete. 
Guided by this theoretical outline of the glacial recession, 
the story of the lake history is greatly simplified. For by fol- 
lowing up the related pairs of moraines in the two diverging 
series, the successive positions of the great ice-dams may be 
easily noted in close connection with the principal concurrent 
changes of area, outlet and level which took place in the con- 
temporary lakes. 
Tlieorctical Relations of Ice-dams, Outlets and Beaches. 
A word or two on the theoretical relations between ice- 
dams, outlets and beaches will give a clearer setting to the 
facts to be considered later. As it moved over the uneven 
country, the plasticity of the glacier caused it to yield readily 
to the influences of topography. Elevated areas of land held 
the advancing ice-front back. The ice pushed forward in the 
wide, open valleys and lagged behind on the intervening high- 
laTids. Thus, the line of the ice-front was made uneven ; lobes 
projected forward in the valleys and re-entrant angles set back 
upon the hills.* New outlets were opened only as the retreat- 
ing ice uncovered lower ways of overflow at the sides of the 
basin. The moment a passage lower than the water surface 
was opened on either side across the retaining land barrier, the 
water rushed through and a new outlet was established. If 
the new outlet stood at the same level as the old one the dis- 
charge would be divided between them and the lake would 
thereafter have two active outlets, but if the new outlet stood 
much lower it would take the whole discharge. In either case 
*It should be borne in mind that whether the ice-front was retreat- 
ing or advancing, the ice itself was always moving forward. Thus, 
topography modified the line of the ice-front during movements of re- 
treat in precisely the same way that it did during forward or ad- 
vance movements. 
'Note toFicf.l. In New York, Pennsylvania, Ohio and Indiana the moraines are 
sliown as mapped by Levorett. In northwestern Ohio they foUow tlie earlier work 
of Gilbert and N. H. Winchell and in northeastern Indiana that of Dryer. In Mich- 
igan, except along the eastern border, the mapping follows Cliamherlin. The 
writer's contributions are mainly along tlio eastern border of Micliigan and in 
Ontario. The hypothetical continuation of moraines across the water-filled lake 
basins and across certain regions which remain unexplored, or where tlio identity 
of the individual moraines is doubtful, are represented by two parallel broken lines. 
The several old outlet channels are shown by pairs of heavy parallel lines. 
