768 GEOLOGY OF OHIO 
Cod; Rockaway, in New Jersey; Salamanca, in New York; and in 
Ohio, near Canton; near the northeast corner of Knox county ; near 
Lancaster; and at Adelphi. The map of Indiana likewise shows a 
remarkable bend to the north from the neighborhood of Louisville, 
inaking the glacial border sweep around a large, unglaciated triangle in 
the southern part of the State, of which the northern part of Brown 
county is the apex, the Ohio river from Louisville to Mount Vernon 
the base. These changes of direction are so abrupt as to cause much 
trouble in discovering the line. In searching for a cause of these 
sudden changes in direction, one soon finds, however, that they are not 
so difficult to explain as they seem. Ice behaves not like a solid, but 
like a semi-fluid. If an oblong block of ice be suspended upon the 
ends it will gradually sag in the middle. If a strong hollow sphere be 
filled with water, and a good-sized orifice be left through which the ice 
may escape, and the whole be subjected to intense cold, the ice will 
project through the hole for a considerable distance. As a matter of | 
fact, ice flows like cold molasses or half-hardened lava. 
It is not necessary to have a steey declivity in order to secure 
glacial motion. Ice can move in any broad valley where water would 
run. In our conceptions of glacial movement we are in danger of 
having our ideas cramped by the contemplation of Alpine glaciers. 
The demands made upon our imagination by the glacial phenomena of 
North America are, to some, almost staggering to reason. We are 
called upon to believe that along a line thousands of miles in extent 
the ice-front of the great glacier rested upon land which is nowhere 
much lower, and in many places is actually higher than the region from 
which it was dispersed. Boulders in many cases have been raised to a 
higher level than their native ledges. 
Upon reflection, however, this is not so paradoxical nor so ex- 
travagant as at first glance it seems. It should be remembered that 
glacial ice is formed not by the freezing of water upon lakes and oceans, 
but by the accumulation of snow, which, under its own pressure, becomes 
converted into ice. If, now, over an extensive level surface, there 
should annually accumulate six feet more of snow than melted, 6,000 
feet of ice would accumulate after a thousand years. It is thus easy to 
see that after a time the ice might form a mountain plateau by itself, 
and, owing to its semi-fluid character, it would gradually move along 
whatever lines presented the least resistance. Such accumulations about 
