DEPOSITS ALONG THE MISSISSIPPI 351 
also to be sought, a reason why a stream should be turned from a 
wide, deep channel, into a narrow gorge which lay higher than the 
stream itself. 
At Thebes, near Cairo, is a similar gorge; for fifteen miles 
_there are bluffs along both sides of the river which is bordered 
only by narrow strips of alluvial land, while at one place, “The 
Grand Chain,” masses of rock proj jecting above the surface compel 
pilots to hold their boats in a very narrow channel. At Cape 
Girardeau, a few miles above, the loess caps a hill fully 170 feet 
above the level of the river bank. Here, again was an obstructed 
ancient channel; and the question of a solid ice-dam was answered 
in the negative at once, for the greatest southern extension of the 
glacier was many miles above. It was deemed possible, though 
not at all probable, that icebergs or floes may have found some 
obstacle to hold them at this point until they had formed a nee 
pack. Beginning near Cape Girardeau is a swamp fully three 
miles in width and terminating more than fifteen miles below, 
which was the former course of the Mississippi. Bluffs border it 
on both sides, in most places precipices of solid limestone. As at 
Grand Tower, no trace of glacial drift could be found above high- 
water mark; besides, the valley of the swamp is too wide and too 
deep for ice to have jammed. Below here, are reached the ex- 
tensive swamps that fill the former prolongation northward of the 
Gulf of Mexico; and farther research was useless. 
Tt thus was evident that by no possibility could loess deposits 
south of the Missouri river be due to a dam of either earth or ice; 
and some other explanation must be worked out and investigated. 
Wright has calculated, and brought forward proof of his 
figures, that at its greatest discharge during the melting of the 
continental glacier, the Missouri reached a flood height of at least 
two hundred feet. 
On the Illinois river, sixty miles above its mouth, are bluffs 
of loess fully one hundred feet high, proving this stream also 
subject to great floods. 
At the same time the Mississippi was draining a large area of 
ice-covered country. 
The border of the ice-sheet being in this region along a line 
approximately east and west, these rivers orld discharge their 
immense volumes of summer water at practically the same time, 
and not with intervals between flood height, as is now the case. 
It is quite probable that the rise in the Mississippi, when reinforced 
by that from the Illinois, fully equalled that in the Missouri. 
When_all these waters united, in a channel not much wider in many 
places than in one of the three, it follows that, either the current 
must flow with great velocity or the water must rise to a level 
greater below the junction than it would naturally. Bearing upon 
this point is an observation by Leverett, who says that “between 
than at the present day. Yet it seems probable that at times the 
