FOWKE] ANTIQUITIES OF MISSOURI 97 
only of Missouri, it may be said they exist by thousands, amid sur- 
roundings of every sort. Some extend in close succession for miles at 
the rate of two to seven or eight on every acre, broken, perhaps, now 
and then, by a slough or a cypress brake, beyond this continuing as 
before. Nearly all the smaller ones.are made either of reddish, sandy 
clay, such as forms the upland subsoil, or of gumbo from the low 
grounds, consequently they are almost invariably infertile, so much so 
that in a field covered with vegetation their location may be deter- 
mined by the sparser growth on them. This feature, which was 
repeatedly stated as a fact by farmers, entirely destroys the hypoth- 
esis that the mounds were erected for agricultural purposes, though 
this is probably true of some mounds here and there on which good 
crops are raised. 
Nor is the theory more plausible that these mounds were erected 
for camping places by hunters, for there are more of them than 
there would be animals to hunt at one time. 
They were not residence mounds, for there are lacking around 
them the usual indications of an Indian domicile—broken pottery, 
bones, fragments of implements, etc. 
Nor are they burial mounds, for with rare exceptions they are 
barren of contents. 
The mounds are found in various situations from well-drained high 
levels to swamps where, after a period of prolonged dry weather when 
the roads become hard and even dusty, they can be reached only by 
carefully stepping on roots and stumps to keep out of the mud and 
water surrounding them. Moreover, scores, even hundreds of them, 
in the compass of a few square miles, are built on sticky clay low- 
lands within a few yards or rods of sandy ridges, never muddy, where 
the ground is dry on the surface within a very short time after the 
heaviest rains, and where various crops yield abundantly. 
It is evident that all the land where the mounds now stand, and 
perhaps the entire country except the present swamps, was prairie 
land, destitute of timber at the time they were built. The sloughs 
and bayous mark the courses of former streams, and it may be that 
when the mounds were made these were open channels clear of timber, 
affording free passage in canoes. This would explain the location of 
village sites along their margins. With the actual levels of upland 
and water courses remaining as at present, if there were no forest 
growth at all there would be no great overflows except backwater 
from the Mississippi, because local rainfall would rapidly drain away. 
Many floods now are caused by the choking up of streams with drift- 
wood; if this were all cleared out, to give an unobstructed flow, water 
would pass away in half the time or less. 
5780—Bull. 37—10——7 
