CONANT ARCHAEOLOGY OF MISSOURI. 365 



First, as to the different degrees of preservation of the skele- 

 tons. While all were so decayed that it was possible to secure 

 only three or four skulls — the under jaw seldom or never com- 

 plete — the bones were usually so far gone as to have no more 

 consistency than the sand which covered them. In many instan- 

 ces, only a faint streak, or whitish line, in the earth, as successive 

 sections were shaved down by the spade, would disclose the form 

 of the head. And, again, the jugs and bowls, or pans, would be 

 found in their relative positions, but no trace of the skeleton what- 

 ever ! These facts prove, as remarked above, that many years 

 must have intervened between the first and last interments. 



Again, the high bank which marks the boundary of the bayou 

 shows the long continued action of flowing water. To those who 

 are acquainted with the habits of this capricious river, the state- 

 ment that here was once the channel of the Mississippi, will cause 

 no surprise. The town of New Madrid, eighteen miles away, is 

 a moving town, and slowly retiring before its resistless encroach- 

 ments. In 1804 its site was one mile farther east than now, and 

 upon what is now the eastern shore. The towns and cities of the 

 mound builders were almost always upon the banks of some 

 river. If it be granted that this high terrace upon which they 

 stand was once the bank of the Mississippi, and that they were 

 abandoned when the river left its ancient bed, the question arises, 

 is there any ascertainable uniform rate of recession by which the 

 centuries may be estimated which have passed since that time? 

 Perhaps the question admits of no satisfactory answer. Still, we 

 will venture one speculative estimate. The river has receded (as 

 at New Madrid) about one mile in seventy years. Assuming the 

 mean distance of these works to be now fifteen miles inland, the 

 computation upon these premises would give, in round numbers 

 one thousand years as the time of their desertion. Another ques- 

 tion arises here relating to the length of time required for the 

 cypress swamp to take possession of the deserted channel of the 

 river. Does it follow it up by slow degrees, or appear at one and 

 the same time throughout its length ? 



One other fact only can be noted in this connection. This whole 

 region has been terribly shaken and disturbed by earthquakes. 

 During the earthquake of 1S11 the old town of New Madrid dis- 

 appeared, the river changed its southward course and for several 



