JO A N CM E N T M N U M E N T S . 



Streams, or cut off the headlands formed by their junction with each other. 

 In such cases a fosse and wall are thrown across the isthmus, or diagonally from 

 the bank of one stream to the bank of the other. In some, the wall is double, and 

 extends along the bank of the stream some distance inwardly, as if designed to 

 prevent an enemy from turning the flanks of the defence. 



To understand clearly the nature of the works last mentioned, it should be 

 remembered that the banks of the western rivers are always steep, and where 

 these works are located, invariably high. The banks of the various terraces are 

 also steep, and vary from ten to thirty and more feet in height. The rivers are 

 constantly shifting their channels ; and they frequently cut their way through all the 

 intermediate up to the earliest-formed, or highest terrace, presenting bold banks, 

 inaccessibly steep, and from sixty to one hundred feet high. At such points, from 

 which the river has, in some instances, receded to the distance of half a mile or 

 more, works of this description are oftenest found. 



And it is a fact of much importance, and worthy of special note, that within the 

 scope of a pretty extended observation, no work of any kind has been found occu- 

 pying the first, or latest-formed terrace. This terrace alone, except at periods of 

 extraordinary freshets, is subject to overflow.* The formation of each terrace 

 constitutes a sort of semi-geological era in the history of the valley ; and the fact 

 that none of the ancient works occur upon the lowest or latest-formed of these, 

 while they are found indiscriminately upon all the others, bears directly upon the 

 question of their antiquity. 



In addition to the several descriptions of defensive works above enumerated, 

 there are others presenting peculiar features, which will be sufficiently noticed in 

 the plans and explanations that follow. These plans are all drawn from actual and 

 minute, and in most instances personal survey, and are presented, unless otherwise 

 specially noted, on a uniform scale of five hundred feet to the inch. When there 

 are interesting features too minute to be satisfactorily indicated on so small a 

 scale, enlarged plans have been adopted. This is the case with the very first plan 

 presented. Sections and supplementary plans are given, whenever it is supposed 

 they may illustrate the description, or assist the comprehension of the reader. To 

 shorten the text, the admeasurements are often placed upon the plans, and the 

 "Field Books" of survey wholly omitted. The greatest care has, in all cases, 

 been taken to secure perfect fidelity in all essential particulars. In the sectional maps, 

 in order to show something of the character as well as the positions of the works, 

 it has been found necessary to exaggerate them beyond their proportionate size. 

 Some of the minor features of a few works are also slightly exaggerated, but 

 in no case where it would be apt to lead to misapprehension or wrong concep- 

 tions of their character. 



* This observation is confirmed by all who have given attention to the subjecl, in the Ohio and Upper 

 Mississippi valleys. Along the Gulf, and at points on the Lower JVIississippi, where the entire country 

 is low, and subject to inundation, and where the operation of natural causes is rather to elevate than 

 depress the beds of the streams, some of the ancient works are invaded by water. 



