J38 ANCIENT MONUMENTS. 



base, one of which is six thousand feet long and extends across the isthmus, the 

 other runs along the bank of the stream. The extremity of one of the long 

 walls terminates in a species of citadel, of semi-circular shape, which has horn- 

 works and curtains defending the gateways, and also protected ways to the river. 

 Here are a number of mounds and excavations similar, in all respects, to those 

 characterizing the defensive works generally. This work is not entire, having 

 evidently been greatly encroached upon by the river, which is constantly under- 

 mining the elevated terrace upon which it stands. Still beyond this point, upon 

 the Platte, Kanzas, and the numerous other tributaries of the Missouri, many large 

 and interesting works are said to occur.* They have been remarked high up the 

 streams, in the valleys overlooked by the Rocky Mountains. But little more 

 than the fact of their existence is known ; of their character we are ignorant. 



In the vicinity of the city of St. Louis formerly existed a very large and 

 interesting group of works, consisting mainly of a series of mounds so arranged 

 as to constitute the sides of a parallelogram. These mounds were generally square 

 or oblong, with level summits. Some were terraced, bearing a close relationship to 

 those in Mississippi and Louisiana. A few conical mounds occurred in connection 

 with them, but there was neither embankment nor ditch. All were situated upon 

 the second terrace. 



The most interesting feature of the group is the singular work yet preserved, 

 denominated the " Falling Garden." This, as described by James, consists of a 

 succession of terraces, artificially formed from the bank of the natural terrace, 

 which is here upwards of fifty feet in height. The lowest of these terraces is 

 eighty-seven feet broad by one hundred and fourteen long; the second is fifty-one 

 feet, and the third thirty feet wide. Their slopes are regular, and the aspect of 

 the structure that of a Mexican teocalli of four stages.f 



Structures of brick are mentioned as occurring both in Missouri Und Arkansas, 

 but their existence is not sufiiciently well authenticated. There are, without 

 doubt, numerous remains scattered over the territory embraced in Arkansas, 

 Texas, and New Mexico ; and it is not impossible that their investigation would 

 result in developins the fact that there is a gradual transition, from the earth- 

 works of the Mississippi to the more imposing structures of brick and stone of 

 Mexico and Central America.:j: 



* Lewis and Clarke, p. 48. 



■f Views of Louisiana, p. 189 : James, Expedition to Roci<3' Mountains, vol. i. p. 314. 

 I " The distance from the large mounds on the Red liver to thosp in New Spain is not so a^roat. (nit. 

 that they maj' he rpacarded as existinaj within the same country." — Hrerkenriflye. 



