THOMAS.] ARKANSAS. 587 



fragments of pottery and occasionally whole vessels, stone chips, broken 

 i){)ues of animals, and other refuse material; immediately below this 

 is sometimes a layer of hardened muck or dark clay; at this depth 

 there is often found, in the eastern pai-t of the state, a skeleton, or 

 sometimes two. 



The burnt clay often contains impressions of grass or small twigs, 

 and sometimes is ornamented on one side by being stamped apparently 

 with an implement made of split reeds of large size. Hard-burned 

 fragments of this clay have given origin to the statement often made 

 that brick is found in southern mounds. 



That this clay is plaster from the houses of the mound-builders is 

 shown by the fact that remains of the upright posts and of the cane 

 lathing forming the walls of the building have been found, and from 

 other facts mentioned hereafter. 



These remains of the houses of the mound-builders of Arkansas are 

 not confined to the small mounds of the type mentioned, as they fre- 

 quently occur in tlie larger mounds, both of the pyramidal and conical 

 form. Sometimes the i-epeated building, burning, and covering on the 

 same spot results in forming a mound of considerable size, as, for 

 example, the one in Butler county, southeast Missouri, explored by Col. 

 Norris and heretofore described in this volume. So far as the explora- 

 tions which have been made indicate, these small, domiciliary mounds 

 appear to be CDuflned chiefly to the low alluvial sections of Arkansas, 

 and seem to have resulted from the following customs: First, that of 

 forming an eartheru platform a foot or two in height on which was 

 placed a dwelling, and, second, that of burying the dead in the floor, 

 then burning the house over them and covering the ruins with earth. 

 Examples illustrating these statements will l)e found in the first part 

 of this report. 



Kemains of ancient hoiises, apparently of the mound-building age, 

 and connected with or pertaining to mound groups, but not in mounds, 

 were discovered in a few localities. These were some 2 or .3 feet 

 beneath the surface of the ground with the usual mass of burned clay 

 plastering, remains of the posts of the walls, etc. In the two or three 

 cases observed, the buildings consisted of two or three rooms, each 

 about 10 or li' feet square. 



The ordinary conical burial mounds of Arkansas and also of Mis- 

 sissippi present marked differences from those of southern Missouri. 

 These are fully illustrated by the examples given in the reports of the 

 Bureau assistants. The chief variations were found in the condition 

 of the skeletons ; in some groups nearly or quite all were closely folded up, 

 though seldom in a sitting posture; in other localities they were found 

 chiefly lying at full length. In a few instances promiscuous masses, 

 like those of the northern section heretofore described, were found, but 

 these appear to be comparatively rare in the south. In some cases, 

 especially in the graves, remains ot the bark wrapping (apparently 



