134 MOUND EXPLORATIONS. 



been destroyed, as pieces of human bones are very common in the 

 plowed fields. 



The location, forms, and heights of the various mounds of the Cahokia 

 group as given in theannexed diagram (PI. vi) are from Mr. McAdams's 

 survey and are believed to be strictly correct. The figures on or by 

 the mounds indicate the height. 



The next excavations were made in the mounds at Mitchell, on Long 

 lake. The principal digging was done in the base of the large mound 

 through which the railroad tracks rnn. Bones and sea shells had beeu 

 discovered here by some workmen in digging a trench through the base 

 of the mound between two railroad tracks for the purpose of laying a 

 water pipe to the lake. This ditch was reopened, then widened out, and 

 the spot fully exploited. There seemed to have been 4 or 5 skeletons of 

 adults, which lay east and west. A great number of whorls of sea 

 shells had been buried with them, probably taken out in forming drink- 

 ing cups or water vessels. These shells are from a few inches to a foot 

 or more in length and belong mostly to the genus Busycon. 



The mound from which these shells were taken was nearly square in 

 shape, 100 paces on each of its sides, 25 or 30 feet in height, with a flat, 

 level summit. It is now, with the exception of a small portion in the 

 center between two railroad tracks, obliterated, a part only of the base 

 remaining. 



In removing the western side of the mound a few years ago, to make 

 a road across Long lake, many human remains were found and, with 

 them, implements of stone, bone, and copper. The mound was composed 

 principally of black dirt or soil, and wherever excavations were made 

 in the base, at the depth of 3 or 4 feet, the original under soil of the 

 surrounding praiiie, a yellowish sandy loam, was reached. This is the 

 mound ftom which Mr. Henry R. Howland obtained the copper articles 

 described and figured in his paper in the bulletin of the Buft'alo Acad- 

 emy of Sciences, 1877. 



In addition to the maps already given, Mr. McAdams prepared a map 

 of the western part of Mswlison county, including one range of sections 

 in the northern part of St. Clair county, showing the location and rel- 

 ative positions of the various groups of mounds named. This map is 

 shown (on a reduced scale) in PI. vii. 



RANDOLPH COUNTY. 

 STONE GRAVES ON THE MILL TRACT. 



These are situated aboul half a mile north of Prairie du Rocher, on 

 a long ridge that runs in a westerly direction nearly across the nar- 

 row bottom of a small creek that flows through the village. This 

 ridge, which is about 25 feet higher than the bottom land, descends 

 gradually from the hills to the west, having a steep slope on each side. 

 The soil is yellow, tenacious clay. The graves were on the rounded top, 



