The great majority of the scraping implements obtained from the mounds, graves, and shell heaps are simply valves of _Unio_ or clam-shells, unaltered except by use; yet there is a widely distributed class of worked specimens, which have been altered by making a rough perforation near the center of the valve, and by the grinding down and notching of the edges. A very fine specimen is illustrated in Fig. 3, Plate XXVI. It is formed of the left valve of a _Unio tuberculosus_. It was taken from a mound at Madisonville, Ohio, and is now in the national collection. A similar specimen from the same locality is illustrated in an account of the exploration conducted by the Scientific and Literary Society of Madisonville.[41] I have seen four other fine specimens from the same locality; all are made of the shell of the _Unio tuberculosus_ (?). It will be seen by reference to Fig. 3 that the posterior point of the shell is much worn, as if by use, while at the opposite end, near the hinge, the margin has been slightly notched. The large specimen, figured in the Madisonville pamphlet, as well as all other examples from this locality, are also much worn at the posterior end, and slightly notched on the anterior margin. The perforations are roughly made, and nearly one-half an inch in diameter. I have carefully examined all the specimens of this class within my reach, probably twenty-five in all, most of which are in the national collection, and I find them all very much alike. They are from two to five inches in length, have rude central perforations, and are worn by use at the posterior point, and notched on the anterior margin. The blunting of one end by use calls for no explanation, but the purpose of the perforation is a little obscure. It may have been used for convenience in transportation, but more probably for attaching a handle. On discovering that a notch had in all cases been made at the upper end, I became convinced that the latter use was intended. Whether the supposed handle has been long or short, or attached longitudinally or transversely, I am unable to determine. In Plate XXVI, Figs. 4 and 5, two methods of hafting are illustrated. If used for striking, the long handle would be the more suitable, but if for scraping, dressing skins, scaling fish, or shaping wood or clay, the handle suggested in Fig. 5 would be the most convenient. The clam-shell agricultural implements, so frequently mentioned by explorers along the Atlantic coast, were attached to handles in the manner of hoes or adzes, as shown in Fig. 2, Plate XXVII. It is possible that the specimens under consideration may have been hafted in this manner. A perforated valve of a _Unio gibbosus_, which has probably been used as a knife or scraper, is shown in Fig. 1, Plate XXVII. It was obtained from a cave near Nashville, and is now in the national collection.