_Celts._ Celts made of stone such as were common in the Thompson River region[158] were not found by us in the Yakima region; but one typical specimen which apparently resembles the celts found on Puget Sound, more than it does those found in the Thompson River region is shown in Fig. 45. It may be seen in the collection of Mr. McCandless who secured it from an Indian at Ellensburg. This celt is made of serpentine and is 190 mm. long.[159] A similar specimen, in the same collection, resembles this one but shows grooves along the side edges by means of which it was cut out. There is a celt made of green serpentine, only about 3 mm. thick in the collection of Mr. Owen, but it was found at Umatilla, Oregon. [158] Smith, (d), p. 142; (c), p. 415. [159] Museum negative no. 44507, 6-8. Celts of jadeite (?) narrow and oblong were found on Snake River above Lewiston in the Nez Perce region to the east.[160] Spinden states that these were evidently acquired by trade from natives of the northwest coast and that they have been cut by grooving and breaking. Also, that this method and material was not employed by the Nez Perce who considered the objects to have been used as wedges. I am inclined to believe, therefore, that these more nearly resemble the celts of the Thompson River country[161] than they do those of the coast. At least one celt of this general style has been found near Lake Chelan lying between the Thompson River region and both the Yakima and Nez Perce regions. It is a long stone celt and was found in an ancient grave on the bank of the Chelan River near the house of Hon. Amos Edmunds, of Chelan, Washington. In the graves of this group, according to Mr. C. G. Ridout, who cooperated with Mr. Edmunds in excavating at this place, and from whom all of our information on this specimen was obtained, stone knives and skinning and scraping tools were found. This celt is of a mottled green "marble resembling onyx" (probably serpentine or nephrite) 400 mm. long, 47 mm. wide and 15 mm. thick. It is slightly concave on the two sides, while one side edge is flat and the other is concavely bevelled. The poll is of the natural unworked stone and judging from the drawing furnished by Mr. Ridout, was broken off. It is raggedly diagonal. The cutting edge is sharpened by long convexly ground surfaces of nearly equal size and curve. The bevel of the side edge suggests that the material for the celt was cut out by grooving and breaking as was the case in the Thompson River region, where the celts showed similar traces of grooving.[162] It is true that similar grooving may be seen on celts from the Coast, but in that region the celts are short, while in the Thompson River area they are long like this one and the material is more often of the mottled green color than on the coast. The specimen is owned by Mr. Edmunds and is in the collection of Mr. Ridout. [160] Spinden, p. 182 and Figs. 1, 2, Plate IX. [161] Smith, (c), Fig. 349.