first, except that the upper third is broken off and missing. The cutting edge is somewhat curved and is fractured twice on each side; the bit is oval in section and its sides, which bevel suddenly from the shaft, bulge so slightly that they seem concave and apparently are somewhat reworked; and the pits on both sides are pronounced and very smooth. The specimen, which is covered with yellow clay, was found by the Oregon Alpine Club, and in November, 1900, was deposited in the museum of the Oregon Historical Society. (Cat. no. 140, List no. 26; Am. Mus. Nat. Hist., Neg. no. 12. No cast.)

The fourth hand-hammer-adze (d) differs from the first three in that it has no top knob, facets, or reworked surface. The cutting edge is curved, convex on one side and less so on the other, giving the implement a form similar to that of some of the celts of the Mississippi valley. The surface is polished very smooth but still shows some of the marks of pecking by means of which the object was fashioned. The specimen is 87 inches (226 mm.) long and of a yellowish brown color. It was found in Washington county, Oregon, and was presented in November, 1900, to the museum of the Oregon Historical Society, by Mr A. H. Garrison of Hillsboro. (Cat. no. 29 (10029), List no. 28; Am. Mus. Nat. Hist., Cast cat. no. 16/9854, Neg. no. 12.)

There is a similar hand-hammer-adze in the collection of Mr D. W. Owen, of Kennewick, Washington, which he says is from Umatilla, Oregon. This specimen, so far as is known at present, indicates the eastern limit of distribution of this form.

The first specimen of this kind that came to my notice is in the James Terry collection in the American Museum of Natural History. It is catalogued under no. T-22774 as a "chisel stone, plummet shaped, Columbia City, Columbia river, Oregon . . . collected by Dr C. G. Capler on October 4, 1882." In general it (g) resembles the first hand-hammer-adze described in this paper, but the object as a whole is of a slightly different shape, the neck being short, the lateral bulge of the body high up near the neck, and the bit long and slender; one facet merges into the flat surface of the side, while the other is rough, apparently having been made by pecking. The entire surface from the top to the side is curved

continuously, the neck being formed by grooving the side edges and carrying the groove around nearly to the middle of the sides, but leaving a small surface standing out like a ridge connecting the top with the side. The bit is oval in cross-section and the celtlike edge is convex. On each face of the bit are four grooves, two on each side. They extend from points between the side and the edge, near the middle of the object, to the bevel for the blade. The grooves on the left part of each side extend farther to the right at the blade, causing the object to suggest a spiral or screw. The specimen is 8 inches (213 mm.) long, made of heavy stone of a light bluishgray color; the surface is smooth in some places but shows marks of pecking in others.

Mr E. D. Zimmerman, of Philadelphia, informs me that in his private collection at Monterey, Pa., are six or seven hand-hammeradzes. Judging from a photograph of a portion of the collection, these are of the type here described; one of them has a hat-shaped top; two at least are of the long-bitted variety.

In a photograph of the H. C. Stevens collection, recently offered for sale, may be seen at least three hand-hammer-adzes. One of these has a simple knob at the top; another, a hat-shaped top, bulging body, and long bit; while a third specimen, which appears to be of the type above described, has a long bit. The top is grooved around twice (cf. fig. 23 e, Mem. Am. Mus. Nat. Hist., Iv) and on the side of the body shown in the photograph are two grooves which meet near the neck and then diverge, passing on each side of the spot where the facet is usually found, toward the edges of the side. No facet shows in the picture. The grooves just described give the object an appearance suggesting the lower side of a fish, the grooves indicating the gill slits.

Rev. Myron Eells probably refers to this type of object in his statement that "still another seems to have been a pestle at the handle end, and a blunt edge at the other." He also doubtless alludes to this type when, referring to chisels and wedges, he states:

"Dr Rafferty has nine whole ones, or parts, about which there is no doubt. They mostly come from Sauvies Island, and are generally of hard

1 Smithsonian Report for 1886, p. 286.

AM ANTH., N. S., 8-20

volcanic rock. They vary in weight from 2 pounds 14 ounces to 5 pounds 11 ounces; in length from 6 to 1334 inches, and width from 234 to 31⁄2 inches, and in thickness from 2% to 234 inches. The edges

are sharp, but the stone is thick a short distance from the edge. I know of none from other parts of Oregon."1

The fact that some of the grooves on the Terry specimen looked as if recently made, taken in connection with its peculiar shape, led the writer at first to regard it as a questionable specimen, or at least as a "sport" not at all characteristic of the region. The number of similar specimens from a restricted area which have since come to the author's notice, however, prove that they constitute a type characteristic of the archeology of Willamette valley and vicinity.

The facets suggest that these specimens have been used as hammers. The writer found similar objects only a short distance to the northwest of Portland, from Copalis head southward to Shoalwater bay, Washington, which are of the same type as those known to have been used by canoe-makers as hammers, that were secured in 1898 by Dr Livingston Farrand among the Indians at Quinault. However, all the specimens found from Quinault to Shoalwater bay, so far as the author is aware, have plain ends instead of celtlike ends and may be called hand-hammers. Probably these hand-hammer-adzes were used by canoe-makers as combination hammers and adzes, the blows being delivered in such a way as to form the facets.

Two specimens (pl. xx111, e-e', f-f'), which may be designated hand-adzes, have celtlike ends and tops resembling pestles.

The first hand-adze (e, e') resembles in its upper portion a pestle, with a circular body, somewhat larger at the base than at the top, a disk-shaped striking-head, and a convex top. The surface of this portion of the object shows very slight scars or flutings, reminding one of the surface of a whittled stick or of a pared vegetable. The shaft expands suddenly into the disk-shaped strikinghead, which in turn coalesces into a celtlike form projecting from the base of the upper portion. The line of demarkation between the upper (cylindrical) and the lower (celtlike) portions of the specimen is obscure except along part of one edge (e'). The celtlike

part is somewhat convex on one side (shown in e), concave on the other, especially at its base, but elliptical in cross-section; it tapers gradually from its large base toward what was once the cutting edge, but which is broken off. There are many signs of fluting on the convex face. The whole object is 67 inches (174 mm.) long. It is hard and heavy and appears to be basalt, although the surface, except where broken, is much weathered and resembles yellowishgray chalk.

This hand-adze was found in the garden of Mr E. D. Nelson, Portland, Oregon, and was presented by him on February 5, 1903, to the Museum of the Oregon Historical Society. (Cat. no. 382 (380), List no. 35; Am. Mus. Nat. Hist., Cast cat. no. 16/9860, Neg. no. 1 (edge) and 3 (side).)

The second hand-adze (ƒf, f') resembles the first, but the top and the convex exterior of the bit present the natural surface of a water-worn pebble, while the remainder of the surface shows marks of pecking, by which process the object was fashioned from the pebble. In some places these marks are partially obliterated by grinding and polishing. There are no flutings on the surface. The disk shape of the striking-head shows plainly for fully half the circumference of the specimen, but the convex side of the celtlike part extends nearly half-way up the shaft of the pestle-like part. It is set, as it were, about half-way its length on the side of the lower half of the pestle-like part. The bit is lenticular in crosssection and oval in outline. The cutting edge is semicircular, sharp and beveled to an edge, chiefly from the concave side. The whole object is 74 inches (184 mm.) long and is made of heavy grayish or milky blue mottled stone, possibly slate.

This specimen was found on Columbia slough about ten miles below Portland and was deposited in the Museum of the Oregon Historical Society on Nov. 30, 1902. (Cat. no. 383, List no. 36; Am. Mus. of Nat. Hist., Cast cat. no. 16/9861, Neg. no. 1 (edge) and 3 (side).)

There is a specimen of this type (cat. no. 25) in the collection of Mr Louis O. Janeck, North Yakima, Washington. The natural surface of the pebble from which the implement was made shows on the ridge, or the part which corresponds to the sides of the

striking-head of the pestle-like section. The specimen is 61⁄2 inches (165 mm.) long and made of rock resembling diorite or diabase. (Am. Mus. Nat. Hist., Neg. cat. no. 44452 (2-1), 44503 (6–4).) It is perhaps the most nearly perfect form of this type ever seen by the writer. The upper end corresponds closely in form to a pestle, with a slight indication of a knob at the top, a flaring body, and a short striking-head the periphery of which extends as a ridge nearly, if not quite, around the specimen. The celtlike part is toward one edge, so that one side expands to meet the ridge above mentioned, forming a concavity; the other contracts to meet it, forming a convex sweep from the cutting edge to the beginning of the body of the pestle-like part. The specimen was found near the surface in an old burial ground of the Indians near the mouth of Yakima river on what is known as McNeals island. This specimen marks the present known eastern limit of the occurrence of the form. Mr Zimmerman has informed me that there are five or six specimens of this type in his collection.

The region north of Portland has yielded a pestle,' shaped like the upper part of the present specimens, which was used as a hammer, the blow being delivered with the end instead of with the side as in the previous case. From the same area come stone celts hafted in handles made of antler (see fig. 29 d, p. 164, Mem. Am. Mus. Nat. Hist., IV). From the region south of Portland are such celts, hafted by being lashed to stone handles (for specimens of such handles see Oreg. Hist. Soc. cat. no. 381, list 34, and the collections of the American Museum of Natural History, the United States National Museum, and the Peabody Museum of Harvard). It seems to the writer, therefore, that this type (the hand-adze) is a combination of the pestle-shaped hammer of the north and the stone celt-handle of the south with the celt of both regions, and that it resulted from a modification and combination of the same ideas that produced these neighboring forms with which it may be compared.

It is also interesting to compare this form with one from the gravel at Oregon bar, California, shown in plates II and VIII of the paper by Professor William H. Holmes on Auriferous Gravel

See American Anthropologist, N. S., I, fig. 10, p. 364, 1899.