ETHNOLOGY. 373 



pottery. Precisely similar fragments are in the National Museum from 

 San Domingo, and, indeed, many of the pieces from southern Central 

 America closely resemble them in quality and material. 



CELTS. 



The celts, one hundred and thirty-five in number, are of the very 

 highest order of workmanship, being beautifully shaped, and many of 

 them the most highly polished stoue implements in the National Mu- 

 seum. The material is fine grained, and varies in color from black to 

 nearly white, many of them being of a jadeite green. In shape, nearly 

 all of them belong to Evans's third class, or oval-sectioned, and the 

 great majority resemble his figure 75 so closely that I feel sure he is 

 right in hesitating to believe the celt figured in his work to have been 

 made in Scotland. (Fig. 10.)— (Evaus, Stone Imp. p. 118. The use 

 to which these polished celts was put, or, more correctly speaking, 

 the manner of haftiug them, is graphically illustrated in the accom- 

 panying sketch of a celt inserted in a mortise in a handle of hard red 

 wood and found in a cave in Caicos or Turk's Islaud, by Mr. George J. 

 Gibbs, and kindly lent by him to be cast and engraved. (Fig. 11.) 

 A still more interesting and precious relic, from the same locality, and 

 found by the same gentleman, is that given in figure 12, which repre- 

 sents a celt in the handle, the whole being gracefully carved out of a 

 single piece of jadeite. 



A beautiful ax, similarly carved from a single piece, is figured and 

 described in Jones's Aboriginal Remains of Tennessee. — (Smithsonian 

 Contributions, No. 259.) 



Those interested in comparative archaeology will take great pleasure 

 in compariug these with figures 91, 92, and 93, of Evans's Ancient Stone 

 Implements. This mode of hafting suggests that these oval-sectioned 

 celts, set in their handles with the edge in a line with the haft, were 

 rather battle-axes than industrial tools, although this is mere conjecture. 



The celts iuthe Latimer collection vary in section from circular to oblong 

 elliptical, in length from 1.75 to 12 inches, in width from .75 to G.5 

 inches. The chord of the edge is often oblique to the axis of the stone. 

 Some have almost semicircular edges; of others the edge is nearly a 

 straight Hue. A few are so unique as to deserve especial mention. The 

 figures iu the margin represent, throughout this paper, the number of 

 the specimen in the ethnological collection of the National Museum. 



1G89S. A large, mottled, greenish, flat celt, pointed at the butt, frac- 

 tured. Length 11.2, width G.5, thickness 1.95 inches. Mr. Gibbs also 

 sends drawings of two large flat celts, similar to this one and the three 

 following, trom Turk's and Caicos Islands. The occurrence of these 

 large polished celts over so wide an area, corresponding in fact to that of 

 the Caribs in Columbus's day, coupled with the frequent allusions of 

 Herrera, Peter Martyr, and others to dug-out canoes, shaped like trays, 

 and capable of holding from one to one hundred and fifty persons, leads 



