NO. 14 ARCHEOLOGY OF BAY ISLANDS, HONDURAS STRONG 61 



suggesting the use of wire in the building process, although actual 

 wire does not seem to have been used. The loop is round and solid. 

 There is one smaller bell of similar type (pi. 9, h). The most com- 

 mon type has a more marked pseudo-wire technique around the neck 

 and has a round wire loop for suspension as shown in the illustration 

 (pi. 9). A few specimens lack this neck (pi. 9, b), and some pass 

 directly from the pear-shaped body to the wire loop (pi. 9, g). Seven 

 small bells have animal faces in good relief (pi. 10, a, b, d, f), one 

 has a complex wire scroll in quadruplicate at the top (pi. 10, e) 

 and one bell, the smallest, is globular (pi. 10, c). A considerable 

 number of all types have been broken and bent up to retain the 

 clapper. This, in conjunction with the broken and worn appearance 

 of the majority, suggests that they had been used for a long tmne 

 prior to their deposition as an offering. As to their original source, 

 all that can be said at present is that they closely resemble certain 

 of the copper bells found in a large cave deposit on the Uloa River, 

 which were described by Blackeston (1910 b). Presumably, the 

 bells had been cast in molds, but I do not pretend to understand the 

 exact techniques employed. Aside from the bells the two rings are 

 the only definite copper artifacts. The larger of these (pi. 9, c) 

 is 4 cm in diameter and has two holes for fastening in a central 

 disk, which has disappeared. The small ring (pi. 9, /) has only 

 a slightly concave edge and lacks perforations. It is not welded 

 together as in the first case, and the two edges overlap. Its use is 

 uncertain, but it may be a broken or unfinished ear spool. The flat 

 disk of copper (pi. 9, ^) is hammered all over its surfaces, and the 

 edges are smooth. Whether it is an unfinished piece or once served 

 as a pendant is uncertain. 



Through the courtesy of the Bureau of Standards, a spectro- 

 chemical analysis was made of (a) the copper disk (pi. 9, e), (b) 

 a bell with a definite neck and a face in relief (like pi. 10, /) and 

 (c) a pear-shaped bell (like pi. 9, g). For these data I am grateful 

 to Dr. Lyman J. Briggs, Director of the Bureau of Standards, and to 

 B. F. Scribner, who made the analyses. Aside from their future value 

 in any comprehensive study of aboriginal copper bells, the analyses 

 indicate that (a) and (b) are similar in composition, but that 

 (c) differs rather markedly. It seems probable, therefore, that the 

 ultimate source of the pear-shaped bells (pi. 9, g) was different from 

 that of the majority, which have definite necks and in some cases 

 decoration in relief. 



