HOLMES. I 



KNIVES — CELTS. 203 



and tbeir Axes sharp Stones bound to the end of a Stick, and glued in 

 with Turpentine. By the help of these they made their Bows of the 

 Locust Tree."' 



Drake, in his "World Encompassed," speaking of some of the south- 

 efn tribes of South America, probably the Patagoniaus, says that " their 

 hatchetts and knives are umde of mussel-shells, being great and a foot 

 in length, the brickie part whereof being broken ofi', they grind them 

 by great labor to a line edge and very sharpe, and as it seemeth, very 

 durable.^ # * » Tiieir working tools, which they use in cutting 

 these things and such other, are knives made of most huge and mon- 

 strous mussell shels (the like whereof have not been seen or heard of 

 lightly by any travelers, the meate thereof being very savourie and good 

 in eating), which, after they have broken off the thinne and brittle 

 substance of the edge, they rub and grind them upon stones had for the 

 puri^ose, till they have tempered and set such au edge upon them, that 

 no wood is so hard but they will cut it at pleasure with the same.'" 



According to Sproat, shell knives were used by the Indians of Van- 

 couver's Island in carving the curious wooden images placed over 

 graves.'' 



Ancient shell knives are very rarely found iu collections. Such spec- 

 imens as have come to my notice could as well be classed as scrapers 

 or celts. We will probably not be far wrong in concluding that such 

 implements were used for scraping and digging as well as for cutting. 

 As a rule, knives proper were simply sharjjened bivalve shells. The 

 scrapers so frequently mentioned were doubtless often the same, but 

 probably more frequently portions of the lower whorl of the large uni- 

 valves. 



CELTS. 



Implements of this class are generally made from the lower part of 

 large univalves. They were probably used in a variety of ways, with 

 handles and without. The spine-like base of the shell forms the shaft, 

 the blade being cut from the broadly expanded wall of the lower whorl. 

 Nearly all the specimens in the national collection have been obtained 

 in this way. In Plate XXV three very fine examples are figured. The 

 specimen illustrated in Fig. 1 is more than usually well fashioned, and 

 is extremely massive, having the proportions and almost the weight of 

 typical stone celts. It is five inches in length, two and three-fourths iu 

 width, aud nearly one inch through at the thickest part. 



'Beverly: History of Virginia, 1722, p. 197. 



''Drake, iu Hakluyt Society Publications, vol. XVI, p. 74. 



3 Hid, p. 78. 



* Sproat's Savage Life, p. 86. 



