1896.] 0.)4 [Cushing. 



istic of the environment we found them in, thej' are the product of a 

 very slow growth. Certainly, while this art of the keys may not have 

 been, nay, was not, altogether of a strictly local origin, it was in the 

 main, of a kind whicli one might expect to find developing or developed 

 in such an environment. Everywhere, for example, evidence of the 

 influence of shell, shark-tooth, and other sea-produced materials — used 

 as implements in the working of wood, bone and horn, and of shell 

 itself — could, as I have shown, be traced here ; and had jilainly, as I 

 have also shown, given rise to special ornaments on particular parts of 

 things thus made. But the point of interest is, that these ornaments were 

 not only conventional, but that they had already become conventionally 

 speciiiUzed ; were, many of them, indeed, so highly conventionalized and 

 thus so specialized, that except for the completeness of our series, they 

 could not have been traced to their simple, incidental origin in the kinds 

 of tools used, modes of working employed, and materials worked. I 

 have said that this kind of conventionalization in art and localization 

 of decorations, is of exceedingly slow growth. This is because genera- 

 tions, if not ages, are required for the radical modification of a single 

 specialized ornament on any particular part of a specialized tool or 

 implement, weapon or ceremonial appliance, among primitive peoples ; 

 owing to such peculiar conceptions of the meaning and potency of form 

 as I have already discussed in its relation to ceremonial objects, and will 

 presently again refer to as particularly relating to things practically used. 

 By way of a single example, I may instance the circular obvolute, or 

 navel ornament (as I have called it), in its relation to the ends of the hard- 

 wood handles of certain classes of tools in the collection. I have referred 

 to this as having been derived directly from the double spiral or obvo- 

 lute observable on the cut-otf apices or ends of conch- or busycon-shells 

 and other univalvular shells. I have also suggested that the use of 

 kingfish jaws and shark-tooth knives in girdling sticks, bj'^ a process of 

 cutting around and around the sticks always in the same direction, with- 

 these sharp, yet jagged tools, produced, as shown by many specimens in 

 the collection, rough, spiral rosettes at the ends of the sticks. Now 

 when the sticks were severed in the same way, but first from one side 

 then the other, the figures produced at the ends of them strikingly 

 resembled the involuted spirals at the ends of the worked shells. Thus, 

 although the figure when associated with purely ceremonial objects 

 doubtless signified the "navel" or "middle "—as earlier suggested — yet 

 it came to be associated also with the ends of the handles of tools the 

 working parts of which were made of the columelUe of shells on the 

 ends of which it naturally occurred. Thus, for mythic reasons, the figure 

 was doubtless considered not only appropriate, but even essential to the 

 handle, no less than to the shell armature of such a tool, in order to 

 harmonize its parts, to give potency or etfectiveness to it as a whole. 

 So too, with the radiate or rosette figures found on the ends of very 

 small liandles made from saplings. It was observed that when suitable 



