208 ISLAXD CULTURE AREA OF AMERICA [eth. anx. 34 



brandies of a tree bent into a hoop and fastened at their ends. He 

 was the first to associate the stone collar with ' tree worship ' — 

 an important advance in the solution of the enigma. Mr. Joyce 

 described a stone collar in the British Museum in which there is no 

 shoulder ridge, but what appear to be the two ends of branches 

 'overlapped' and 'hooked together' at the point where the shoulder 

 ridge ordinarily is found. This led him to suggest that in studying 

 a stone collar we must ' retranslate ' it to its wooden prototype, and 

 recognize that the juncture of the ends in this case, and perhaps in 

 all, was effected as follows : ' When the limbs of the fork [of a tree] 

 were trimmed, the stump of a small subsidiary branch growing in a 

 convenient position toward the end of each, was left projecting; the 

 longer limb was bent round, and the projection toward its termina- 

 tion was hooked round the projection on the shorter limb; the ad- 

 dition of a cotton bandage would hide the joint and make all 

 secure ' (p. 410) . 



" ' It is perfectly obvious,' Mr. Joyce says, ' that these collars were 

 constructed originally of wood. A young tree was selected and cut 

 off immediately below a fork ; the two ends of the fork were trimmed 

 into unequal lengths, the longer bent round so as to overlap the 

 shorter, and the two fastened together by a band of cotton smilar to 

 the leg bandages worn by the natives.' He also states (p. 410) : ' Start- 

 ing with the supposition that they were originally constructed of 

 wood (which seems to me to be almost certain) , it seems possible that 

 a clew might be found in the prevalence of tree worship in the 

 Antilles.' 



" ' The heavy collars,' continues Mr. Joyce, ' which appear to have 

 been formed of a single and comparatively stout stem bent into a hoop 

 and the ends secured by a bandage may represent a zemi made origi- 

 nally from the straight trunk of a tree without a fork.' Thus a second 

 and important step in the interpretation of the meaning of the collar 

 was taken by Joyce in the recognition of the collar as a zemi con- 

 nected with tree worship, its original prototype being made of wood, 

 the stone form being a more lasting one, but one in which certain 

 characters of the wooden or archaic form still persisted. 



" In Eamon Pane's account of how Antillean wooden zemis were 

 made, as directed by a tree spirit, we have evidence of tree worship 

 in Haiti ; but the testimony afforded by this account is too meager to 

 prove that when the tree referred to by the Catalan father was felled 

 it was made into the form of a collar. The author suspects that the 

 idol referred to by Ramon Pane represented the Yuca Spirit,^' but 

 this suspicion is still subject to proof. 



*The Yuca spirit or Tocahu (Yocabana) Maorocon, Mahouya, Huracan, or Great 

 Serpent, whot^e idol Maroio was one of the two stone images in the Cave of the Sun, 

 worshipped for rain and blessings. 



