Page:Folklore1919.djvu/669

Rh ghost-society which is certainly a close relative of the Dukduk, wear head-coverings which are sometimes cylindrical in form. There is reason to believe that the hats of the Banks Islands are derived from masks which are often more than head-coverings, but both the Tamate societies and the Rukruk show how important a place is taken in the beliefs of those who wear them by that part of the mask which covers the head, a belief which follows naturally from the vast importance attached to the head throughout Oceania.

If I am right in my general position that the statues of Easter Island are only hypertrophied examples of the images of the people who introduced the art of stone-working into the Pacific, we are provided with a fully sufficient motive for the presence of crowns or hats on the statues of Easter Island. Moreover, there is the clearest evidence that the hats of Melanesia are associated with a cult of the dead, and it therefore becomes natural that these crowns should only occur on the statues of the burial-places of Easter Island and should not be present on the statues which are shown by their locality to have had a different purpose. Moreover, the head-coverings of the Rukruk of Bougainville and the tamate of the Banks Islands provide parallels for the cylindrical form of the crowns of Easter Island. It is not necessary to assume, as does Mr. Balfour, that this shape was adopted in order that the crowns might be rolled from their place of manufacture to the images upon which they were to be placed. Their form is one proper to their function as head-coverings of images representing the dead.

If the sculptors of Easter Island put hats on their statues because these were the proper attribute of representations of the dead, we have an all-sufficient motive for their choice of a material different from that of which the other parts