DIXON— SWAN-MAIDEN THEME 



stances farther eastward. The first of these lies in the Banks and New 

 Hebrides groups in eastern Melanesia. In these tales 1 the hero steals 

 the wings of a sky-maiden, who, with her companions, flies down to 

 bathe. He hides the wings and thus secures the maiden for his wife. 

 She produces food magically, and for this is scolded by her mother- 

 in-law or brother-in-law. As she sits weeping, her tears wash away 

 the earth which covers her wings where they were hidden. Rejoiced 

 at the discovery which gives her once more her freedom, she puts 

 the wings on, and flies away. The husband then ascends to the upper 

 world by means of the well-known arrow-chain, and as he is descend- 

 ing again, thinking that she is following him, she cuts the chain above 

 him, and he is killed by the fall. It is to be observed that in one 

 version the sky-maidens are called Vinmara, a term so similar to 

 Widadari as to be very suggestive. If these New Hebridean versions 

 had been encountered anywhere in the Indonesian area, one would 

 be justified in regarding them as probably historically related to the 

 other Indonesian versions, and as, with these, showing affiliation with 

 Indian prototypes. Isolated as they are, however, the problem is not 

 so simple. So far as I know, no versions of the tale have been recorded 

 nearer than Nufoor; but it must be remembered that we have but the 

 scantiest materials in the way of mythology from the Solomon islands, 

 where, on a priori grounds, one might expect to find a linking form. 

 On the other hand, inasmuch as there is considerable material now 

 published from the Bismarck archipelago, the Admiralty islands, and 

 parts of eastern New Guinea, it would seem that as no trace of the 

 story has been found there, it can hardly be regarded as character- 

 istic of the more purely Papuan or aboriginal population of the 

 Melanesian area. The people of the New Hebrides are admittedly of 

 complex origin, an older aboriginal stratum being overlaid by one or 

 more immigrant strata, derived ultimately from Indonesia. This being 

 so, the possibility of the tale under discussion being one of the many 

 culture elements carried by the migrating ancestors from Indonesia 

 to Melanesia, must be admitted. If we were dealing only with a few 

 scattering cases of similarities in myths, the explanation of trans- 

 ference or transmission could hardly be regarded seriously. But as 

 the number of cultural elements of all sorts, apparently thus derived, 

 is large and varied, the conclusion is greatly strengthened. 



The second isolated area in which the tale occurs is in New South 

 Wales and Victoria. The first version, 2 from the Euahlayi tribe, runs 

 as follo ws: 



1 Codrington, The Melanesians, pp. 172, 397 sq.; Suas in Anthropos, vol. VII, pp. 54 sq.; Mac- 

 donald in Australasian Asso. Adv. Science, vol. IV, pp. 731 sq. 

 * Parker, Australian Legendary Tales, pp. 44 sq. 



[85] 



