THE material on the mythology of Melanesia, though incomplete and fragmentary, appears rather clearly to prove the existence of two distinct strata, one of which may be called Papuan, the other Melanesian. The former is best represented among the Kai tribes of the region north of Huon Gulf in German New Guinea, as well as by the Baining and Sulka of northern New Britain, and may be traced, more or less plainly, among the remaining coastal tribes of both German and British New Guinea; whereas it is much less apparent in the Banks Islands, the New Hebrides, and Fiji. The Melanesian stratum, on the other hand, is perhaps best developed in eastern Melanesia, i. e. Santa Cruz, the Banks Islands, the New Hebrides, and Fiji; though it is well represented throughout the New Guinea littoral districts, among the coast tribes of northern New Britain and in the Admiralty Islands. What has been called the Papuan type of mythology seems to be characterized. by a relative absence of cosmogonic myths, by the prominence of ghosts, and by a general simplicity and naïveté; and this category also appears to show an extensive development of tales of local distribution only, corresponding to the discreteness and lack of relationship on the linguistic side. The Melanesian stratum, on the other hand, exhibits a considerably greater evolution on the side of cosmogony, an especial fondness for cannibalistic tales, and a rudimentary dualistic character which is revealed in the many stories of the wise and foolish culture hero brothers. Further examination of this Melanesian type seems to indicate that
it is by no means a unit, although, because of the character of the material, any conclusions must be wholly tentative. The following grouping is suggested: (1) myths of general distribution throughout Melanesia; (2) those confined more or less strictly to New Guinea and the immediate vicinity; and (3) those similarly restricted in their distribution to Fiji, the New Hebrides, and the Banks and Santa Cruz Islands.
If now, instead of limiting our view to Melanesia alone, we include the whole of the Oceanic area and endeavour to discover the relationship of Melanesian mythology to that of the adjacent sections, it appears that, whereas of the two main types (the Papuan and Melanesian) the former shows little in common with any of the other Oceanic regions, the latter, on the contrary, exhibits numerous and interesting relationships with Indonesia, Micronesia, and Polynesia, and some even with Australia. The Melanesian type of incidents which reveal similarities with these other areas may be divided into four groups: (1) those whose resemblances are only with Indonesia; (2) only with Polynesia; (3) with both Indonesia and Polynesia; and (4) with Micronesia. The first of these groups is represented much more strongly in New Guinea than in the eastern archipelago; and in New Guinea it is far more prominent on the northern coast than on the southern. It would seem to manifest influences from Indonesia which, in the course of migrations eastward, did not extend beyond Melanesia, and which were greater in New Guinea and its vicinity than in the eastern and more distant archipelagos. The second group--rather unexpectedly--is, like the first, more prominent in New Guinea than farther east, but is better represented on the south coast than is the first group. From the character of the incidents and their distribution in Melanesia and Polynesia this group itself would appear to comprise (a) incidents preponderantly Melanesian, borrowed by the Polynesian ancestors and carried with them into Polynesia, and (b) incidents of Polynesian development which have been
transmitted westward as a result of the probable late reflex of Polynesian peoples into parts of eastern Melanesia.
The third group, comprising myth-incidents from Indonesia, Melanesia, and Polynesia, is contrasted with both the others in that it is best represented in eastern Melanesia. Theoretically, these incidents may be regarded as a portion of those brought by the Polynesian ancestors from their Indonesian homes and still preserved by them in Polynesia. Their presence in Melanesia would thus be hypothetically due to their having been taken over from the migrant Polynesians, and their greater prominence in the eastern archipelago would be expected, as it was presumably in this area, rather than in New Guinea, that, during their migration, the Polynesian ancestors made their longest halt and exerted their greatest influence on the aboriginal population. The last group, which is composed of those incidents common to Melanesia and Micronesia, is about equally represented in New Guinea and the eastern archipelago. The relatively large number of similarities between Micronesia and Melanesia is only what we should expect, owing to the many evidences derived from other sources, of relationship between the peoples of the two areas; but the amount of agreement with eastern Melanesia is rather striking.