144 SISSANO. 
( 
Tongafiti kava, since that migration had had no communication with 
Melanesia. Wehave not been able to lay out the early track of Tongafiti 
migration, except in this particular of Melanesian exclusion. I have 
suggested that it is possible that such careful examination of the speech 
material of Micronesia as we have been able to expend upon the 
speech material of Melanesia may reveal the traces of Tongafiti 
entrance into the Pacific by an equatorial or northern traverse. It 
is at this point proper to note that in Micronesia we have the kava 
custom more or less widely extended, and in this case intercommunica- 
tion with Melanesia is expressly contraindicated. 
We have botanical evidence of the existence of the physical basis of 
the kava custom in Indonesia; the Piper methysticum is reported from 
several islands, although it is not now in use. ‘The conditions of the 
problem are most satisfactorily met by assuming the origin of the kava 
custom among the Proto-Polynesians during their residence in Indo- 
nesia anterior to the Christian era, and further assuming its expulsion 
therefrom in the company of its addicts by the sirih-chewing Malayan 
invaders. From this source we can see the Proto-Samoan migrants 
carrying the kava custom through Melanesia (this without prejudice 
to a possible independent development of a kava usage in Melanesia, 
which would be neutral so far as regards our greater problem), and 
we can suggest for later determination the explanation of Tongafiti 
migrants carrying the kava custom into such subdivisions of Micro- 
nesia aS we know to practice it. We have already bestowed consider- 
able attention upon the advance of sirih-chewing upon peoples addicted 
to the kava custom. By combining these two elements, the older 
and the more modern customs, we shall find an explanation of the 
present distribution of sirih and kava which seems to entail no serious 
difficulty at any point. 
gi. suk nose. 
39. iru Bahasa, Ambon, Kolon. 54. ison East Ceram. 
40. hiru Bahasa, Ambon. 55. lisu Bahasa. 
41. irun Minahassa. 56. iri Bahasa, Ambon. 
42. irung Java. 57. ili Bahasa, Ambon. 
43. idung = Java. 58. ileng Bontoc Igorot. 
44. inu North Ceram. 59. nien Buru. 
45. niru North Ceram, Allor, Mina- | 60. nieni Buru. 
hassa. 61. ninin East Ceram. 
46. nirun-- Kei. 62. uruna Malagasy. 
47. ngirung Minahassa. 63. urong  Dyak. 
48. ngilung Minahassa. 64. kamuru Macassar. 
49. ngijung Celebes Alfuro. 65. ngunu  MHalmaheira. 
50. ilu Bima. 66. nunu Ternate. 
51. nilunu’ Kei, Banda. 67. usnut Gani. 
52. ninu Bahasa. 68. soong Subanu. 
53. ninura Ambon. 69. nge Buru. 
The Indonesian material shows the same interlacing of isu and 
usu forms as in Melanesia and at the edges of Polynesia in Viti and 
Kapingamarangi, the isu type being represented by the forms 
39-55, the usu by 62-68. In this speech area we find a significant 
group of transition, or at least intermediate, forms in which the vowel 
