230 THE POLYNESIAN WANDERINGS. 



kit. This may be the basic ki with a persisting radical final consonant, 

 or the / maybe an irregular accretion ; it is impossible in the paucity of data 

 to come to a conclusion. If Pak, Leon tiktik be kit under metathesis this 

 is the simplest form. In Norbarbar and Retan, with floating vowel, we 

 have this stem with a prefix sc, so, common to both. In Gog I see such 

 resemblance to Retan that I diagram its composition as we-s-kit. 



From this stem, by a common frontal abrasion, we may perhaps derive 

 the common word of Polynesia for small; Samoa, Tahiti, Marquesas, 

 Rapanui, Mangareva iti; Tonga jii (metathetic), Moiki itiiti by and by, 

 and Hawaii iki (a kappation of radical t and not to be confounded with 

 the stem iki). This would restore to Polynesia the possession of the 

 third Efate" stem, and it now seems to be proved by Rarotonga ngiti. 



Omba mbiti we lack data to coordinate. The first impression that this 

 is iti with a prefixed modulant will hardly stand against the fact that */* 

 is an eastern form nowhere found in the west, and the other fact that we 

 know of no well attested instance in which a stem which has had its head 

 rubbed off acquires a new one by random selection. 



We now are left with Vuras and Aneityum, which have the common 

 element net or tin according as we consider the metathesis to have been 

 applied to one or to the other. If we regard net as primal and as equivalent 

 to nit we shall find any attempt to associate it with iti by frontal accretion 

 blocked by the same bar as at Omba, and this somewhat considerable 

 array of data affords no example of a k-n mutation by which we might 

 account for it. 



Uvea sii, with Aniwa aiihi, may open for us a small group of Melanesian 

 relationships in Aneityum sisi, Efate ses, sos; Epi takisi, Lakon sik, Volow 

 siwi, Motlav su, and thus connect with the se, so, element of the compound 

 forms of Norbarbar, Retan and Gog. It should be observed, however, 

 that Uvea sii devolves normally through Tonga jii from iti. 



Dr. Codrington (Melanesian Languages, 81) comments "as is the case 

 with most adjectives there is but scanty agreement in the words meaning 

 little." I think that in working over his material quite another conclusion 

 has been reached and that the Polynesian content is distinctly marked 

 from end to end of Melanesia. 



Indonesian affiliations are rare. We find Wahai kiiti, Bouton kidikidi, 

 Salayer kedi, and beyond these three we may scarcely venture, perhaps 

 not even so far. 



In none of the foregoing is there aught which points to the Semitic 

 identification. 



150. 



kiat, the sticks which cross from the canoe to the outrigger. 



Samoa: 'iato, the outrigger struts. Tonga, Nuguria: kiato, id. 

 Futuna: kiato, id. Fotuna: akiato, id. Maori: kiato, the 

 thwart of a canoe. Tahiti : iato, the transverse beams which 

 connect the outrigger to the canoe. Hawaii : iako, the arched 

 sticks which connect a canoe with its outrigger. Mangareva : 

 kiato, a large raft. Mangaia: kiato, the outrigger. Paumotu: 

 kiato, to pierce and cross for joining. 



