134 THE SUBANU. 
Malayan section I have aimed to order the material in the progress of 
variation, principally with respect of the final syllable or consonant. 
The agreement is so overwhelming in favor of mata that we need have 
no hesitation in postulating that open stem as primal. ‘The residual 
forms, each concurring in but a single speech or at most in but two or 
three, will fall into the two classes of suffixed composition members and 
closing consonants added in conformity with the régime of the several 
dialects. 
55- mate to die; Subanu matay id. P. W. 373. 
mate Samoa, Tonga, Fakaofo, Futuna, { make Hawaii. 
Niué, Uvea, Maori, Tahiti, 
Rarotonga, Rapanui, Mar- | mate Macassar. 
quesas, Mangareva, Pau- | mati Malay. 
motu, Manahiki, Bukabuka, | maty Malagasy. 
Tongarewa, Nukuoro. matay Magindano, Subanu, Visayan. 
u-mate Nuguria. patay Visayan, Java. 
ko—mate Aniwa. maki Kisa. 
kono—mate Fotuna. matei Kayan. 
Strange to say, this word in all its recorded occurrences must be a 
conditional, for that is the only basis upon which we can comprehend 
the form patay which exists in Java and is found in Visayan simultane- 
ous with matay. We should then postulate a primal at, recognizing 
the considerable predominance in Indonesia of forms in 1; this might 
then serve to account for the of7 of Samoan. In that language mate 
and ot are synonyma relative to the fact, but, relative to the subject, 
mate is employed of the beast, oft of the man. In my former notes upon 
these two words (The Polynesian Languages, pages 274, 374) I was 
forced, in the lack of this fuller information, to assign to the courtesy 
speech the use of of as death. It will be far simpler to consider it as 
probably a primal, and the discovery of ate or ati in Indonesia will serve 
to establish this view beyond peradventure. 
Since the writing of this note my collation of the Bontoc Igorot 
in Seidenadel’s vocabulary has disclosed in the noun signifying death 
the primitive z7déy and variants itiy, édoy, doy. ‘Thus, having reached 
a hypothetical primitive by deduction, it is interesting to find that 
access to additional data brings confirmation. 
56. mati-kuku nail, claw; Subanu kanuku id. 
kuku Viti. mai—uu Tahiti, Marquesas, Hawaii. 
mati-kuku Mangareva, Futuna, Maori. 
mai-kuku Maori, Paumotu, Rapanui, Mar- | kuku Malay, Savu, Pampangas. 
quesas. cuco Tagalog. 
beji-kuku Tonga. coco Visayan. 
pasi-kuku Uvea. ké6ko Bontoc Igorot. 
mati—‘u‘u Samoa. 
In Polynesia we encounter the primal stem only in Viti; elsewhere 
it is involved with a formative agent, principally mat: or mai. ‘This 
occurrence of kuku in Viti should serve to set aside Hazlewood’s note 
that it derives from kuku the name of a small shell; this shell name 
extends beyond Viti into Polynesia in its own independent existence. 
The primal is well established in Indonesia, for the vowel mutation is 
there negligible, particularly the interchange of o and u. 
