108 THE SUBANU. 
attributive in adjective sense descriptive of some appearance of the 
claws of this crab; the only meaning which might seem applicable is 
derivable from Mangarevan ngongo ‘‘a conical hole,’’ with which we 
may associate Hawaiian no ‘‘a hole.”’ ‘The Samoan alimango is the 
Portunid Lupea; we are by no means sure to what extent the dotted and 
pockmarked appearance of this crab would warrant the designation 
“pitted claw’’ when the marking is carried over the whole carapace, 
but we are no more than at the beginning of our understanding of 
selectivity of definitive characters as authorizing name-creation by these 
beginners of speech. 
7. apunga-—leveleve spider; Visayan laoalaoa id. P. W. 361. 
apunga—leveleve Sainoa. punga—werewere Maori. 
ka—leveleve Tonga, Futuna, Niué. pua—verevere Tahiti. 
hala~neveneve Nukuoro. puna—welewele Hawaii. 
buta-lawalawa Viti. puna-—vevee Marquesas. 
ka—velevele Uvea. 
punga—verevere Mangareva, Paumotu, lawalawa Malay. 
Mangaia. kaaowa Bontoc Igorot. 
The primal sense appears to be that of the web, but the passage to 
the webster is not difficult, therefore we find the word indifferently 
applied to the spider. Our three Indonesian terminals being found in 
agreement upon the form which characterizes Nuclear Polynesia, we 
may argue that the concordant inversion which marks the Tongafiti 
use 1s of later development than the exit of the Proto-Samoans from 
the Malayan archipelago. 
8. alelo tongue; Subanu dila id. 
alelo Samoa, Futuna, Niué, Fakaofo, | rera Bima. 
Manahiki, Hawaii. rilah Ratahan. 
arero Maori, Tahiti, Paumotu, Mangaia, | lila Sanguir, Bugis. 
Rapanui. ela Malagasy. 
warero Moriori. lilah Bouton, Salayer, Menado. 
aledo Sikaiana. ledah Malay. 
alel Rotuma. lidah Kayan, Basakrama. 
aeo Marquesas. ilat Java. 
elelo Tonga, Hawaii. dila Bolanghitam, Sulu, Tagalog, Ilo- 
erero Mangareva. cano, Pampangas, Visayan. 
lelo Hawaii. delah Baju. 
eo Marquesas. djila Bontoc Igorot. 
In the Polynesian we have no difficulty in picking out the stem /elo, 
nude in Hawaiian and Marquesan, elsewhere prefaced by the formative 
a, concerning which I have already made sufficient note in item 6. 
The presence of the simple stem in Hawaii and the Marquesan is not of 
critical value, inasmuch as each has the augmented stem as well. In 
general we note that this augment has been acquired since contact with 
the Indonesians ceased. In the Indonesian languages the final vowel 
has passed from o toa, a mutation of no moment in the vowel uncer- 
tainty of that area. We find, then, the first five items sufficiently 
representative of the Jelo stem. The remaining forms fall into two 
groups according as the initial or the medial liquid undergoes mutation 
