130 THE SUBANU. 
in Torres Strait, therefore at points of our earliest information of the 
Samoa Stream and the Viti Stream respectively. 
45. lua hole; Subanu luang id. 
lua Samoa, Futuna, Hawaii. 
luo Niué, Tonga. loaka Malagasy. 
rua Rapanui, Paumotu, Mangareva, | rua Malay. 
Tahiti, Maori, Mangaia. luwang Java. 
ua Marquesas. luang Subanu. 
lue Mota. 
From the Indonesian evidence (to which we must add the anomal- 
ous Malay lubang and the Bontoc Igorot kazpan) we are justified in 
regarding this as a stem closed in ng and the Malagasy is a normal 
mutation therefrom. 
46. ma conditional prefix; Subanu, Visayan ma id. 
This particle is general throughout the three Oceanic areas. It 
undergoes the normal vocalic mutations; it is paralleled by at least two 
similar particles (ta and pa) with differences in the consonantal modu- 
lant. JI am forced to postpone discussion of the variety in use in this 
matter to a later period of my researches. Ina general way it may be 
said that given a signification of an action or a state in a primal diffuse 
attributive, when the need arises for particularity the employment of 
ma prefaced to the attributive stem conveys the sense that a given 
object exists in the condition stated in the stem signification. Such 
forms are in essence adjectival in their employment. We may illus- 
trate this from the Samoan fola to spread out and mafola applied to an 
object which has been extended and therefore is spread out; we are 
forced to employ passive forms, but no such voice sense is yet within 
the power of these languages. 
47. masakit, makit sick; Visayan saquit id. P. W. 379. 
masaki Futuna. mai Rapanui, Tahiti, Hawaii. 
mahaki ‘Tonga, Uvea, Niué, Maori. 
madhake Viti. masaquit Ilocano. 
maki Marquesas, Rapanui, Mangaia, | sakit Malay, Visayan, Bontoc Igorot. 
Mangareva, Paumotu, Nu- | makit Silong. 
guria, Fotuna. maki Kisa. 
ma‘i Samoa. 
The strong concord in Indonesia leads me to postulate a final t. 
In the masakit forms this is quite clear and finds confirmation in the 
Melanesian King miseit. In the Malayan we have in one form for 
maki this final and in the other instance it does not appear; testimony 
toward the establishment upon Melanesian authority of the final t is 
derived from Baravonmait. The masakit type is a conditional; we find 
the primitive sakit in Malay, Visayan, and Bontoc Igorot. ‘This is the 
elder type; itis Proto-Samoan. For the Tongafiti makit we have been 
able to discover no instance of a primitive, but analogy leads us to the 
conclusion that this also is a conditional ma—akit. Interpolating a term, 
we may infer the descent from sakit through hakit to akit, then prefacing 
akit with the conditional ma the concurrent vowels might coalesce 
through crasis. It will be observed that now in dealing with the 
