MELANESIAN ANNOTATIONS ON THE VOCABULARY. 91 
came equipped for their studies with a practical knowledge of the 
speech of Tahiti, from which the k had then completely dropped 
out. But when they listened to a word from Samoan lips with which 
they were familiar in Tahiti they noted a difference in this particular 
sufficiently great to call for note in recording the new speech. We 
find an illustration in this breadfruit word, in Tahiti uru with a com- 
plete abolition of the k, but in Samoan they found themselves under 
some phonetic necessity to particularize attention upon the dropping 
of the initial and to write the word ‘ulu. The inverted comma is 
used consistently in the written Samoan to record the absence of that 
palatal mute. 
The phonetic problem is for the most part disregarded by those 
foreigners who assume to speak some sort of Samoan, yet it is essential 
to the proper use of the speech and involves no insurmountable dif- 
ficulty. The vowel introduced by inverted comma, to which has 
been assigned the name “‘break,”’ is pronounced as of its proper quality 
but from the palatal mute position. Effectively the speech organs 
are set in the position for the production of the k and immediately 
without emitting the k sound, pass to the voicing of the vowel. That 
this is of importance is illustrated by the fact that between ulu the 
head and ‘ulu the breadfruit the only means of distinction is the 
break. Its presence in Samoan is proof that the dropping of the k is 
most recent. While upon this theme we note the still more modern 
swing back to the resumption of the palatal mute in pronunciation, 
but now addressed upon the lingual mute with the result that in 
written Hawaiian and in spoken Samoan the t has been completely 
replaced by k, a phonetic development fitly described as the process 
of kappation. 
The dropping of k from a primal kulu is thus easily accounted for. 
It would be more difficult to predicate a primal ulu and to account 
for the assumption by widely sundered peoples of k and none other, 
as shown in items 13-16 and 25-33. ‘The only objection to postulating 
a primal kulu lies in the labial group 17-19 and 31. To pass from 
palatal to labial is not in the scheme of the mutations of these lan- 
guages, all the more difficult to consider by reason of the extreme dif- 
ficulty of using the lips in the most of the Melanesian languages. It 
is easier to regard these labial forms as sporadically developed from 
ulu as a secondary stage of kuru. 
In establishing the un forms as derivative from kulu we encounter 
no phonetic difficulty, for the 1-n mutation is abundantly determined. 
We find confirmation in the close parallelism of the two types of form, 
kulu-kunu 15 and 27, kul—kun 13 and 26, gul-gun 16 and 29, ulu—unu 
11 and 23, ul-un 1 and 20; the only case in which we do not find a 
clear parallel is 30 gunu for which gulu is not yet discovered. Second- 
ary to the un type we find in 22 ongoi a mutation of n—-ng as well as a 
