6 AN AUSTBALIAX LAXGUAGE.
Ill Samoan and in other Polynesian dialects, nrj is very common as an initial, and as a final too in the whole of Melanesia. In this respect the Polynesian and the Melanesian langnages are akin to the Anstralian. The Malay also uses ufj both as an initial and as a final. Some Australian dialects nasalise the A', as in the English word ' ink '; to this there are parallels in the Melanesia!! languages, and there the sound is represented by h or q.
In Tamil, one of the Dravidian languages of India, with which our Australian language is supposed to be connected, one forma- tive suffix is y?/, nasalised into ngu ; it is used as the initial sound of a syllable, as in ni-ngu, 'to quit' ; to this extent it corresjioiids with our ncj.
Our author, in his edition of 1831, has in some words a doubled fjuttural-nasal, as in bungngai. As the second of these is only a cj attracted by the nasal that precedes it, I have written such words with g-g. In fact, the double so!ind proceeds froi!! the one nasal, as iii our English word ' finger.' Some of the Melanesian languages have this double sound both with cj and with Tc.
But in both of its uses, initial and final, the Australian ncj arises from the nasalisation of the guttural (] ; it is a simple sound, and should therefore be represented by only one letter, not by the digraph nrj. In Sanskrit, the symbol for it as a filial, for there it is never used as an initial and seldom as a final, is ii- ; but, as the Australian ng comes from y, I prefer to use g as its symbol. If we compare the Dravidian pag-al, 'a day,' with the Melane- sian bung, 'a day,' it is clear that the ng proceeds from a g^ for the original root of both words is the verb bha, 'to sine.' Eurther examination may, perhaps, show that our ng is, in some cases, a modification of the sound oin, as in the Erench ' bon,' 'bieii,' or even of a final vowel, but at present that does not seem to me at all likely.
Besides ng, there are the two subdued nasal sounds of n and in — that is, n before d, and m before & ; these harden the con- sonant that follows, and produce such sounds as nda^ mba. The same sounds are common in Eiji — a Melanesian region — but not in Polynesia.
Of the palataJs, the language has cli, as in the English word 'church,' and y, as in 'jam'; to these may be added the conso- nant //. The ch and the ./ sounds are, in some vocabularies, printed as tcli and Oj ; that is quite unnecessary. I have adopted 6 as the symbol for cZ*, because it is a simple sound.
The only cerebral that we have is r, although the sound of it is often so asperated as to resemble the Dravidian rough and hard r. Our r is neither the Arabic vibrating gliv, nor the Xorthumbrian hurr, but is more like the rolled r of the Parisians.
�� �
