2. Nevertheless, several dialects have forms which show the agglutinative words on the way to become inflexional. In the dialect of Western Australia, 'the woman's staff' is yago-ȧk wanna, in which the -ȧk has lost its independence, aud is as much a case-ending as the <«, i, or is of the Latin genitive. So also in Awabakal ; the -umba of kokara cmoumba, 'my house,' may be regarded as inflexional ; for, although the -ba can be de- tached and used as a separate word, not so the -um. I believe the -limba to be a weathering for gu-mba, the gu being a dialect form of the post-position ko, as in AViradhari ; yet the -XI cannot stand alone ; the m belongs to the ba.
3. As to the Cases of nouns and pronouns, they are shown by separable post-positions which are themselves nouns, adjectives, or verbs. The post-position birun g, for example, meaning 'away from,' is an adjective in the Wiradhari dialect, and means ' far distant,' while birandi, another form from the same root, is the post-position, 'from.' The other post-positions in the paradigm on page 1(5 are all taken from the monosyllables ka and ko. Of these, I take ko to be a root- verb, implying ' motion to,' and ka another, meaning ' to be ' in a certain state or place ; but of their origin I can give no account, unless ka be related to the Dravidian verb agu, already noticed, and ko be a modiflcdform of ka. These tAvo roots, variously combined, beeom.e the post- positions kai, kin-ko, ka-ko, kin-ba, ka-ba, ka-birung, kin- birung on page 16 ; by the influence of the final consonant of the words to Avhich they are joined, the initial k of those becomes t, I, or r.
A similar account of the post-positions in the ISTarrinyeri, the Diyeri, and other distant dialects could, no doubt, be given, but from the scantiness of our knoAvledge, that is at present im- possible.
4. As to the Gender of nouns, that is either implied in the meaning of the Avord or to be guessed from the context. In Fijian, a word is added to mark the gender ; for example, gone is ' child,' and, from it, a gone tagane is ' a boy,' but a gone alewa is 'a girl.' The Samoans say ull po'a and ull fafine to mean a ' male dog ' and a ' female dog,' and the Ebudans something similar. Our Australians have no such devices, but they iiave some Avords in which the gender is clearly distinguished by an ending added on, or by a change of the voAvel sound of the final syllable of the word. The most commtm feminine sufiix is -gun; as, mobi, 'a blind man,' mobi-gun, 'a blind Avoman'; yinal, 'a son,' yinal-kun, 'a daughter'; another suflix is -in; as, AAvaba- kal, 'amanof AAvaba,' AAvaba-kal-in, 'a Avomanof Awaba '; ma- koro-bau, makoro-bin, 'a fisher-man,' 'a fisher-Avoman,' showa change in the vowel sound. I think that, in proportion to the extent of the language, instances of this kind — the expression of
�� �
