80 



according to the above, be also transcribed [trge.-^a, 

 vakka^, kom^]. The above examples go to prove that the 

 same difficulty about the phonetic transcription of cert-ain 

 sounds exists wherever the reflexed or inverted [r] is found. 

 I question, however, whether the Somersetshire "hum" for 

 ''run," or the American ''purty" for /'pretty," would be 

 correctly represented by [ha : nj and [pa : ^i]. Either [harn] and 

 [parti], or [hrn] and [p^ti], with a syllabic [r], would appear 

 preferable. 



The following remarks by Teichelmann and Schiirmann 

 show the difficulties they encountered in dealing with the 

 question : 'V is changed with I or d; as kurlana, kullana 

 (recent) ; garia, gadla (fire) ; murla, mulla (dry) ; r is omitted 

 before n, as marnkutye, mankutye (three) ; marngandi, man- 

 gandi (to request) ; nurnti, nunti (away) ; r before t is 

 changed into t, as ngartendi, ngattendi (to ask)." 



[f] is the bilabial v, the ordinary sound of German w 

 (was, ewig), and of the Spanish h and v between vowels 

 (caho, ave). It is rare in Australian languages, but was 

 distinctly heard (only in the middle of words), and is per- 

 haps a variant of [w] or [b]. 



The stressed or accented syllable is indicated by the sign 

 ['] placed before it, but as the great majority of native words 

 are accented on the first syllable, the accent mark is omitted 

 whenever such is the case. Thus the words [parkana] and 

 [gamidi] are accented like the English words "perjury" and 

 "company.''' Where there are four syllables, as in 

 [bindjanidja], ['>7arin7;J6ri], and [jerabula], the main stress is 

 still on the first syllalDle and there is a slight secondary stress 

 on the third syllable, as in the English "tantalizing." Why 

 certai^ words, such as [tan' dan ja, ka'rawi, pu'lombi, 

 ba'lardu] are accented on the second syllable, is a question 

 whose solution would require a more thorough knowledge of 

 native languages than we possess. 



Letters and words in the phonetic alphabet occurring in 

 the ordinary text are placed within square brackets. 



Voiced and voiceless consonants. — The sounds of j) and h 

 are alike to the native, and it is the same with t and d, k and 

 g. He pronounces [kaka, kaga] or [gaga], ''head," in- 

 differently; so also [tidna] or [didna] for "foot," and [papa] 

 or [baba] for dog. It is impossible to imagine the Australian 

 aboriginal taking an interest in his own language or reducing 

 it to writing and using it for literary purposes (as has been 

 done by one or two of the North American tribes), but, if 

 such a thing were done, it is probable that only the voiceless 

 consonants fp, t, and k) would be used, at least at the begin- 

 ning of words; otherwise much confusion would occur. 



