ON THE ETHNOLOGICAL SUR\'EY OF CANADA. 371 



substituted for each other ; indeed, the same collector will find himself Writing the 

 same word sometimes with one vowel, sometimes with another. 



CONSONAN'tS. 



I pointed out in my notes on Sk-qo'mic that our division of certain consonants 

 into surds and sonants was not applicable to that tongue. The same remarks apply- 

 here. I believe I should be quite within the truth if I asserted that the speech of 

 the coast and delta Salish possessed neither surds nor sonants as we understand 

 those distinctions in English. I gravely question whether their t's, k's, s"s and p's 

 are accurately represented by our surds t, k, s, and p (/ they do not possess at all, 

 nor its corresponding sonant v). They seem rather to occupy an intermediate posi- 

 tion, and partake of the quality of both surd and sonant. At times this is plainly 

 discernible, and one is at a loss to render the sound accurately. In the tribal names 

 of the Kwa'ntlEn, for example, the k is certainly not our 7c, nor is it better rendered 

 by the corresponding sonant </. It is rather a rare quality of sound combining both 

 elements. The native^ is another example. Rarely is it uttered like our p, and still 

 less is it our b. It can only be described as something between the two, and must be 

 heard in the mouth of a native to be fully appreciated. The native pronunciation 

 of English terms makes it quite clear that some at least of their ' surds ' do not 

 correspond to ours ; and that, although they have no sonants, as distinct from surds, 

 in their own speech, they yet invariably convert a surd into a sonant when speaking 

 English. For example, David calls bacon, bagon ; pit he calls bit ; and bite, bide. 

 Veal he pronounces as beal, and barrow as farrow. These are but a few examples, 

 but they serve to illustrate my point : 



I find the following consonantal sounds in Tcil'Qe'uk : — 



t, approximately as in English. 



t, intermediate between our t and d. (I do not distinguish this from the other.) 



k, approximately as in English. 



k, intermediate between our k and g. (I do not distinguish this from the othen) 



k', palatised k, almost Tiy. 



q, as cli in loch in broad Scotch. 



Q, approximately as wh is uttered in North Britain. 



H as the German oh in icli. 



h, m, y, as in English. 



n, sometimes as in English, sometimes with a suspicion of I about it, 



1, ,1 }, J) )> )j »» '^ -it 



S, ,, ,, M J) ,, ,) *" »» 



ss, as in English M&s. 



p, rarely as in English ; generally intermediate between our p and &j 



c, as in English sh ; initial and medial. 



c, as th in the word thin ; initial, medial, and final. 



tc, as cli in the word church ; initial, medial, and final. 



ts as in English Fitz ; initial, medial, and final. 



kl, as cl in the word climb ; initial and rriedial. 



tl, an explosive Z, approximately like the II in Welsh ; initial, medial, and final. 



si, as in English ; initial and medial. 



kw, as qu in the word quantity ; initial and medial. The combination of these 

 two consonants occurs of tener than any other in Tcil'Qe'uk. It is the predominating 

 element of its vocables. Some writers treat them as consonant and vowel. I 

 cannot think this to be correct ; w is here as much a consonant as the h is, being 

 followed in every instance by a vowel. 



The consonants in Tcil'Qe'uk are fairly determinate in character. Certain of 

 them, however, show a tendency to permutation ; tc and ts are perhaps the 

 commonest interchanges. I was at a loss for a long time whether to write the 

 tribal name Tcil'Qe'uk or Tsil'Qe'uk. In the mouths of some Indians s runs 

 uniformly into c. s, q, and tl are also common interchanges. To mark the hiatus 

 which occurs in the uttering of some words I have employed the apostrophic sign, 

 placed over the word, thus : — Tcil'Qe'uk. 



ACCENT. 



Accentuation is as marked in Tcil'Qe'uk as in Sk'qo'mic, and as difBcult to bring 

 under rule. The most general rule is that which seems to hold good throughout all 



B B 2 



