5 
In these texts I have used one, and only one, symbol for each Cree 
phoneme (a digraph for is), and have not adopted extra symbols for vari- 
ations within any Cree phoneme, such as fronted and backed a, or voiced 
k; such distinctions, however precious to a foreign ear, are irrelevant to 
the language we are considering. 
But where variation involves different Cree phonemes, I have tried to 
record it; as, when a is fronted all the way to (Cree) i or backed all the 
way to (Cree) u; dh-apiyan, ah-apiyin : when thou sittest there ; dsawaham 
he crosses it (water), dsiwaham, dsuwaham, dsowaham. 
In these texts I have kept the forms as they were actually taken from 
dictation. The only exceptions are, first, that my notes are here reduced 
to the distinctive sounds of Cree, irrelevant variations (such as attempts 
to note the fluctuating quantities of s) being eliminated; second, I have 
corrected forms which repeated experience showed to be, beyond reason- 
able doubt, errors of mine, the correct form being, again beyond reasonable 
doubt, available. In this I have been extremely conservative; the reader 
will be able to correct many a form that I have left in the text either 
because I am not absolutely sure that the deviation is due to my error, or 
because, having erred, I was not sure of the proper correction. Where the 
eccentric form seems to interfere with the sense, a note is appended; it 
is just in those cases that alteration is dangerous: the form may be rare or 
archaic; if not, even an informant's “slip of the tongue” may be of interest, 
and lies beyond our right of correction. Or again, in some of the texts a 
word will appear consistently in divergent form; I have left it so, for it 
may have been the speaker and not my ear that deviated. 
I have tried to adapt punctuation and word division to the structure 
of Cree, making the former as intelligible and the latter as complete as 
possible. 
The hyphen is used between members of compounds. In noun com- 
pounds the first member usually adds -i ; paskwdwi-mustus : buffalo 
(paskwdw : prairie). This merges with preceding postconsonantal w to u; 
masku-pimiy : bear’s fat {maskwah bear, stem maskw-). In the remaining 
compound words the first member is a particle, e.g. poni-kimiwan : it 
stops raining. The hyphen distinguishes these compounds from unit 
words, such as pondyimdw : he stops thinking of him {-dyimdw does not 
occur as an independent word). 
This distinction has been partly troubled in Cree. For instance, 
historically, atim : dog, horse, is an independent word, whereas -astim 
in the same meaning, is a medial stem, as in wdpastim : white dog or horse. 
But actually we find, on the one hand, misaiim : horse, where historically 
we should have either a compound word, *misih-atim.y *7nis-dtini {misih : 
big), or a unit word *7nistasiim. On the other hand, we have 7natsastim : 
evil dog, where historically we should have either the compound *matsih- 
atim, *mats-dtim, or the unit word *7natasti7n. Similarly, we find forms 
like kihtsdyimow : he is conceited, where the initial stem kiht- : big, has 
taken on the form of the independent particle kihtsih : big, beside the 
historically justified kihtdyimow. 
Consequently my use of hyphen will not always be found consistent. 
A form like nahapiw : he sits down, might well be taken as nah-apiw, 
since there is a particle nahih .* well adjusted, and a verb apiw : he sits 
there. Yet we write nahapiw because the ordinary type of sandhi for a 
