248 CLASSIFICATION OF ALGONQUIAN TRIBES [ETR. ANN, 28 
son pl. an. subj., third person sing. an. object) as compared with 
Cree -atcik, Fox (participial) -atcig'; sk{i]k[i] becomes ss, -ss (subj. 
mode, third pl. an. subj., second person sing. object), Cree -skik, Fox 
-kig' (part.). Further, it may be noted that final -w*, w¢ after con- 
sonants, has a history in Montagnais different from that in Cree. 
Observe Montagnais ni—ku HE—ME (independent mode), Cree ne—k, 
Fox ne—gw*, tshi—ku HE—THEE (independent mode), Cree ke—k, 
Fox ke—gwa, -iku (first person pl. incl. of subjunctive), Cree -yuk, 
Fox -yagw’. These phonetic changes are of extremely wide appli- 
cation. It is unnecessary to give tables showing the verbal termina- 
tions as they agree with those of Cree. It may be noted that -v 
corresponds to Cree -w and -au to Cree -ow, except in the first person 
pl. incl., where we find -w. The reason for the latter is not clear. 
After emphasizing the essential unity of Cree and Montagnais it 
may be well to point out some individual traits of the latter. In the 
first place though there is a pan (Cree pun) preterite, it is confined to 
the indicative and does not occur in the subjunctive. Another point 
is that the ‘‘suppositif’’ of the mode “subjonctif”’ is clearly allied 
to the Fox potential subjunctive for which there is no correspondent in 
Cree (compare Mont. -iakukue we (excl.),-ikuakue we (incl.),-iekwekue 
YE with Fox -yagdge®, -yagagu “, -yagdgu'%, respectively). The other 
intransitive persons in Montagnais have the characteristic ku but 
have no correspondents in Fox. The transitive forms do not corre- 
spond closely, though there are resemblances between the two lan- 
guages; hence tables are not given. In closing, it may be added that 
the Montagnais on—me, etc., has the appearance of a passive in 
structure, but there are several points which are not clear. (The 
above examples of Montagnais and Cree are taken, respectively, from 
Lemoine and Horden, with the exception of Cree askiy, which is from 
Lacombe. It will be seen by consulting the tables of Fort Totten. 
Cree that the terminal & of Horden is doubtless the strong (impure) 
sonant g of the former, Fox, Sauk, Kickapoo, Ottawa, etc. A couple 
of examples of Sapir’s Montagnais, ts*inipaha@’wats THOU KILLEST 
THEM an. (Fox kenepahdwag'), tsimpahé'wats HE KILLED THEM an. 
(Fox kicinepahdwag'), ickwe'wats WOMEN (Fox vkwawag’'), illustrate 
the principles mentioned above. The writer suspects that Skinner’s 
tci a (Rupert’s House Cree) THOU is really tstya. The initial ¢s* at 
once classes the word as Montagnais. It is true that according to 
Lemoine the ordinary Montagnais correspondent has 7, not y; but it 
should be noticed that in Cree dialectically kiya occurs (see Horden, 
Cree Grammar, p. 3, London, 1881; Lacombe, Dictionnaire de la 
Langue des Cris, p. xv, Montreal, 1874). The Rupert’s House Cree 
then would correspond to this.) 
In discussing the relations of other Eastern-Central Algonquian 
languages, it is understood that Montagnais agrees with Cree unless 
