MICHELSON. ] LIST OF STEMS. 617 
378, 501), though also from Oliver Lincoln and Thomas Scott, the evi- 
dence contained in the Fox text previously published by me (Bulletin 
72 of the Bureau of American Ethnology) and grammatical notes on 
them, as well as the evidence contained in Jones’ Fox Texts, and the 
published Cree, Ojibwa, and Algonkin material. In a few instances I 
have used the facts shown by my unpublished Sauk, Shawnee, Cree, 
Ottawa, Potawatomi, etc., texts and grammatical notes. In one or 
two cases the Kickapoo tales collected by William Jones and trans- 
lated with comments by myself (Vol. 1x, Publications of the American 
Ethnological Society) have been of value, but the rather inadequate 
phonetics haye prevented me from using them systematically. As 
a whole, it may be stated that the number of uncertain stems has 
been reduced to a minimum. In every case where possible a few 
references are given to the texts by page and line to further facilitate 
its study. Where no such references are given, it is to be understood 
that such stems were abstracted from the grammatical notes on them 
(vide supra) or from my unpublished Fox texts. 
Stems which can not occur in initial positions are given with a 
hyphen before and after them. The list shows how much more 
numerous the stems which occur in initial positions are than those 
which can not. A word may be here said about certain stems which 
are given with an alternate terminal 4 4, e. g., nepa&-, nepii-. This 
means that the first and second persons singular of the independent 
mode end in -*", that 4 is used in all persons of all subordinate modes, 
changing to ai when immediately followed by such suffixed pronouns 
as begin with y (not taking into account cases where an auxiliary occurs 
between the stem and suflixed pronoun), that 4 is used in the first 
person plural exclusive and inclusive as well as second person plural— 
all of the independent mode; but that & is used in the third person 
animate and inanimate (where this occurs), singular and plural, of 
the independent mode. Stems given with terminal 4 change this 
to -"’ when the 4 occurs at the end of a word, but retain the 4 other- 
wise throughout save in the third person animate singular and 
plural of the participial mode, in which cases 4 is substituted for it; 
also this a appears as & before g. Broadly speaking, stems with & 
are noninitial, while those with 4 4 are initial. There are, however, 
some exceptions to this rule, e. g., kiwi-, return. There is evidence 
to show that this dual classification is not confined to Fox but likewise 
occurs in a number of Algonquian languages (for example, Shawnee 
and Peoria). 
It will be noticed that many auxiliaries and copulas are given with 
final t or n, but variants without these are also given. Up to the 
present time it has not been possible to formulate laws governing this 
3599°—257——i40 
