180 THE ROYAL SOCIETY OF CANADA 
des Provinces Maritimes empruntés aux Langues Sauvages (Quebec, 1906) ,* 
and by P. G. Roy’s Les Noms Géographiques de la Province de Québec 
(Lévis, 1906). At its best this stage involves the collection of the 
earliest forms of the names, not only from printed works, as in the case 
of the above-mentioned books, but also from MS. records in land offices, 
ete., and it involves also the use of data obtained by interviewing the 
living Indians. My own earlier work upon the Place-nomenclature of 
New Brunswick, (published in these Transactions, II, 1896, ii, 175-289) 
belongs in this class. 
Ill. Toe INVESTIGATIONAL STAGE.—In this the strictly scientific 
method of induction is followed. All preferences, preconceptions, and 
prejudices are eliminated so far as can be done. All possible sources 
of information are searched and their relative values are carefully 
weighed. Particular importance is attached to data from original and 
disinterested sources, especially the earliest recorded forms of names, 
which are sought in such matter-of-fact places as ancient deeds, grants, 
etc. The psychological and etymological principles involved in name- 
giving and word-changing are taken into account. All available sources 
of information about the native language are made use of, especially 
all existent dictionaries, vocabularies, and grammars,—for the con- 
struction of the native language as well as its vocabulary is important. 
The testimony of the living Indians is sought, but with the understanding 
that this is subject to many errors, and that the trained white philo- 
logist, with his knowledge of the psychological and philological, as well 
as the etymological and geographical data, can reach conclusions far 
more likely to be correct than those of even the most intelligent Indian. 
The investigator also makes himself acquainted, personally and thor- 
oughly, with the places in question, since the connection between 
Indian place-nomenclature and the natural characteristics of a place 
is usually very close. All of these data taken together constitute the 
evidence, which the investigator marshals and weighs, precisely as in 
any other scientific investigation or in any court of law (for the methods 
and aims of the two are substantially identical), deducing therefrom a 
verdict, which may be “proven,” “probable,” “possible,” or “uncer- 
tain.”’ This is the final stage in the study of the subject. It has not 
yet been applied in any part of Canada. 
* Partly in the interest of historical accuracy, and partly for my own satisfac- 
tion, I wish to point out that a large proportion of the interpretations of New Bruns- 
wick place-names in Rouillard’s work, although credited to Father Bourgeois (of 
St. Joseph’s College, Memramcook) are identical with those in my earlier work (in 
these Transactions) on New Brunswick Place-nomenclature, and were evidently 
taken from that work, though without the customary acknowledgment, by Father 
Bourgeois. 
