474 Flom 2 pers. pi. form dokker is used in polite address to one as well as to more than one just as in Bergen. Han and ho are used as substitutory pronouns for words for animals and things as well as in reference to persons, "i overensstemmelse med det natur- lige eller det grammatiske kjp'n." That means: in agreement with either the natural or the grammatical gender; or does it mean: in agreement with the natural gender, or (if natural gender is not present) with the grammatical gender, (as in words for things and concepts). Possibly the latter, for one says (must say?) han about ei slaastkj&mpa as well as about oxen, b dten, varmen, etc. That this is the meaning is seen in the later sentence which points out, that "Tr. tar her bestemt avstand fra rnalf^re som kan bruke det grammatiske kj0n i strid med det naturlige." The extensive use of han in the dialect is well illustrated in the prevailing preference of han som for den som (han so si det, han lyg). The account of the use of the subjective-reflexive pronoun is one of the most interesting and important parts of the study. There has here evidently been a conflict in progress for a long time between the levelling of two forms under one against the demands of clearness, which latter tended to maintain both. Now one, now the other, has been in the ascendency. And it is not a little surprizing that as recently as the close of the XlXth century the personal pronoun was displacing the old reflexive. Then we learn, further, that a reaction set in, and at the present time the reflexive seems definitely victorious. I hope elsewhere to be able to go into this question more fully than I could now, in connection with dialectal usage elsewhere, and shall therefore confine myself here to the above mention of the condition in the Tr. dialect. The indefinite det is much more restricted in Tr. than in either literary Nw. or the dialects. In the first place, it is not used in reference to natural phenomena or the weather; there the pronoun han is used, exactly as in WNu. dialects, and in considerable measure elsewhere I am inclined to think. In the second place, many impersonal verbs and impersonal uses of verbs have been replaced by personal constructions, as & kl0r so i haue (but det feile han ingenting) . The tendency is in this direction in Nw. dialects in general but possibly Tr. has gone rather farther than most. In connection with the Verbs, pp. 46-69, it is to be noted that the auxiliary of the perfect is always ha, not only in, e.g., har kommet f#r seint, but also with bli, as ho har blidd. The omission of the auxiliary, as in many Nw. dialects and as in Swedish, is not evidenced; the pure future tense always requires the auxiliary komme til at. Finally as to the passive. With regard to the compound passive it should be noted as an
especially interesting matter that, while the auxiliary bli is the
