抄録
The liniguistic geography has not founded its strong foothold in geogra-phy. In order to make the actual state of linguistic phenomena clear from a geographic stand-point, we researched the distribution of twenty-five words and sotalds in an area teared to pieces by many dendriform. mountain chains. We chose 1.47 old women as the informants at random including every settlement in the area.
It is very interesting that words change their form from place to place, and, in addition, it seems that the existence or non-existence of words should be noticed most. Especially the latter may tell us the economic and cul-tural differences between seaboard settleirtents and inland ones, e. g., tyoro (a kind of fishing boat) (Fig. 8) etc. But non-existences of words appear even in the seaboard settlements quite often. This probably shows us the actual state of linguistic life in connection with economy or culture. In some cases, at least, the distribution area of a word has an indiscriminate borderline, i.e., so entangled with the others. Even in a smalll settlen-ment, in which people seen quite homogeneous culturally, different words to on object are used in practice.
According to distribution of the the word kousi (calf) whose kinds of calling were more than twenty-five (Fig. 2) and to the vowel system of this dialect we can presume a relation of this area to western Japan connected by the marine communication in past days. And as the words of the common Ja-panese entered this dialectal and educationally homogeneous area, they had no prominent core regions (Kernlandschaft) but a scattered distribution.
