Page:Folklore1919.djvu/398

32 two ways of obtaining reproductions are mixed, and interchange of material under known conditions is effected. The present paper will deal briefly with results obtained from an application of the first two methods only.

The material chiefly employed in those experiments which are here to be described, consisted of folk-stories developed in a community very different from that to which my subjects belonged, and containing striking, curious, and often unfamiliar incidents and names. Picture material was also given to be reproduced, care being taken that the mere drawing of the pictures employed presented no great difficulty. Subjects read the stories over twice to themselves at their own normal reading pace. The pictures were studied for a period of four minutes. First reproductions were in all instances begun fifteen minutes after the original study of the material. In cases where a subject gave repeated reproductions, no reference was allowed to the original, or to his own earlier renderings, in any of the tests following upon the first. Detailed analysis and discussion of the results are impossible within the limits of this paper, but will be published later.

2.

The results under this head will be very briefly summarised:—

(a) The repeated reproduction of stories by the same individual revealed definite widespread tendencies toward change. These were largely dependent upon typical differences in the use of the various types of cue upon which reproductions may depend. Many subjects rely chiefly, for the details of their remembering, upon the use of words. In such cases the most important determining factors in the reproductions were the length and style of the original, together with the actual construction of the phrases. A subject of this class will often preserve some peculiar turn of phrase intact, even when incidents much more