478 



tilting a group, the time required by the last syllable of a group 

 indicates the rhythm used by the observer. 



As to the tables illustrating the changes which the rhythm undergoes 

 when the familiarity with the material increases, we have divided 

 the number of repetitions required in learning and repeating a series 

 into three groups of successive repetitions in the manner just described 

 (Vide Proceedings of the meeting of 31 March 1917, Vol. XXV 

 p. 1316). We then determined the mean time of every syllable of 

 the various groups and intervals. R. and D. learned the series in 

 three groups of four syllables. M. in two groups of six, each sub- 

 divided into three groups of two. The stress fell invariably on the 

 last of the two syllables. For the rest the tables are self-evident. All 

 timevalues are expressed in seconds. Tables I, II, and 111 refer to 

 the rhythm of the first repetition, while Tables IV to IX show the 

 changes in the rhythm resulting from the observer's greater famili- 

 arity with the material. 



When studying the data of Table 1, II and III we see directly 

 that our observers used rhythm already at the first repetition, with 



I as well as with II. M. alone, seems to make an exception with 



II ; anyhow with him the time falling to the last syllable of the 

 group is shorter than the mean duration of the remaining syllables 

 constituting the group. With all our observers the rhythm with I 

 undergoes a slight acceleration from the very beginning, which is 

 rather peculiar. (M 0.82, 0.80 ; R 1 .07, 1.04, 1.01 ; D 1.02, 0.99, 0.86). 



While the mean time of the syllables, constituting the various 

 groups, is gradually diminishing from the beginning to the end, the 

 intervals decrease for R also (1.51, 1.40); for D however they remain 

 approximately constant (1.43, 1.45). Since M learned the series in two 

 groups of six syllables, no data referring to him could be procured. 



In the repetition experiments the duration of every syllable is, as 

 a rule, shorter than the corresponding duration in the learning- 

 experiments. The same may be observed regarding the intervals, 

 viz. a decrease with all observers. The acceleration of learning in 

 the repetition-experiments benefits the group as well as the inter- 

 vals. Just as the rhythm of the learning-experiments that of the 

 repetition-experiments is also slightly quickened from the beginning 

 towards the end, as the acceleration, observed in the learning-and 

 in the repetition-experiments, follows, almost without exception, the 

 order of the groups and the intervals, i. e. the farther group and 

 interval are advanced in the series, the greater is the decrease. 



With II there is in the learning-experiments a tendency to 

 constantly increase the velocity while retaining the rhythm, as the 



