Page:Memory (1913).djvu/47

2. In the later series: m = 620, P.E._o = ± 44.

Distribution of the Separate Values

Within the limits			i.e. within the deviation		Number of deviations			Of these deviations there occur
					By actual count	Calculated from theory		Below	Above
1⁄10		P.E.	±	4	3	4	.5	1	2
⅙		P.E.	±	7	5	7	.6	3	2
¼		P.E.	±	11	11	11	.3	6	5
½		P.E.	±	22	25	22	.2	13	12
		P.E.	±	44	44	42	.0	21	23
1	½	P.E.	±	66	56	57	.8	29	27
2		P.E.	±	88	71	69	.0	38	33
2	½	P.E.	±	110	76	76	.0	41	35
3		P.E.	±	132	79	80	.0	42	37

By both tables the supposition mentioned above of the existence of a less perfect but still perceptible correspondence between the observed and calculated distribution of the numbers is well confirmed.

Exactly the same approximate correspondence must be presupposed if, instead of decreasing the number of series combined into a test, the total number of tests is made smaller. In this case also I will add some confirmatory summaries.

I possess two long test series, made at the time of the earlier tests, which were obtained under the same conditions as the above mentioned series but at the later times of the day, B. and C.

One of these, B, comprised 39 tests of 6 series each, the other, C, 38 tests of 8 series each, each series containing 13 syllables. The results obtained were as follows: