[324] Elements, part i, chap. ii, fin.
[325] Lectures on Metaphysics, lecture xiv.
[326] Nature, vol. iii, p. 281 (1871).
[327] If a lot of dots or strokes on a piece of paper be exhibited for a moment to a person in normal condition, with the request that he say how many are there, he will find that they break into groups in his mind's eye, and that whilst he is analyzing and counting one group in his memory the others dissolve. In short, the impression made by the dots changes rapidly into something else. In the trance-subject, on the contrary, it seems to stick; I find that persons in the hypnotic state easily count the dots in the mind's eye so long as they do not much exceed twenty in number.
[328] Mr. Cattell made Jevons's experiment in a much more precise way (Philosophische Studien, iii, 121 ff.). Cards were ruled with short lines, varying in number from four to fifteen, and exposed to the eye for a hundredth of a second. When the number was but four or five, no mistakes as a rule were made. For higher numbers the tendency was to under- rather than to over-estimate. Similar experiments were tried with letters and figures, and gave the same result. When the letters formed familiar words, three times as many of them could be named as when their combination was meaningless. If the words formed a sentence, twice as many of them could be caught as when they had no connection. "The sentence was then apprehended as a whole. If not apprehended thus, almost nothing is apprehended of the several words; but if the sentence as a whole is apprehended, then the words appear very distinct."—Wundt and his pupil Dietze had tried similar experiments on rapidly repeated strokes of sound. Wundt made them follow each other in groups, and found that groups of twelve strokes at most could be recognized and identified when they succeeded each other at the most favorable rate, namely, from three to five tenths of a second (Phys. Psych., ii, 215). Dietze found that by mentally subdividing the groups into sub-groups as one listened, as many as forty strokes could be identified as a whole. They were then grasped as eight sub-groups of five, or as five of eight strokes each. (Philosophische Studien, ii, 362.)—Later in Wundt's Laboratory, Bechterew made observations on two simultaneously elapsing series of metronome strokes, of which one contained one stroke more than the other. The most favorable rate of succession was 0.3 sec., and he then discriminated a group of 18 from one of 18 + 1, apparently. (Neurologisches Centralblatt, 1889, 272.)
[329] Revue Scientifique, vol. 39, p. 684 (May 28, 1887).
[330] Cf. Chr. Wolff: Psychologia Empirica, § 245. Wolff's account of the phenomena of attention is in general excellent.
[331] Pflüger's Archiv, xi, 429-31.
[332] Physiol. Psych., 2d ed. ii, pp. 238-40.
[333] Ib. p. 262.