[573] Cf. Fechner, Psychophysik, ii, 499.
[574] The primary after-image itself cannot be utilized if the stimulus is too brief. Mr. Cattell found (Psychologische Studien, iii, p. 93 ff.) that the color of a light must fall upon the eye for a period varying from 0.00275 to 0.006 of a second, in order to be recognized for what it is. Letters of the alphabet and familiar words require from 0.00075 to 0.00175 sec.—truly an interval extremely short. Some letters, E for example, are harder than others. In 1871 Helmholtz and Baxt had ascertained that when an impression was immediately followed by another, the latter quenched the former and prevented it from being known to later consciousness. The first stimulus was letters of the alphabet, the second a bright white disk. "With an interval of 0.0048 sec. between the two excitations [I copy here the abstract in Ladd's Physiological Psychology, p. 480], the disk appeared as scarcely a trace of a weak shimmer; with an interval of 0.0096 sec., letters appeared in the shimmer—one or two which could be partially recognized when the interval increased to 0.0144 sec. When the interval was made 0.0192 sec. the objects were a little more clearly discerned; at 0.00336 sec. four letters could be well recognized; at 0.0432 sec., five letters; and at 0.0528 sec. all the letters could be read." (Pflüger's Archiv, iv, 325 ff.)
[575] When the past is recalled symbolically, or conceptually only, it is true that no such copy need be there. In no sort of conceptual knowledge is it requisite that definitely resembling images be there (cf. [pp. 471] ff.). But as all conceptual knowledge stands for intuitive knowledge, and terminates therein, I abstract from this complication, and confine myself to those memories in which the past is directly imaged in the mind, or, as we say, intuitively known.
[576] E.g. Spencer, Psychology, i, p. 448. How do the believers in the sufficiency of the 'image' formulate the cases where we remember that something did not happen—that we did not wind our watch, did not lock the door, etc.? It is very hard to account for these memories of omission. The image of winding the watch is just as present to my mind now when I remember that I did not wind it as if I remembered that I did. It must be a difference in the mode of feeling the image which leads me to such different conclusions in the two cases. When I remember that I did wind it, I feel it grown together with its associates of past date and place. When I remember that I did not, it keeps aloof; the associates fuse with each other, but not with it. This sense of fusion, of the belonging together of things, is a most subtle relation; the sense of non-fusion is an equally subtle one. Both relations demand most complex mental processes to know them, processes quite different from that mere presence or absence of an image which does such service in the cruder books.
[577] Psychologia Empirica, § 174.
[578] Analysis, i, 330-1. Mill believed that the various things remembered, the self included, enter consciousness in the form of separate ideas, but so rapidly that they are 'all clustered into one.' "Ideas called up in close conjunction ... assume, even when there is the greatest complexity, the appearance, not of many ideas, but of one" (vol. i, p. 123). This mythology does not impair the accuracy of his description of memory's object.
[579] Compare, however, [p. 251], Chapter IX.
[580] Professor Bain adds, in a note to this passage of Mill's: "This process seems best expressed by laying down a law of Compound or Composite Association, under which a plurality of feeble links of connection may be a substitute for one powerful and self-sufficing link."
[581] Analysis, chap. x.
[582] H. Maudsley, The Physiology of Mind (London, 1876), p. 513.