More important than the assimilations and dissimilations between directly given impressions are those that arise out of the reciprocal action of a direct impression and of ideational elements, which belong to previous impressions, and therefore arise by means of an act of memory. Reproductive assimilations of this kind we have already met with in our reading experiments (see p. 26). We saw there that a well-known word can in general be read almost instantaneously, although its scope greatly exceeds that of the focus of attention. It is clear that this great facilitation in apprehension is only possible owing to the familiarity of the object, because by its action it gives rise to the reproduction of former corresponding impressions, and thereby causes the completion of the image only partially perceived. We can convince ourselves of this in a striking manner by means of reading experiments, in which certain letters of a fairly long word have been voluntarily altered. Such changes are then in general only partially or not at all perceived in these quick reading experiments. It may easily happen that we take the following combination of letters "Miscaldoniousness" for the word "Miscellaneousness," although four out of the seventeen letters of the word have been changed, If by chance our attention is very strongly concentrated upon one of the wrong letters, we can perceive the mistake, but for the other wrong letters the right ones are as a rule substituted. It is obvious that this phenomenon is exactly the same as the one we continually meet with when we overlook misprints in a book, only that in our experiments a false reading is greatly favoured by the shortness of the exposition-time. In all these cases we generally take it for granted that it is nothing more nor less than an inaccurate apprehension, as the expression "overlook" suggests. Yet our rapid reading experiments convince us that this expression is really incorrect. In reality it is not a mere not-seeing of the wrong letters, but a seeing of the right ones in the place of the wrong ones. If we call into our mind directly after the experiment the image we have seen, we can see very often in those very places, where a wrong letter stands, the right letter in the full distinctness of an immediate impression. This is, of course, only possible if the wrong letter is displaced by the reproduction of the right one. Such a process is obviously made up of two parts--firstly, the displacement of the wrong letter, and secondly, the reproduction of the right one. Naturally both acts take place quite simultaneously, and therefore we may look upon the displacement as an effect of the reproduction. In this combination of the two acts an assimilation process and a dissimilation process are joined together. By means of an assimilation caused by the other letters the right letter is reproduced, and this together with all the rest of the word has a dissimilating effect upon the wrong letter.