Page:Mind (Old Series) Volume 11.djvu/301

 IX. NOTES AND COERESPONDENCE. A SUPPOSED LAW OP MEMORY. Mr. J. Jacobs, in his careful and benevolent account of my little book, Ueber das Gediichtnis (Mixo, Vol. x. 454), tries to deduce from some of my results a very simple numerical relation. I had investigated the depen- dence existing between different quantities of nonsense-syllables and the number of repetitions required to learn them by heart. Though the resulting numbers proved remarkably uniform in two sets of experiments, I was not able to unite them by a simple mathematical formula. Mr. Jacobs thinks he has found out the following one : The number of repetitii >ns required to learn a given set of syllables by heart is (in my case) treble the excess of syllables over the threshold (i.e., over the number of syllables learnt without repetition). The latter number being, for instance, 6, a row of 12 syllables would be reproduced after 3 (12-6) = 18 repetitions, a row of 20 syllables after 3 (20 -6) = 42 repetitions. Perhaps this relation, though seemingly confirmed to some extent by my experimental results, will appear rather too simple to be very credible. The conditions of my experiments were simple enough in comparison with the intricacy of repro- duction in ordinary life, but they were still very complicated in comparison with elementary psychological relations ; my numerical results therefore can hardly be expected to be expressible by very simple formulas. On this account I refrained when I first read Mr. Jacobs's hypothesis from com- menting upon it. But since Mr. Jacobs, following up his idea, makes it the starting-point of some further remarks in a paper read before the British Association and published in MIXD 41, I must say a word about it, in order to avert further confusion. The supposed law is based upon a very unfortunate and very regrettable misprint in my book. The Table upon which Mr. Jacobs founds his suggestion, and which he correctly re- prints from page 64 of my book, contains the number 26 instead of 36. By introducing the latter into the above formula, the calculated result becomes too different from the observed one to leave any possibility for the supposed relation. In the first place, of course, I myself am responsible for thus leading my critic astray, and I sincerely regret what- waste of time and thought has been caused by the incorrect figure (by the way, the only one I have hitherto detected in the numerous Tables oi' my book). But it will only be fair to acknowledge that the fault rests not with me alone, and that the error of Mr. Jacobs was not necessarily determined by the slight neglect I committed. On page 63, immediately preceding the one in question, the right number 36 is mentioned twice, and on page 65, im- mediately following the mischievous Table, there is first a diagram and secondly a paraphrase of the results obtained, which both contain again, explicitly and implicitly, the correct number 36. Berlin. H. EBBINGHAUS. 'FALSEHOOD' AND 'KIXOHAXCK' ix PLATO. I am surprised that neither Dr. Martineau nor 1'rof. Sidgwick (in MINI> 41) has appealed to the well-known pa.-.-agc /,', // /c3S2, which is obviously referred to in Jt'ji. 53"), one of the pl;i-es under discussion. 1 do not see how the former passage c;in leave any doubt at all asto what the conviction is which these two places are intended to express, n';.., that ignorance, f] ev