Page:Mind (Old Series) Volume 11.djvu/401

400 L. T. STEVENS : VI. 1-480 1-548 + 4*6 per cent. 750 836 + 11-5 744 792 + 6-5 660 655 - -8 590 571 - 3'2 500 489 - 2'2 360 345 - 4-2 260 253 - 2-7 VII. 1-458 1-615 + 10-8 979 992 + 1-3 730 733 + -4 600 588 - 2-0 472 446 - 5-5 400 354 - 11-5 These Tables show very plainly the main point of the paper, namely, that the error made in the reproduction of the longer intervals is plus, of the shorter ones, minus; and that at some point between the two extremes the error approaches zero. The value of this indifference-point is found not to be constant, but to vary for different individuals between % 53 and -87 sec., the average being about % 71 sec. The position of the indifference- point is not so accurately defined in Case III. as it is in the others. The reason for this is not evident. Of the six regular experiments made for intervals from '55 to '70 sec., the mean value of the standard interval is -61 sec., and the average of the error of reproduction is -3 per cent., the individual variations not being, algebraically, more than 1 per cent. With reference to the amount of variation, the number of experiments is, in the majority of cases, not sufficient to give the results a quantitative value. The amount of variation depends upon the fixedness of the attention on the work, and upon the experience which one has had in estimating and holding intervals of time. The complaint was made quite frequently by several of the gentlemen of their inability, at times, to keep their whole attention upon the experiment ; and subsequent examination of such experiments showed that the error made vas either contrary to the law above stated, or was in the proper direction but abnormally large in amount. The effect of practice was well seen iu Case III., where a great number of observation swas made. In the first few experiments very large errors were made ; but as the work proceeded, they gradually diminished in amount, until they reached, on the average, the size that is represented, for the various intervals, in the Tables. Tables I. and II. were obtained from gentlemen who are musicians by profession ; in the second case, where numerous experiments were made, the uni- formity in the amount of variation was a noted feature from the very start. Moreover, in both these cases there was not that difficulty which was experienced by certain of the others, in grasping the longer intervals of time.