412 PSYCHOLOGY. 



a heterogeneous impression is suddenly brought. Then comes the 

 question, with which member of the series do we perceive the additional 

 impression to coincide ? with that member with whose presence it 

 really coexists, or is there some aberration? ... If the a<lditional 

 stimulus belongs to a different sense very considerable aberrations may' 

 occur. 



"The best way to experiment is with a number of visual impressions 

 (which one can easily get from a moving object) for the series, and 

 with a sound as the disparate impression. Let, e.g., an index-hand 

 move over a circular scale with uniform and sufficiently slow velocity, 

 so that the impressions it gives will not fuse, but permit its position at 

 any instant to be distinctly seen. Let the clockwork which turns it 

 have an arrangement which rings a bell once in every revolution, but 

 at a point which can be varied, so that the observer need never know 

 in advance just when the bell-stroke takes place. In such observations 

 three cases are possible. The beil-stroke can be perceived either ex- 

 actly at the moment to which the index points when it sounds — in this 

 case there will be no time-displacement ; or we can combine it with a 

 later position of the index — . . . poaitive time-displacement, as we 

 shall call it ; or finally we can combine it with a position of the index 

 earlier than that at which the sound occurred — and this we will call a 

 negative displacement. The most natural displacement would appa- 

 rently be the positive, since for apperception a certain time is always re- 

 quired. . . . But experience shows that the opposite is the case : it 

 happens most frequently that the sound appears earlier than its real 

 date — far less often coincident with it, or later. It should be observed 

 that in all these experiments it takes some time to get a distinctly per- 

 ceived combination of the sound with a particular position of the in- 

 dex, and that a single revolution of the latter is never enough for the 

 purpose. The motion must go on long enough for the sounds them- 

 selves to form a regular series — the outcome being a simultapeous per- 

 ception of two distinct series of events, of which either may by changes 

 in its rapidity modify the result. The first thing one remarks is that 

 the sound belongs in a certain region of the scale ; only gradually is it 

 perceived to combine with a particular position of the index. But even 

 a result gained by observation of many revolutions may be deficient in 

 certainty, for accidental combinations of attention have a great influ- 

 ence upon it. If we deliberately try to combine the bell-stroke with 

 an arbitrarily chosen position of the index, we succeed without diffi- 

 culty, provided this position be not too remote from the true one. If, 

 again, we cover the whole scale, except a single division over which we 

 may see the index pass, we have a strong tendency to combine the 

 beli-stroke with this actually seen position ; and in so doing may easily 

 overlook more than i of a second of time. Results, therefore, to be of 

 any value, must be drawn from long-continued and very numerous ob- 

 servations, in which such irregular oscillations of the attention neutral- 

 ize each other according to the law of great numbers, and allow the 



