September 12, 1890.] 



SCIENCE. 



149 



proficiency in thus doing- several things at once that consti- 

 tutes much of the diiference between the expert and the 

 novice; and it is this "telescoping" process that seems to be 

 the method by which complicated operations are at length 

 performed in short times. It is for this reason that the time 

 per word of reading 100 words is shorter than the time of 

 reading a single word. Cattell reads a short word in 409(7, 

 a long one in 451(7, but 100 such in 2550' per word, and, if 

 the words are in construction, in 125(7 per word ; thus indicat- 

 ing how much of the difference between ordinary reading, and 

 reading single words, is due to the continuity of the experi- 

 ment, how much to the association between the words. So, 

 also, Cattell reads a single letter in 4240', but 100 such in 

 224(7 (compare XVII., XV., with LIH., LIV.). When the 

 series is too long continued, fatigue sets in, and the time is_ 

 again longer; it is longer for 500 than for 100 words and 

 letters; and for colors and pictures there is no saving in 

 naming 100 above naming a single color or picture. 



A special study of this power of grasping several things at 

 once was made by Cattell by having letters move along on 

 the surface of a rotating drum, and varying the width of a 

 slit in a screen through which they were read. When the 

 slit just allowed one letter to be seen at a time, they could 

 be read at the rate of one letter in 228(T; and as the slit was 

 widened to admit two, three, four, five, and six letters at 

 once, the rate increased to one letter in 200(T, 178(7, 166(7, 

 160(7, and 160(7. As it takes 424(7 to name a letter singly 

 (XV.), it would seem that the whole of a letter need not be 

 seen at once to be recognized, — an inference corroborated by 

 the fact, that, when the slit admits only one-tenth of a let- 

 ter at a time, the letters can be read at 400(7 per letter. The 

 result also indicates that there is a limit to the power in 

 question. M. Paulhan finds similar results in more complex 

 opei'ations. He multiplies numbers and recites a verse or 

 two at the same time ; and the time needed for this is shorter 

 than the sum of the times required to do each separately. 

 In very simple cases the time of doing both together is not longer 

 than the time for doing the more difficult of the two separately. 

 The mind should accordingly not be likened to a point at 

 which but a single object can impinge at one time, but 

 rather to a surface of variable extension. It should like- 

 wise be noted that the performance of a complex and ex- 

 tended mental task is not the same thing as the separate 

 performance of the several elements into which that task 

 may be analyzed. 



The distinction between the sensory and motor iorva of re- 

 action requires mention in this connection, because, when 

 applied to complicated adaptive re-actions, it seems to in- 

 volve overlapping of mental processes. The times cited in 

 the table (XXIV.-XXIX.) in Munsterberg's experiments (and 

 they are the only ones available for the present purpose) re- 

 fer to sensory re-actions. In these the attention is directed 

 to the word about to be uttered. It is recognized, and re- 

 ferred to its group. The corresponding movement is then 

 aroused and performed, the several processes being succes- 

 sive in time. In the motor form the word is thought of as a 

 "forefinger-moving'' word; and the movement upon which 

 the attention is kept fixed is expectantly kept ready to be 

 set off at the slightest notice. The several processes thus 

 play into one another, some perhaps entirely falling away. 

 Both anticipatory movements and errors (moving the finger 



next to the correct one) are not infrequent. The motor 

 times for the series XXIV.-XXIX. are 289, 355, 430, 432, 432, 

 and 437 (f; the differences between motor and sensory times, 

 94, 110, 258, 280, 461, 685(7. Until these very important and 

 striking results are better understood, it would be unwise to 

 enter into a discussion of them; but it may be noted (o) that 

 the increase in the complexity of the processes is more regu- 

 lar and prominent in the "sensory" times, the "motor" 

 times of the last four experiments of the series being about 

 alike; (6) that the "motor" complexity seems to be related 

 to the range of the impressions; and (c) that the differences 

 in time between the two modes of re-actions increase as the 

 processes become more complex. 



(7) Practice and Fatigue. What was said under these 

 headings of simple re-actions applies with equal force to 

 complex ones. Various experimenters notice the decrease 

 in time as the experiments proceed. They note that this 

 decrease is relatively greatest at first, and in those individu- 

 als and processes whose time is relatively longest at*lhe out- 

 set; also that it soon reaches a limit, and, when once thor- 

 oughly acquired, is not liable to be lost after a moderate 

 degree of disuse; and that it at times seems to be confused 

 with a transition from a sensory to a motor form of re-ac- 

 tion. As illustrative of one or other of these points, it may 

 be mentioned that Tischer finds as a rather typical case the 

 decrease of a distinction time from 160(7 in the first set to 

 95(7 in the second, and 86(7 in the third, all reduction ceas- 

 ing on the average after 5.5 sets; that Trautscholdt, in re- 

 actions consisting of repeating a word, finds times of 299, 

 273, and 258, and in another case of 205, 176, and 1550", in 

 three successive periods of fourteen days each ; that for Ber- 

 ger and Cattell, beginning with some practice in experi- 

 ments of this kind, find the time for incomplete re-actions 

 reduced by 30 and 20(7 after four months' experimentation; 

 and, finally, that the great decrease in the incomplete re- 

 actions of Kries and Auerbach (from 64 and 117 to 21(7, 

 from 153 and 109 to 36(7, from 104 and 97 to 49 and Uff, in 

 various experiments) strongly suggests a radical change in the- 

 mode of re-action. Another aspect of the effect of practice 

 appears in a study by Berger of the times required by the 

 boys of the nine classes of a German Gymnasium, and of 

 the class preparatory to the Gymnasium, to read 100 and 

 500 words in construction in German and in Latin at a 

 maximum and at a normal rate. There is a constant de- 

 crease in time as the boys advance in age. In Latin the several 

 times per word were 262, 135, 100, 84, 79, 57, 54, 49, 48, 430"; 

 in German 72, 55, 43, 37. 39, 28, 27, 26, 25, 23(r; the great 

 difference between the first two times in Latin being due to 

 the fact that the boys who required 262(7 to read a Latin, 

 word had never learned Latin at all. That these differences 

 are to be referred to specific practice rather than to general 

 mental maturity, appears from a comparison of the above 

 times with the times required by those boys to name colors ; 

 viz., 135, 99, 119, 123, 100, 91, 112^, 99, 86 (f. 



The results regarding fatigue are not equally definite. 

 Many mention the general fact of fatigue, and to avoid ifc 

 perform but few experiments in a series. We have already 

 seen that it takes relatively longer to read 500 letters, words, 

 colors, pictures, than to read 100. On the other hand, Cat- 

 tell, after a very long series of re-actions, found no serious 

 or constant increase in the time, but seemed to feel the ef- 



