688 



SCIENCE. 



[N. S. Vol. III. No. 71. 



viduals for sucli a reaction is 588 a, with a 

 variation of 84 <t. It is thus quite clear 

 that, wliile the simple reaction time for the 

 two special subjects is much shorter than 

 the normal, their time is just about normal 

 in a reaction involving a simple distinc- 

 tion and choice, and is considerably longer 

 than the normal in a reaction involving a 

 complex distinction and choice. 



The incident related of Houdin, the 

 ' king of the conjurers,' regarding his re- 

 markable powers of taking in at a glance 

 the miscellaneous contents of a shop win- 

 dow, suggests another power of great use 

 to the prestidigitator. Mr. Hermann claims 

 to possess a similar power, although he 

 does nothing in his stage performances 

 that demands such a comprehensiveness of 

 perception. I exposed for h a second 10 

 patches of color requiring him to name as 

 many he could see ; in each of two trials 

 he named five correctly. When the color 

 patches were different in shape as well as 

 in color he was able to see three in i a 

 second and describe them correctly. He 

 was also able to read two words in the same 

 time. I also counted the number of con- 

 secutive exposures of \ second each needed 

 for the reading of a sentence containing 17 

 words; it required 10 exposures or 1.7 

 words per exposure. In one-second expo- 

 sures Mr. Hermann could read 3 isolated 

 words, and required 8 exposures to read a 

 sentence of 29 words or 3.6 words for each 

 exposure. 



Similiar averages for a group of about 40 

 persons indicate about the same quickness 

 of perception for color 4.5 as compared with 

 5; an inferior perception for combined 

 color and form 1 .8 as compared with 3, only 

 12% of those tested recognizing as many as 

 three color forms ; and likewise for words 

 seen separately 1.4 as compared with 2 

 (22% reading 2 words), but a distinctly 

 higher average of the number of words read 

 in one exposure. On the whole, these few 



experiments would indicate that, as regards 

 the quickness and scope of perception, Mr. 

 Hermann would rank well (except in read- 

 ing words in a sentence), but by no means 

 exceptionally well in the general average. 



For Mr. Kellar the tests were somewhat 

 differently arranged. The patches of color 

 and the various forms were arranged con- 

 secutively and were read in order as one 

 would read words on a line. In exposures 

 of one second Mr. Kellar could read cor- 

 rectly four colors and three forms. In 

 reading words scattered over the page he 

 read 2 correctly in his first trial and 3 

 in second trial. In four successives expo- 

 sures of 1 second each he read a sen- 

 tence containing 27 words, or an average 

 of 6.75 words per second. Mr. Kellar would 

 thus rank below Hermann in all but the 

 reading of words in a sentence, in which he 

 far exceeds him, but would be equalled by 

 about 86% of a group of college students. 



Another form of testing this capacity was 

 attempted, but with no success. Mr. Her- 

 mann was shown 10 pictures, and asked to 

 study them for about 45 seconds ; he was 

 then shown a card containing 40 pictures 

 and requested to mark off which of the 40 

 were also seen on the former card. He 

 marked off 7, 4 of which were correct and 

 3 wrong. For Mr. Kellar this test was 

 more systematicallj' made. He was shown 

 a card containing 40 pictures and at the 

 same time a slip containing 10 words, the 

 names of a certain 10 of the pictures; and 

 asked to find the pictures named by the 

 words as rapidly as possible. This took 

 him just 45 seconds, the average of a mis- 

 cellaneous group of individuals being 64 

 seconds. A few minutes later he was asked 

 to note on a card containing 60 pictures as 

 many as he could remember having seen 

 on the former card containing 40 pictures. 

 He succeeded in recognizing but 11, the 

 average in this test being 17.5 pictures. I 

 also tested Mr. Kellar's visual memory by 



