MEMORY. 667 



All improvement of memory consists, then, in the in- 

 provement of one's habitual methods of recording facts. 



by heart will shorten the time it takes to learn an entirely different kind of 

 poetry. During eight successive days I learned 158 lines of Victor Hugo's 

 ' Satyr.' The total number of minutes required for this was 131 1 — it should 

 be said that I had learned nothing by heart for many years. I then, work- 

 ing for twenty- odd minutes daily, learned the entire first book of Paradise 

 Lost, occupying 38 days in the process. After this training I went back to 

 Victor Hugo's poem, and found that 158 additional lines (divided exactlj'as 

 on the former occasion) took me 151| minutes. In other words, I commit- 

 ted my Victor Hugo to memory before the training at the rate of a line in 

 50 seconds, after the training at the rate of a line in 57 seconds, just the 

 opposite result from that which the popular view would lead one to expect. 

 But as I was peceptibly fagged with other work at the time of the second 

 batch of Victor Hugo, I thought that might explain the retardation ; so 1 

 persuaded several other persons to repeat the test. 



Dr. W. H. Burnham learned 16 lines of In Memoriam for 8 days ; time, 

 li-17 minutes — daily average 14f. He then trained himself on Schiller's 

 translation of the second book of the ^neid into German, 16 lines daily 

 for 26 consecutive days. On returning to the same quantity of In Memo- 

 riam again, he found his maximum time 20 minutes, minimum 10, average 

 14||. As he feared the outer conditions might not have been as favorable 

 this time as the first, he waited a few days and got conditions as near as 

 possible identical. The result was . minimum time 8 minutes ; maximum 

 19| ; average 14^\. 



Mr. E. S. Drown tested himself on Virgil for 16 days, then again for 

 10 days, after training himself on Scott. Average time before training, 

 13 minutes 26 seconds ; after training, 12 minutes 16 seconds. [Sixteen 

 days is too long for the test , it gives time for training on the test-verse.] 



Mr. C. H. Baldwin took 10 lines for 15 days as his test, trained himself 

 on 450 lines 'of an entirely different verse,' and then took 15 days more 

 of the former verse 10 lines a day. Average result: 3 minutes 41 seconds 

 before, 3 minutes 2 seconds after, training. [Same criticism as before.] 



Mr. E. A. Pease tested himself on Idyls of the King, and trained him- 

 self on Paradise Lost. Average result of 6 days each time : 14 minutes 34 

 seconds before, 14 minutes 55 seconds after, training. Mr. Burnham hav- 

 ing suggested that to eliminate facilitating effect entirely from the training 

 verses one ought to test one's self a la Ebbinghaus on series of nonsense- 

 syllables, having no analogy whatever with any system of expressive verses, 

 I induced two of my students to perform that experiment also. The record 

 is imfortunately lost ; but the result was a very considerable shortening of 

 the average time of the second series of nonsense-syllables, learned after 

 training. This seems to me, however, more to show the effects of rapid 

 habituation to the nonsense-verses themselves than those of the poetry 

 used between them. But I mean to prosecute the experiments farther, 

 and will report in another place. 



One of my students having quoted a clergyman of his acquaintance 

 who had marvellously improved by practice his power of learning his 



