408 psYCHOLoar. 



fore us breaks iuto parts disconuected with each other, and 

 forming each as it were a separate object or system, not 

 conceivable in union with the rest, it becomes harder to 

 apprehend all these parts at once, and the mind tends to 

 let go of one whilst it attends to another. Still, within 

 limits this can be done. M. Paulhan has experimented 

 oavefully on the matter by declaiming one poem aloud 

 whilst he repeated a different one mentally, or by writing 

 one sentence whilst speaking another, or by performing 

 calculations on paper whilst reciting poetry.* He found 

 that 



■" the most favorable condition for the doubling of the mind was its 

 sinultaneous application to two easy and heterogeneous operations. 

 Two operations of the same sort, two multiplications, two recitations, or 

 the reciting one poem and writing another, render the process more 

 uncertain and difficult." 



The attention often, but not always, oscillates during 

 these performances ; and sometimes a word from one part 

 of the task slips into another. I myself find when I try to 

 simultaneously recite one thing and write another that the 

 beginning of each word or segment of a phrase is what re- 

 quires the attention. Once started, my pen runs on for a 

 word or two as if by its own momentum. M. Paulhan 

 compared the time occupied by the same tAvo operations 

 <lone simultaneously or in succession, and found that there 

 w^as often a considerable gain of time from doing them 

 simultaneously. For instance : 



" I write the first four verses of Athalie, whilst reciting eleven of 

 Musset. The whole performance occupies 40 seconds. But reciting 

 alone takes 22 and writing alone 31, or 53 altogether, so that there is a 

 difference in favor of the simultatieous operations." 



Or again : 



"I multiply 421 312 212 by 2; the operation takes 6 seconds; the 

 recitation of 4 verses also takes 6 seconds. But the two operations 

 done at once only take 6 seconds, so that tbere is no loss of time from 

 combining them." 



Of course these time-measurements lack precision. 

 "With three systems of object (writing with each hand whilst 

 reciting) the operation became much more difficult. 



* Revue Scientifique, vol. 39, p. 684 (May 28, 1887). 



