UTAH ACADEMY OF SCIENCES 211 
this law in an experimental situation has always seemed 
to me a practical and valuable experiment. But to do so 
one would need to employ an activity which involved little 
or no forgetting, otherwise, the loss during the recess from 
forgetting would far more than compensate any gain that 
would come from the recess. Little progress was made 
in the matter of finding a learning exercise which involved 
no forgetting until the writer came across a study made by 
Judd, while still a teacher at Yale, which dealt with practice 
on some of the optical illusions until the illusion would 
be overcome. Judd found that after the illusion had been 
overcome by constant practice that the learning, and such 
it actually is, would persist without modification for a 
period of a year. 
Judd had given us a clue. But, what were we to do? 
Judd’s apparatus could not be duplicated because of the 
enormous expense. We hit upon the device known as 
mirror tracing. This seemed to involve the illusion and 
was open to accurate and easy use in the laboratory. Our 
first experiment with the mirror tracing was a complete 
success. We found that after a certain degree of efficiency 
had been secured, that this was not only retained but often- 
times improved upon after a lapse of nine months. James 
was verified and we found ourselves in possession of data 
far richer than James had ever suspected, for it was not 
only shown that improvement would occur over the recess 
interval, but that some recess was absolutely necessary for 
improvement to take place at all. It was also shown that 
the amount of recess or time out between practices which 
was necessary for the highest efficiency varied enormously 
with individuals. This led us to feel that we had found 
a device by which we could measure the time element of 
voluntary control. The results from futher experimenta- 
tion more than justified our hopes. Below, are indicated 
some of the conclusions from our studies which bear di- 
rectly upon the problem of control. 
To summarize our findings here: In the early stages 
of the learning the recess period is absolutely necessary for 
any improvement in accuracy; the least improvement can- 
