ive brain weight of the f ena] e ia greater tljan that of the 

 male. Fxperimente upon individual rata v/ere conducted as near- 

 ly as possible at the sane time of day, thus fornin^-, feeding 

 rhythms. 



There are three methods of estimotinti perfection 

 in experiments relatin^i to the behavior of animals; the num- 

 ber of erroi's, the distance traversed and the time required. 

 In any case it is hard to form a judtgrnent as to what consti- 

 tutes an error and especially so in a comparative study of 

 this kind where it is possible for the personal prejudice of 

 the experimenter to enter. At the time this investigation v/as 

 begun there was no adequate means for measurint? the distance 

 traversed. This left the time consumed as the only criterion 

 of perfection. However, time is the criterion used by most 

 investigators in the animal field. Hicks^, in summing up the 

 experimental results of several investigators concludes, that 

 "time is the best single criterion, inasmuch a,s it represents 

 all phases of the process of lefnrning, and since it will yield 

 the most, comparable results at the hands of different inves- 

 tigators." In timing the rats a very accurate Sv/iss split-sec- 

 ond 8top-v,'a,tch was used. Under ideal conditions, perhaps, the 

 animal should be presented to the problem by one person, tim- 

 ed by another, while the experimenter himself should merely 

 record results. But timing very soon becomes automatic; when 

 the rat is crossing the starting line it is almost impossible 

 to inhibit the impulse to press the button. 



Hicks, The Relative Values of the Different Curves of Learn- 

 ing, Jl. of An. ^eh., vol. I, pp. 138 ff. 



At the conclusion of the experimentn all the rwts 



