218 E. C. MACDOWELL AND E. M. VICARI 



of averages might be demanded throughout, but the doubUng 

 of every table would result in unjustifiable bulk, especially in 

 the light of the clearness of the results as they stand. 



When it became necessary to make drastic reductions in the 

 size of the experiment, eleven rats in this generation, six tests 

 and five controls, were taken out of the training after they had 

 been given the eight days of training. They lack the training 

 in the multiple-choice apparatus as well as retention on the maze. 

 In order to have the same group of rats summarized for the 

 maze and the multiple-choice apparatus, these eleven rats were 

 not included in the major part of the calculating that has been 

 done with the maze data. However, from the standpoint of 

 this paper, these rats, as far as they went, are as good as others; 

 accordingly, they have been included in the most important 

 summaries. The time data are summarized, including, as well 

 as excluding, these rats which we call 'incompletes,' since their 

 training was not completed. There are then three sets of rats 

 to be designated 'completes,' 'incompletes,' and 'failures.' 



/. Irregularities of data 



The number of rats included in the various averages is indi- 

 cated in the tables; these mmabers vary for different averages 

 even within one group of rats. The details explaining each of 

 these irregularities are given in the following paragraphs. Al- 

 though we feel fully justified in making the minor exclusions 

 and inclusions besides the main ones treated above, we wish 

 to remove all suspicion of our having tiried to influence the result 

 one way or the other in so doing. These explanations are made 

 here all together to avoid many scattered foot-notes. 



Rat 1279: Record sheets for the first day lost; this rat is not 

 included in the averages for the first day or for first half of train- 

 ing, all training, training and retention. 



Rat 1311: Record sheets for third day lost; this rat is not 

 included in averages for the third day, first half of training, all 

 training, training and retention. 



