30 RUTLEDGE T. WILTBANK 



learnings showed a large saving relative to the group with 

 one, this is deprived of any significance by the fact that, 

 in the case of the other four mazes, two of them proved 

 easier for the group with one prior learning than for the 

 group with two. A comparison of the groups with three 

 prior learnings and the groups with four discloses the fact 

 that in two instances — namely, in mazes B and C, the 

 groups with three prior learnings learned the maze more 

 readily than the groups with four; and, if there were 

 three instances where the groups with four prior learnings 

 did better than those with three, one of these was but 

 .1 of a trial better. This table might accordingly be brought 

 forward in support of the statement that three maze- 

 learnings are as good as four, measured by the beneficial 

 effect upon the learning of a subsequent maze according 

 to the standard of trials. But the actual differences in 

 these figures representing the number of trials necessary 

 to learn the third and fourth mazes are not large enough 

 to give much weight to such a statement. 



The D maze presents the least evidence of any beneficial 

 effect accruing from the learning of preceding mazes, and 

 the principle which is to be invoked in explanation is the 

 one used with reference to the negative figure representing 

 the transfer in trials, as may be seen on pp. 18 ff. The num- 

 ber of trials needed by the groups coming to this maze with 

 two and with three prior learnings was greater than the 

 number made by the control-group, and the number needed 

 by the group coming with four learnings was also high. 

 These figures were all greater than any in the A and B 

 mazes, the control-figures of which show them to be some- 



TABLE 24 



The Rankings of the Groups upon the Mazes, According to the Average 

 Number of Trials per Rat 



Maze A Maze B Maze C Maze D Maze E Av. Rank 

 Group having learned 



four preceding mazes 1 3 2 2 1 1.8 



Group having learned 



^Aree preceding mazes 2 1 1 3 2 1.8 



Group having learned 



two preceding mazes 3 4 3 4 4 3.6 



Group having learned 



one preceding maze .4 2 4 1 3 2.8 



