360 JOSEPH PETERSON 



favoring recency and frequency factors is the result of the learn- 

 ing. A completed habit must give 100% of sUch reactions. 14 



It is not contended here that our results show the effects of 

 frequency and recency on behavior to be negligible. On the 

 contrary, their effects are obvious in any detailed study of the 

 rat's learning in the maze, pursued by the method of analysis 

 here developed. But so far as the bringing about of the short 

 cuts (the elimination of useless acts) in learning is concerned, 

 recency and frequency factors do certainly not seem to play the 

 important part that they have been considered to play in maze 

 learning. It is but natural to suspect that the same thing will' 

 hold for other types of learning. It seems that we are in need of 

 searching analyses of the detailed aspects of all sorts of learning. 

 Our initial spurt of progress in the study of learning has passed 

 and mere time, error, discrimination, and average attainment 

 curves of general results can no longer solve the problems that 

 we are coming to as soon as we begin more detailed studies. 



So far as experimental evidence goes at present it would seem 

 that maze learning by rats agrees in the main with the results 

 to be expected on the basis of probability - frequency factors 

 as their general results were pointed out in the early part of 

 this paper; that is, that the blind alleys nearest the food box 

 are first eliminated, and that entrances to blind alleys are great- 

 est near the starting place in the maze and decrease for the 

 successive cut de sacs directly with their nearness to the food box. 

 Recent results published by Hubbert and Lashley 18 seem to agree 

 with this conclusion, though these results raise other problems the 

 solution of which is not yet made clear. Miss Hubbert found in 

 an earlier research 16 no invariable sequence in the elimination 

 of blind alleys, but the more recent article cited admits that 

 " when averages of very large groups of animals are taken there 

 does seem to be progressive elimination of errors for the food 

 compartment to the entrance of the maze." 17 Miss Vincent's 18 



14 See an erroneous conclusion by Hamilton in his interesting monograph, A 

 Study of Perseverance Reactions in Primates and Rodents. Behav. Mon., Ser. 

 No. 13, 1916, pp. 38-46. 



16 Hubbert, H. B., and Lashley, K. S. Retroactive Association and the Elimina- 

 tion of Errors in the Maze. Jour. Animal Behav., 1917, 7, 130-138. 



16 Elimination of Errors in the Maze. Ibid., 1915, 5, 66-72. 



"These averages as given in the later article are, goinp - in the order from en- 

 trance place toward the food box: 30.6, 26.4, 19.7, 19.7, 18.7, 8.3. 



18 Vincent, Stella B. The White Rat and the Maze Problem. IV. The Num- 

 ber and Distribution of Errors: A Comparative Study. Jour Animal. Behav., 

 1915, 5, 367-374. 



