342 J. F. DASHIELL 



as shown in figure 5; and also that, as before remarked, changes 

 in vertical distances on the graph indicate differences in the mazes 

 and alleys used in different trials as well as changes in subjects' 

 ability. 



This matter of orientation to food direction is in line with gen- 

 eral observations on the behavior of very many (perhaps all) 

 higher animals when seeking their homes. The gray rat, for 

 instance, does not always follow the same path back to his hole 

 in the wall; and especially when partly blocked in his efforts to 

 return by obstacles or enemies in his path, will adopt circuitous 

 routes about the side walls or, in rare cases, across the open 

 floor, sometimes even progressing by stages from cover to cover, 

 always in the general direction of his hole. 



The matter may be summarized: In a series of changing mazes 

 with food box in one constant direction from the entrance, a 

 blind alley turning toward the general direction of the food box is 

 more likely to be entered than one turning away from this gen- 

 eral direction, in about the ratio of 3 to 1. This developing 

 tendency forms a factor in the transfer of maze training. 



Two criticisms may have suggested themselves to the reader: 



1. Was not the developing orientation in the food box direction 

 due to olfactory sensations? Unhappily, no experimental check 

 was employed on this question (the fuller possibilities of this 

 method of training group IV in the changing mazes were, indeed, 

 not in the experimenter's mind until the results came to be 

 worked over several weeks after). Internal evidence of a kind is, 

 however, to be found. Inspection of the maze designs used (fig. 

 4) in connection with the frequency table of most of the errors 

 made (table 3) will bring to light many cases of turns into blind 

 alleys at points where olfactory stimuli would surely have led 

 otherwise. See, for instance, these mazes and alleys: 3, C and E; 

 6, B; 7, D and E; 8, B; 9, D; 11, A and C; 12, A and C; 13, C; 

 14,E;16, D;19, D;20,E. 



2. Was not the orientation to the food box really an orientation 

 to cardinal directions (to a point of the compass) ? This matter 

 of orientation to absolute direction, irrespective of any particular 

 points or cues, seems to the writer a highly dubious idea. That 



