xi KNOWLEDGE OF CONCRETE OBJECTS 259 



arising out of a similar reaction in the past, with which a 

 satisfactory result .s associated. It is, on the surface, an 

 argument against this theory, that if the direct or custom- 

 ary way into a room is barred, a dog or cat will at once l 

 betake itself to any other route that there may be. This 

 adoption of alternatives suggests that their action is to be 

 referred, not to an impulse urging them to move in a 

 particular direction, but rather to a desire to be in a par- 

 ticular place, the road to which is indifferent, so long as it 

 is passable. 



By way of testing the point, I made a number of experiments 

 with dogs about the house in which I live, to which all of them 

 were strange. Like most houses that go back for four centuries, 

 this house is full of turnings and windings, and afforded oppor- 

 tunities for devising on a small scale intricate and alternative 

 routes. My method was to get the dog's master to stay outside 

 the house, while I took the dog in, and brought it to the window, 

 whence it could see its master, who would then call it. After- 

 wards I would reverse the process, the master going inside, while 

 the dog remained outside. In all cases, the use of scent was pre- 

 vented, by taking care that the master did not go over the track 

 which the dog would have to follow. The dog, it will be 

 observed, had to respond to its master's call in quite unfamiliar 

 circumstances. It has no association between taking this turn or 

 that turn, and the satisfaction of rejoining its master. 



In these as in other experiments the dogs showed a great 

 variation in intelligence, initiative, and persistence. 2 The most 

 noticeable defect was a failure in " orientation," the result .of 

 which was that on getting out of the house, for example, the 

 dog would turn towards the spot where his master had last been 

 found instead of towards that in which he had last been seen. 3 

 In such cases, the dog would as a rule rapidly return, and either 

 go back into the house, and perhaps to the window again, or (as 



1 As a rule. I confess I have seen dogs fail in this respect. 



2 Putting aside the case of the puppy mentioned above, there were 

 comparatively few complete failures (though there were often initial 

 mistakes) when the master was outside. The dogs were also very fairly 

 successful in finding their master in a ground floor room, but uniformly 

 failed when he called to them from upstairs. This may be attributed 

 partly to greater complexity, but mainly to the impossibility for a dog of 

 identifying an upstair room by its window. The ground floor room the 

 dog could look into and recognise. 



3 I only noted one clear case in which the more familiar route was at 

 once rejected for the right one, against many in which a mistake was 

 made at this point. 



S 2 



