SENSORY CAPACITIES AND INTELLIGENCE OF DOGS 



that the animal was not responding to 

 some cue from himself. The results 

 reported seem to indicate that the smells 

 of nitro-benzol and benzaldehyd, which 

 man cannot distinguish, are also confused 

 by the dog. On the other hand, o.oi per 

 cent nitro-benzol was analyzed out of a 

 mixture of o.i per cent each of eugenal, 

 ionon, linalool, and zimtaldehyd. A 

 number of other mixtures were also suc- 

 cessfully analyzed. As to absolute thresh- 

 holds, vinegar was distinguished in a 

 dilution of i:io 6 , salt and vinegar acids 

 in dilutions of Hio 5 , salt and quinine in 

 dilutions of i'.io 4 . Buytendijk's second 

 method made use of glass dishes which 

 contained the odorous substances each 

 held in a sort of basket with a handle 

 permitting the dog to pick them up singly 

 and retrieve them. The procedure was 

 essentially that used by Henning with the 

 handkerchiefs, except that but three re- 

 ceptacles were used at a time. Solutions 

 of formic and of sulphuric acid containing 

 one part in ten million were correctly 

 responded to, as were also extremely dilute 

 solutions of a number of other acids. All 

 such solutions were said to be quite odor- 

 less to man. Here again there exists the 

 possibility that the dog reacted to cues 

 given by the experimenter. 



To what practical use can the dog put 

 his olfactory ability? In the first place it 

 is so generally accepted that certain dogs 

 are able to follow the trails of other animals 

 that special tests of this ability have 

 apparently seemed unnecessary. This 

 does not mean, however, that a dog can 

 distinguish between the trail of animal x 

 and animal y of the same species nor, 

 perhaps, that he can distinguish between 

 the trails of individuals of closely related 

 species, although most hunters will say 

 that dogs readily do so. Bingham (3) 

 reports that a setter who normally pointed 

 birds very skillfully had on occasions 



followed trails which proved to have 

 been made by turtles, and he states that 

 other cases of this sort have been brought 

 to his attention. It seems probable that 

 olfactory reception is dominantly if not 

 solely responsible for much trailing be- 

 havior. That trails are followed more 

 readily when they have been recently made 

 is doubtless due to the fact that the odor- 

 ous particles, emanating most probably 

 from the footprint of the trailed animal, 

 diffuse rapidly and finally reach a degree 

 of dilution such as to render them below 

 the lower threshold of the dog. Tests 

 made by Buytendijk on a German Shep- 

 herd dog (4) indicate, among other 

 things, the effect of wind on the trail. 

 When following a trail leading straight 

 into the wind, or when following a trail 

 behind an obstruction which cut off the 

 wind the dog followed almost exactly 

 the steps of the man being trailed. But 

 when the trail crossed open country at 

 right angles to the wind the dog did not 

 follow exactly the path of the man but 

 ran along rather uniformly from one to 

 two meters or more to the leeward of this 

 path. 



Although Romanes' tests of the sense of 

 smell in his female setter (36) have fre- 

 quently been referred to, certain interest- 

 ing details have been overlooked. It will 

 be remembered that this animal could 

 trail Romanes successfully even though a 

 number of people were made to walk 

 single file behind him in such a way as 

 apparently to obliterate his trail. By 

 appropriate tests Romanes found that the 

 dog would not follow his trail when he 

 was barefoot, in stocking feet, or wearing 

 a pair of new boots, that the dog would 

 follow the trail of a stranger only if that 

 stranger wore a pair of Romanes' old 

 boots. It was apparently the old boots 

 which were being trailed. Romanes then 

 proceeded to paste heavy brown paper en- 



