MAN-SMELLING BY SAVAGES 91 



in the water. The snowflakes, we may say, have 

 different quahties, according to their chemical com- 

 position or elements, and these elements produce the 

 various stimulating effects on the plants they come 

 in contact with. 



It strikes me as a roundabout or cumbrous way of 

 getting the effect; that it might do in the case of a 

 man or rhinoceros or dog, but in the case of the 

 infinitely more subtle scent in the insect it could not 

 possibly work. Probably it doesn't. 



But I must not go into that imperfectly-known, 

 fascinating insect realm, since the subject we are just 

 now concerned with is man's sense of smell, and of 

 man smelling man. 



That we are not conscious of the smell of people 

 about us is not evidence that it is the same with primi- 

 tive races; nevertheless, I conclude from my observa- 

 tion of both man and animals that savages are not 

 conscious of the smell of those they consort with, 

 except at certain times, as, for instance, after being 

 separated from them for some time, but that they are 

 always conscious of the smell of a stranger. Also that 

 the mother knows the smell of her own babe, and 

 that there is a double advantage in this — the smell 

 stimulates her maternal feeling and prevents infants 

 getting changed by accident, which would otherwise 

 sometimes happen, as the children of savages resemble 

 each other so closely. 



No doubt there are many instances in smell, as 

 in other senses and faculties, of atavism; they would 

 tell us much, but unfortunately they are not recorded. 



