46 SEAT OF SMELL PARTLY IN ANTENNA]. 



circular wall of cardboard. He then put some cock- 

 roaches on the table: they evidently scented the food, 

 and made straight for it. He then removed their 

 antennae, after whii'h, as long as they could not see the 

 food, they failed to find it, even though they wandered 

 about quite close to it. 



On the whole, then, the experiments which have 

 been made seem clearly to prove that in insects the 

 sense of smell resides partly in the antennae and partly 

 in the palpi. This distribution would be manifestly 

 advantageous. The palpi are more suited for the ex- 

 amination of food ; while the antennae are more con- 

 veniently situated for the perception of more distant 

 objects. 



We will now glance at the antennae and palpi 

 themselves, and consider briefly the structures which 

 are suppo-ed to give the sensation of smell. For 

 this three conditions are requisite : (1) an appropriate 

 nerve; (2) free access to air; and perhaps, though 

 this is not so clear, (3) a fluid which can dissolve the 

 odoriferous substance. 



The olfactory organ in Vertebrata consists, as already 

 mentioned, of a mucous membrane containing (1) 

 cylindrical epithelial cells, with a broad, flat termination 

 at the free end ; and (2) of rod-like filaments which, 

 some little distance below the surface, swell out into 

 a nut-shaped expansion, and then contract again into a 

 fine thread, which is probably continuous with the 

 fibrils of the olfactory nerve. 



In Insects and Crustacea the conditions are different. 

 The cellular "underskin," or hypoderm, secretes a hard, 

 horny envelope, and the terminations of the olfac- 

 tory nerves are enclosed in a horny tube with a 



