The Sense of Smell 



vents it from being heard; a bright light 

 eclipses a feeble gleam. These are waves of 

 the same nature. But the roar of thunder 

 cannot cause the least jet of light to pale; nor 

 can the dazzling glory of the sun stifle the 

 least sound. Being of different natures, light 

 and sound do not influence each other. 



The experiment with the lavender-oil, 

 naphthaline and the rest would therefore seem 

 to prove that odour proceeds from two 

 sources. For emission substitute undulation; 

 and the problem of the Great Peacock is ex- 

 plained. Without losing any of its substance, 

 a luminous point shakes the ether with its 

 vibrations and fills a circle of indefinite width 

 with light. This must almost express the 

 working of the mother Bombyx' tell-tale dis- 

 charge. It does not emit molecules: it vi- 

 brates; it sets in motion waves capable of 

 spreading to distances incompatible with a 

 real diffusion of matter. 



In its entirety, smell would thus seem to 

 have two domains: that of the particles dis- 

 solved in the air and that of the ethereal 

 waves. The first alone is known to us. It 

 belongs also to the insect. It is this which 

 informs the Saprinus of the dragon arum's 



329 



