PHYSIOLOGY OF OLFACTION 61 



watery and the olfactory hairs oily, it follows that any 

 substance that gains entrance into the body of the hair 

 must first have been dissolved in water and then in oil. 

 From this standpoint Backman attempted to determine 

 whether there was any relation between the effectiveness 

 of certain olfactory stimuli and their solubility in water 

 and in oil. Water and olive oil, each at 30 degrees centi- 

 grade, were used as the test solvents. Thus methyl alco- 

 hol and ethyl alcohol, which are without strong odor, 

 were found to be freely soluble in water, but only very 

 slightly soluble in oil. Hence while they would dissolve 

 abundantly in the olfactory mucous, they would fail to 

 enter the hairs to any great extent in consequence of which 

 their effectiveness as stimuli must be, according to Back- 

 man, very slight. On the other hand normal butyl alco- 

 hol has a strong odor and its efficiency as a stimulus was 

 believed to depend upon the fact that it is soluble in water 

 to the extent of 8.3 per cent and in oil to an almost indefi- 

 nite amount. Other substances showed somewhat different 

 relations. Thus chlorbenzol could be detected at a dilu- 

 tion of 6.7 x 10.~ gram-molecules per liter of air, and is 

 soluble in water to the extent of 0.25 per cent and in oil 

 indefinitely. Brombenzol could be smelled at the some- 

 what greater dilution of 1.1 x lOr 8 gram-molecules per 

 liter of air; yet it is less soluble in water (0.045 per cent) 

 than chlorbenzol though indefinitely soluble in oil. In 

 these instances the degrees of solubility in water are the 

 reverse of the effectiveness of these two substances as 

 olfactory stimuli. Possibly solubility in oil, as intimated 

 by Larguier des Bancels (1912), is of much more signifi- 

 cance for olfactory stimulation than solubility in water. 

 If the olfactory hairs in man are provided with flagella, 



