PHYSIOLOGY OF OLFACTION 63 



very difficult to explain the variety of olfactory sensa- 

 tions on the basis of stimulation by radio-activity, but 

 the assumption that the stimuating materials act chemi- 

 cally on the substance of the receptor is in easy accord 

 with the di^ejrsity of olfactory experience. 



5. Olfaction in Fishes. It has already been pointed 

 out that most fishes possess paired olfactory sacs whose 

 structure and innervation are essentially identical with 

 the corresponding parts in the air-inhabiting vertebrates. 

 Nevertheless currents of water flow through these sacs 

 and such stimulation as they receive must come from 

 these currents. Nagel (1894), who was one of the most 

 vigorous opponents to the idea that the olfactory organs 

 were stimulated by solutions and believed that gases or 

 vapors were the only real stimuli for these receptors, 

 was led to conclude that the so-called olfactory organs of 

 fishes were fundamentally different from those of the 

 air-inhabiting forms and that they probably more nearly 

 resembled organs of taste than any other receptor pos- 

 sessed by the higher animals. This opinion was based 

 upon theoretic considerations rather than upon any par- 

 ticular observation or test. 



But before these views had been expressed by Nagel, 

 a certain amount of experimental evidence concerning 

 olf action in fishes had been gathered. This was prelimi- 

 nary in character and inconclusive, but it nevertheless 

 paved the way for further advance. Thus the observa- 

 tion of Aronsohn (1884a), that goldfish, which ordinarily 

 will eat ant pupae with avidity, would not take these pupae 

 after they had been smeared with a little oil of cloves, 

 is not final evidence that the fish scented the oil, for 

 it is entirely possible that this oil irritated the skin of the 



