BEHAVIOR OF VERTEBRATES 453 



mechanism to the terminations of the spinal nerves in the epi- 

 dermis and assures that the common sense is independent of 

 the taste buds which are scattered over the suiface of the body 

 and are innervated by the lateral accessories. 



By an examination of the effects of various substances Parker 

 concludes that the ions stimulating the common chemical re- 

 ceptors are the same as those stimulating the human taste buds, 

 and that taste and the common chemical sense are closely re- 

 lated in the vertebrate both with respect to their sensitivity 

 and the nature of the stimuli received. Smell, on the contrary, 

 he finds in fish, as in the land vertebrates, to be much more 

 sensitive to weak stimuli, and to serve, probably, as a distance 

 receptor. 



The author differs from Herrick and Sheldon in holding that 

 the olfactory sense presents the primitive form from which the 

 others have been derived. He is led to this view chiefly by 

 the similarity of the olfactory neurone to sensory cells found 

 in invertebrates. 



In a brief preliminary note Shelford and Allee (2r) describe 

 their apparatus for the study of reactions of fish to solutions 

 of gases and solids and review the results obtained by its 

 use. The device consists of a rectangular tank with an intake 

 at each end and an outlet in the middle. Waters differ- 

 ing in their dissolved contents can be admitted at the two ends, 

 thus forming, in the middle of the tank, a gradient to which the 

 fish react. Eight species of fish were studied. They were 

 found to give indefinite or slightly negative reactions to changes 

 in oxygen pressure and to slight reductions in the salt content. 

 There was no evidence of reaction to nitrogen. All the fishes 

 avoided water containing a per cent, of carbon dioxide greater 

 than that to which they were accustomed (to increases of from 

 five to sixty cc. per liter), and when an increase in carbon diox- 

 ide was accompanied by a decrease .in oxygen the negative 

 reactions became very pronounced. 



The authors conclude that, except in cases of strongly alka- 

 line waters, the content of carbon dioxide furnishes the best 

 single index to the suitability of a water for fishes. The experi- 

 ments seem to have been well controlled by the use of a second 

 tank in which only one quality of water was used. 



