588 W. H. LONGLEY 



The reader may justly challenge the assumption that the two 

 families cited fairly represent fishes of high color. No exception 

 may be taken, however, to an argument based upon analysis 

 of the facts regarding a list of species compiled by one familiar 

 with them, convinced of their conspicuousness, and unaware 

 of the use to which his product is to be applied. Such a list 

 taken from Reighard's ('08) Field Study of Warning Colora- 

 tion, appears in table 5. In the second column its author's 

 comments upon the included species are given in full. The 

 remarks in the third column require no explanation. 



Warning significance has been imputed to the colors of animals 

 for many reasons, su€h as assumed offensiveness in taste or odor, 

 or possession of poisonous qualities, or unpalatability depending 

 upon scaliness, hairiness or the presence of some form of sting. 

 But no claim has been made that one sort of disagreeable attri- 

 bute rather than another may be more fitly associated with types 

 of warning coloration. It is apparent, therefore, that an addi- 

 tional method of testing whether color combinations of dis- 

 tinctive quality are correlated with distastefulness or immunity 

 lies in the examination of species which possess the one sort of 

 defensive attribute whose occurrence may be determined with 

 approximate accuracy, i.e., organs capable of inflicting painful 

 bodily injury. 



The fishes which may be fairly included in this group are listed 

 with their sahent characteristics in the following table. All are 

 countershaded. 



Fishes of high color are freely eaten; 'coUvSpicuous' and speci- 

 ally protected types reduce their visibility as commonly as others 

 by adapting their hues to their surroundings ; and, in pigmentation, 

 the species enumerated in tables 5 and 6, do not differ from groups 

 of the same size selected" at random. Hence, in this class of 

 animals there is not even secondary correlation of bright color 

 with special means of defence, and, in so far as appears from the 

 evidence, the hypothesis of obliterative coloration remains in 

 undisputed possession of the field. 



This situation should surprise no one familiar with the litera- 

 ture of animal coloration, since a student of the subject must be 



