DEPARTMEINT 01^ MARINE) BIOLOGY- II7 



I consider his experiments to be the most conchisive and his results the 

 most uniform and the clearest yet attained by any naturalist in experiments 

 upon the subject of warning coloration. No account of his experiments can 

 give an adequate idea of the dramatic definiteness of his results. As the 

 theory of "warning coloration" has been one of the cornerstones of the 

 theory of natural selection, it is unnecessary to call further attention to the 

 importance of Professor Reighard's results. His preliminary report is as 

 follows : 



An experimental Study of Color Discrimination, Association, and Memory 

 in the Gray Snapper, Lutianus griseiis (Linnceus) , and of Warning Colora- 

 tion in Coral-reef Fishes, by Jacob Reighard, University of Michigan. 



The gray snapper, the commonest predaceous fish of the coral reefs at Tortu- 

 gas, Florida, was made the subject of experim.ents intended to test the follow- 

 ing underlying assumptions of the theory of warning color as applied to coral- 

 reef fishes : ( i ) That there is "a general association of bright colors with 

 poisonous or dangerous qualities" (Beddard), so that certain colors in them- 

 selves warn the predaceous reef fish and are avoided by them; (2) that the 

 predaceous species discriminate colors; (3) that they are able to form asso- 

 ciations between certain colors and disagreeable qualities; (4) to retain 

 these associations (memory). 



The commonest food of the gray snapper, in May or June, is a small 

 silvery fish, Atherina laticeps Poey. The experiments consisted in feed- 

 ing to the gray snappers dead Atherinas, to which different colors had been 

 given by the use of dyes. The tests were made upon a colony of 150 snap- 

 pers, entirely at liberty in their natural environment under wholly normal 

 conditions. 



(i) The snappers took without hesitation Atherinas when colored red 

 yellow, green, blue, or violet. 



(2) By feeding the snappers on Atherinas of one color until they were 

 familiar with it, and then allowing them a choice between that color and 

 another and unfamiliar color, they were found to discriminate white from 

 blue, blue from red, and blue from yellow. Care was taken to eliminate 

 errors due to differences in brightness-values, size, position, and chemical 

 qualities (odors or taste). The colors were impure. 



(3) Red Atherinas were rendered unpalatable by attaching to each a 

 small portion of a nettle-cell bearing tentacle of the medusa Cassiopea; 238 

 tentacled Atherinas were fed to about 150 snappers in i hour and 35 minutes, 

 divided into two nearly equal periods, separated by an interval of 3 days ; 

 60 of the tentacled Atherinas (including the last 16) were refused. That the 

 association thus formed was between disagreeable qualities and the color 

 red, and not between disagreeable qualities and the changed form of the 

 red Atherinas due to the attached tentacles, was shown by the following: 



(a) That red Atherinas without tentacles Vv^ere subsequently refused, while 



(b) white Atherinas with tentacles were readily taken. Twenty days after 

 the red association had been formed, red Atherinas without tentacles were 

 still absolutely refused by all the snappers in the colony, although white and 

 blue were readily taken (retention of a warning association artificially estab- 

 lished). 



