106 A. S. PEARSE 



of colors and color patterns. No indubitable cases of rapid 

 color changes have been reported, and McCook, one of the fore- 

 most students of spiders, points out ('89-'93, vol. 2, p. 325) 

 that a single species may show a number of striking color vari- 

 eties in the same habitat; the coloration is apparently not 

 accurately adjusted to a particular background. 



The insects show striking adaptation to aerial life, and also 

 to a great diversity of habitats. Numerous colors and color 

 patterns have been developed along rather definite lines (Mayer, 

 '97; Tower, '03). A host of insects show protective resemblance 

 and a few have been observed to undergo slow changes w^hich 

 make them more nearly resemble their surroundings (Poulton, 

 '88; Davenport, '03). Kellogg ('05) in his work on American 

 insects says (p. 600) the colors of insects are "fixed by the time 

 they reach the adult stage," but a striking diurnal color change 

 has recently (Schleip, '10) been demonstrated in Dixippus 

 morosus. In this case there is a diurnal migration of pigment 

 granules in a single layer of syncitial hypodermal cells. The chief 

 factor which brings about this migration is the presence or absence 

 of light, but the changes have a strongly developed diurnal 

 periodicity and continue for as much as seventy-eight days 

 in the dark. 



Color discrimination.* — Bateson ('89) and Merejowski ('81) 

 maintained that there is no color discrimination * manifested 

 by the reactions of Crustacea but Minkiewicz ('07) takes an 

 opposite view. On account of the results of the experiments 

 described in this paper (p. 88) the w^riter is disposed to agree 

 with the results of the earlier investigators. The Peckhams 

 ('87, '87a) firmly believe that spiders can discriminate colors, 

 and Lubbock ('79), Lovell ('10) and Turner ('10) are of the 

 same opinion in regard to hymenopterous insects. However, 

 the whole question of color discrimination will bear further 

 investigation. The present evidence is fragmentary, and some 

 of it by no means conclusive. 



Protective coloration. — Di Cesnola ('04) has demonstrated 

 that protective coloration may preserve an insect from the attacks 

 of its enemies, and there is little reason to doubt that protective 



* Color "discrimination" is not intended to assume that arthropods see color 

 as we see it, but only that they may be able to recognize a difference between colors 

 or intensities of color. 



