THE COLORATION OF REEF FISHES 541 



toward the light. They also justify the appeal to ignorance 

 in the case of the remora's unusual coloration, which may not 

 be inconsistent with the development of countershading through 

 natural selection. 



That illumination a,lone cannot account for the distribution 

 of their dark pigment, is shown with additional interesting 

 implications by many countershaded animals, perhaps by all 

 that wear distinct patterns. There are fishes, for example, 

 with alternating light and dark horizontal or vertical stripes. 

 In the former case the dark elements may be so graded that the 

 countershading is largely applied in a series of discrete patches, 

 rather than in a mass appearing unbroken to the naked eye. 



The bass (Roccus lineatus) shows the pattern in question. 

 It is an interesting detail that some of its dark stripes decrease 

 in intensity posteriorly, as they pass more and more into the 

 shadow of the caudal peduncle. The fish might be described 

 with some truth, as dark in color, perfectly countershaded, and 

 marked with a series of horizontal light stripes. This pattern 

 is constant within the species, and hence inherited; or it might 

 be urged that what is inherited is a difference in material in the 

 regions covered by the two sets of stripes, so that one is darkened 

 by sunlight while the other is not. In any case even an indirect 

 effect of light in controlling countershading through the inherit- 

 ance of acquired characters is excluded; for there is obviously 

 a fundamental internal condition upon which the development 

 of the whole system of pigmentation depends, and it cannot be 

 so simply explained. 



But if countershading is no necessary effect of the mere pres- 

 ence of pigment (Echeneis), and is neither directly nor indirectly 

 induced incidentally by external factors (Brachyura and Roccus) , 

 and if, in addition, this system of pigmentation is positively 

 known to contribute most decidedly to the inconspicuousness 

 of its possessors under certain normal conditions, it would seem 

 that if any characters be maintained by natural selection this 

 should be one. Therefore, from a consideration of all the facts, 

 unless indeed one assumes that the countershading is vestigial 

 and antedates the appearance of bright colors, to the conspicu- 



