560 W. H. LONGLEY 



Hence, if large numbers of a species of such fishes can be defi- 

 nitely located by day, one may have almost complete confidence 

 that an inconsiderable proportion of the whole is in other sur- 

 roundings, searching for food and subject to exposure which 

 might result in adaptation to an environment differing from that 

 in which the great majority lies. 



The importance of knowing the exact range of the animals 

 studied should be self-evident, for if one undertakes to as- 

 certain whether their colors repeat those of their normal habitat, 

 it would seem to be axiomatic, that the nature of the places over 

 or through which they wander should be accurately determined. 

 But some writers take it for granted that the observation of 

 individuals of any number of species in what we may loosely 

 describe as the same place at the same time, proves that they 

 have the same habit. This uncritical attitude is manifested in 

 extreme form by Dewar and Finn ('09, p. 88-89), who argue 

 that because the moose, Greenland whale, and a farther mis- 

 cellaneous assortment of mammals, including seals, narwhals 

 and musk-oxen, are colored, the prevailing whiteness of the 

 arctic fauna has been greatly exaggerated, and that the common 

 impression that adaptive coloration is dominant in that region 

 js misleading. This means that to these authors the difference 

 in habit and habitat between terrestrial animals ranging south 

 to the latitude of northern Spain, aquatic animals and polar 

 bears, is so inconsiderable, that their difference in color demon- 

 strates its essential lack of correlation ^^ith habit and its slight 

 biological significance. 



It is of importance to know at what level fishes commonly swmi, 

 for, as they rise or fall, the proportion of the finny population 

 in whose sight they appear against a background of blank water, 

 or against the variegated bottom or its vertical excrescences, is 

 continually changing. 



As a first step toward the separation of the fish fauna of the 

 Tortugas into ecological groups, the time of feeding has been 

 conclusively determined for a number of species by examination 

 of their stomach contents at different times during the day. The 

 record appears in the following table. A few observations are 



