COLORATION OF FISHES LOISTGLEY. 479 



selves in the barren waste. A bit of food will often bring crabs 

 out of hiding, too, to scuttle over the bottom, whose characteristic 

 features are faithfully repeated in their own markings. But if one 

 may judge from their behavior in tanks, the day is not their pre- 

 ferred time for roving, and after a little they almost invariably 

 give what in any animal not covered by a firm exoskeleton would 

 be a shrug or two, and scraping and scratching with their hind 

 legs go down backwards out of sight. 



Thus one finds peculiar structures discharging odd functions, and 

 so interpreting themselves. Proof is obtained, too, that reactions 

 are normal which one sees from time to time among one's captive 

 animals. But often one observes incidents which remain incompre- 

 hensible, as when two yellow grunts (Haemulmi sciurus) approach 

 one another slowly, snout to snout, open their mouths to the limit of 

 their gape, and gaze, as it seems, for several seconds, as if in rapt 

 attention, each at the patch of bright red in the buccal cavity of the 

 other. Nor is it clearer why from tiny holes in dead pieces of coral 

 a small unidentified species of fish with an enormous dorsal fin 

 should protrude half its body and rapidly and repeatedly elevate and 

 depress its great banner, while another seems to respond in kind to 

 the signal. 



One of the most striking things to be observed at favorable places 

 is the abundance of fishes. Not only are hundreds* of individuals to 

 be seen without changing one's station, but as many as 50 species 

 have been noted at one spot within an hour. Since all are so near 

 that almost without exception detailed comparison of appearance 

 and behavior is possible even to the unaided eye, it is obvious that 

 comparable advantages may be enjoyed rarely, if ever, in respect to 

 other groups of animals. 



The places at which species so abound are ideal points at which 

 to learn to distinguish the different forms, and to become familiar 

 with their appearance in life, which is commonly very different 

 from that of dead specimens, even when freshly taken. For ad- 

 vanced students of the behavior of the species concerned such places 

 possess additional advantages, for it is sometimes possible to see 

 there side by side types not occuring elsewhere together, and to 

 verify tentative conclusions regarding their specific difference in 

 habits which may rest upon observations made upon them sepa- 

 rately. As points, however, at which study of behavior should be 

 begun, and particularly as stations at which by study one might 

 hope to determine the significance of the animals' colors and pat- 

 terns, few places could be less propitious. They are average, or 

 typical, environments for a comparatively small number of the 

 species present ; and it is in its typic*al environment that one of 

 these creatures should be seen in order that the significance of its 



