1 84 BULLETIN OF THE BUREAU OF FISHERIES. 



but that they sometimes feed on schools of small fry near the surface, pursuing and 

 capturing them as other fish do. 



These reports were confirmed by observations in the laboratory. It was ordinarily 

 found that the latter method prevailed only in case the fish had not been fed for a few 

 days and were presumably hungry. I am of the opinion that this method of feeding 

 was much more common in fish that had been in the laboratory a few weeks than in those 

 recently brought in. It was observed that flounders, especially Ancylopsetta, frequently 

 stalk their prey. Specimens were repeatedly seen creeping on the bottom with their 

 fins, toward a minnow hovering above them, just out of reach, so slowly and smoothly 

 that their motion was scarcely perceptible and, when they got within striking distance, 

 spring suddenly upward and forward, capturing their prey. They rarely, if ever, take 

 dead minnows or pieces of meat lying on the bottom motionless, but do not hesitate to 

 to take them, even when badly decayed, if they are in motion. In fact, they appear to 

 take decayed food quite as readily as fresh. Thus it would appear that the selection of 

 food depends very largely, if not entirely, upon vision. 



After this work was completed I found a blind specimen which learned to capture 

 minnows, showing that vision can be dispensed with in the process of feeding. This speci- 

 men lost both eyes on September i . From this time until September 1 7 it ate nothing 

 and gave no feeding reaction. On this day, however, it was seen several times to snap 

 at minnows that chanced to come near. On September 19, 15 minnows were put into 

 a rather small inclosure with the blind specimen. In a few minutes it captured and 

 swallowed one of them. After this it fed regularly on living minnows, but it was 

 impossible to make it take dead ones. I am of the opinion that the presence of the 

 minnows was detected only by contact, although I was not able fully to establish this 

 point. 



Owing to their peculiar habits and characteristics and their hardiness, Paralichthys 

 and Aticylopseila are among the most favorable of fishes for experimental work, espe- 

 cially of the sort described in the following pages. 



ADAPTIVE CHANGES IN SHADE, COLOR, AND PATTERN. 



NATURAL BACKGROUND. 



It is well known that Paralichthys and Ancylopsetta usually resemble somewhat 

 the bottom on which they lie — those taken on dark bottoms are usually dark and those 

 taken on light bottoms are usually light. Further than this, however, practically nothing 

 is known. Little or nothing is known concerning the nature, the degree, and the function 

 of this resemblance. In trying to solve some of the problems just referred to, obser- 

 vations were made under various conditions on specimens kept on three essentially 

 different sorts of bottoms: (o) Natural background; (6) black and white background, 

 artificial; and (c) colored background, artificial. 



In studying the adaptive changes on natural backgrounds, some specimens were 

 placed directly on, and others in glass dishes over, fine gray sand; fine black sand; 

 medium, and very coarse yellowish sand, consisting largely of small shells and frag- 

 ments of larger ones; broken oyster shells; coarse black cinders; and bluish-gray 

 pebbles. 



