FEEDING-HABITS OF FISB AND BIRDS. 49 



A general survey of the feeding-habits of trout, where food is plentiful, 

 brings out three distinct facts: (1) the fish prefer mixed diet; (2) fish feed 

 upon one food at a time; (3) fish have a special appetite as apart from a 

 general appetite. 



In waters where food is scarce the fish only grow to a small size. Under 

 these conditions their feeding-habits are entirely different : they bike all 

 food that comes within their reach, they are forced by general hunger to eat 

 that which is- supplied, they have no opportunity of exhibiting preference. 

 Their method of taking food is shown on examination of their stomach 

 contents ; the following are two typical examples: — 



"... Aug. 21, 1905, E. Lambourne, Great Shelford, part of river where food is 

 scarce, stomach contents were 2 house flies, 2 bluebottles, 1 wasp, numerous reed 

 smuts and other small Diptera, 2 shrimps, 1 snail. 



" Sept. 10, 1910, Pennel burn, North Wales, stomach contents were 1 grasshopper, 

 2 wasps, 1 black beetle, several small species of Coleoptera, 1 large Ichneumon fly, 

 1 daddy-longlegs, 2 hoppers, 6 house flies, 2 caddises, numerous small Diptera, 

 numerous red and black ants male and female, several duns and spinners (sub-imago 

 and imago of Ephemeridas)." 



Had observations been confined to such fish, conclusions would have been 

 drawn that trout appear to have no preference for particular foods. 



Evidence that Trout recognize a Difference in Palatability betioeen 



different Food. 



In order to prove that fish do discriminate, a long series of observations 

 must show that, although the fish take both foods, nevertheless, when they 

 have the choice of both, they much more often take one than the other. 

 During May the trout have often choice of two floating insects, the sub- 

 imago of Baetis vernuK, Curtis (the "Olive Dun"), and the imago of Bibio 

 johannis, Meig. (the "Black Gnat"). It will be seen that some fish are 

 takins both insects, others the Bibio only, and others the dun. Each fish 

 thus falls into one of the three classes : by observing a number of fish, 

 distribution of feeding-habit is obtained, as in the following observation : — 



May 16, 1910, "Black gnats and olive duns on the water from 10.0 to 11.30 a.m. : 

 of 20 feeding fish, 11 were taking duns only, 7 black gnats only, and 2 were taking 

 both flies." 



Similar observations made on many days and on different waters show that 

 the dun, Baetis vermis, is taken more frequently than the gnat, Bibio 

 johannis. The number of fish taking both insects was found to depend on 

 the stream on which the observations were made ; where the food-supply is 



LINN. JOUKN. — ZOOLOGY, VOL. XXXIV. 4 



