20 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



first received from the spring. These fry were fed daily with gammari 

 in minute fragments. 



Of the first 90, 4 showed signs of food, 3 in the form of a few grains 

 of dirt in the intestine, and the fourth a fragment of the crust of gam- 

 marus; in the second lot (111), 17 had eaten. I discovered 9 of these^ 

 and found only fragments of Gammarus. Twelve out of the 90 in the 

 third lot had lately taken food ; 4 had eaten fragments of Gammarus ; 

 7, small particles of the leaves and stems of vascular plants ; 2, larvre 

 of Oulicidw and 1 , a Cypris, entire. In the fourth and last lot were 39 

 specimens ; 14 had taken food, 5 in such minute quantity that I did not 

 dissect them ; Gammarus fragments were found in 4 ; larvte of gnats 

 (one Chironomtis) in 3; and a minute vegetable fragment, a Cyclops, a 

 Cypris, and some undetermined entomostracan each in 1. 



To recapitulate : The specimens from the spring ate only vegetable 

 food, but could apparently get nothing else. There was an abundance 

 of vegetation about them, but only 3^ per cent, of them took food at 

 all. They apparently died for want of animal food. 



Fourteen per cent, of those from the hatching-house had taken food 

 (47 out of 340). Of the 35 dissected, 18 had eaten fragments of Gam- 

 marus; 5, minute insect larva^; 4, Entomostraca ; and 8, small i^articles 

 of vegetation. Few of these died, and those not used in the investiga- 

 tion were living a few days since. 



A structural detail observed throws light on the question. With the 

 complete disapi)earance of the egg-sac (and not before) two small, but 

 stout, sharp, recurved teeth are developed on each side of the lower 

 jaw. These are well adapted to the capture of a minute living prey, 

 and apparently could not h ave any other use. I am very well satisfied, 

 on the whole, that the earliest food of this fish will prove to be Living 

 Entomostraca^ with probably some admixture of filamentous algfe. As 

 the gill-rakers are not developed at this early age, I don't see how 

 any smaller forms could be separated from the water, except accident- 

 ally. 



The Gammarus " hash" makes evidently a very good substiute for the 

 Entomostraca. The thicker crust and the necessary loss of much of the 

 soft parts, in pulverizing the animals, make these less nourishing than 

 the natural food — a fact likewise indicated by the greater abundance of 

 the orange oil globules derived from Crustacea^ in the intestines of those 

 which had fed on Entomostraca. 



I shall endeavor to collect some of the fry from Northern Lake, Mich- 

 igan, next month, if I can get away, with a view to putting these con- 

 clusions to a more definite test. 



Very respectfully, your obedient servant. 



S. A. FOEBES. 



Professor S. F. Baird. 



State Laboratory of Natural History, 



Normal, Illinois, 



March 29, 1881. 



