6i 



One of these fishes contained still a large remnant of the egg-sac, 

 showing that the propensity to capture prey must antedate the 

 sensation of hunger. On the 25th the fourteenth and last 

 rem.aining fish captured its Cyclops, and was itself sacrificed in 

 turn. As an indication of the efficiency of the raptatorial teeth, 

 it may be worth while to note that I saw one of the smallest 

 fishes make a spring at a Cyclops, catch it, give three or four 

 violent wriggles, and drop it dead to the bottom of the tank. 



As a general statement of the result of the observations 

 made on these fourteen fishes, we may say that eight of them 

 ate a single Cyclops each, that one took two, and another three 

 of the same, that one took a single Canthocamptus, that two 

 specimens captured two each of this genus, and that finally, a 

 single fish ate Cyclops and Canthocamptus both. The final 

 conclusion was a highly probable inference that the smallest 

 Entomostraca occurring in the lake would prove to be the 

 natural food of the species. 



In order to test this conclusion with precision, I arranged a 

 similar experiment on a larger scale, and under more natural 

 conditions. Through the generosity of the Exposition Com- 

 pany, of Chicago, I was allowed the use of one of the large 

 aquarium tanks in the Exposition building, on the lake shore, 

 and by the repeated kindness of Mr. Clarke, of Northville, 

 Michigan, I was furnished with a much larger number of living 

 whitcfish. Five thousand fry were shipped to me in a can of 

 water, but through unfortunate delays in changing cars at inter- 

 mediate points, about two-thirds of these were dead when they 

 reached my hands. Those living were immediately transferred 

 to the tank, through which the water, taken from the city pipes, 

 had already been allowed to run for several hours. As this 

 water is derived from Lake Michigan at a distance of two miles 

 from the shore, and had at this time the exact temperature of 

 the open lake, the conditions for experiment were as favorable 

 as artificial arrangements could well be made. 



Sending a man with a towing-net out upon the lake with a 

 boat, or upon the remotest breakwaters, immense numbers of 

 all organic objects in the waters were easily obtained. After 

 enclosing the exit of the tank with a fine wire screen, to prevent 



