MILL CREEK. 211 



and then, unable to remain afloat, they come to rest upon 

 the bottom of the stream. If the bottom be of sand or 

 sand} 7 mud, the imprint of their ventral and pectoral fins 

 can be seen ; and always when they make a plunge for 

 some minute crustacean or insect, the print of their pig- 

 like snout can always be seen. 



I began a series of simple experiments by dropping 

 coarse grains of sand into the water directly in front of 

 them. These they seemed to recognize as such, and did 

 not heed them beyond approaching a little nearer to the 

 spot where the sand-grains lodged, or turning so as to 

 directly face the spot if the grains fell a little to one 

 side or the other. Continuing to watch the darters 

 closely, I found that this was a customary movement 

 with reference to all animate objects upon which they 

 fed. The first motion a water -mite or cypris might 

 make caused the darter to assume a make-ready attitude ;. 

 the next assured the fish that the object moving was 

 alive, and immediately it was pounced upon. 



Having a quantity of mustard-seed shot with me, I 

 conceived the plan of controlling the movements of 

 these fishes by dropping single grains of it in front of 

 one or two near together, and while they were waiting 

 for the grains to move a second time, bring up others 

 that were behind. Experience showed that theory and 

 practice were very different, in this as in many another 

 matter, but finally I got seven in a row, and by placing 

 a row of shot-grains on a stick and letting them fall over- 

 board together, I kept these seven together and without 

 confusion ; they faced to the right or left or advanced, in 

 obedience to my order expressed by the dropping of the 



