4 VICTOR E. SHELFORD AND W. C. ALLEE 



side length 110 cm., were painted dull gray and covered with 

 yellow sand while still adhesive. Water was allowed to flow in 

 at both ends at the same rate (usually 600 cc. per minute) through 

 tees made from iron pipe, the cross bar of which contained a 

 number of small holes. The cross bars of the tees rested on the 

 bottoms of the tanks behind the screens. The drains were 

 transverse tubes with their lower sides made of screen, located 

 near the top and opening outside the boxes. The water flowed 

 in at the ends and drifted toward the center at the top and 

 flowed out through the drain. We found no evidence that 

 fishes react to the slight current thus produced. Since each half 

 of the tanks held about twelve liters, it required twenty minutes 

 to fill them or to replace all the water in one of the halves. 



Both tanks were enclosed under a black hood, side by side as 

 shown in the plan (Fig. 1) and were placed about ten centi- 

 meters apart. Two four candle power incandescent lights were 

 fixed above the center of the two halves, i.e., above a point 

 midway between the screen partition and the center drain. The 

 light was thirty centimeters above the surface of the water, 

 which was ten centimeters deep. The lights above a given tank 

 illuminated the outer wall of the other tank (see Fig. 1), while 

 the inner wall of the same tank cast a shadow throughout its 

 entire length. The two tanks were identical longitudinally, but 

 the shadow was reversed with respect to points of the compass. 



The room was darkened during the experiments, which were 

 observed through openings in the hood above the lights. Fishes 

 do not usually note objects separated from them by a light. 

 The fishes not accustomed to aquaria sometimes showed fright 

 and behaved erratically when first put into the tanks, but all 

 such experiments were thrown out. The main stock of fishes 

 was kept in the laboratory in glass-sided aquaria during the period 

 of experimentation. In this way they became accustomed to 

 an aquarium, to the presence of moving objects, and to variously 

 placed lights. 



The purpose of the experiments was to test the reactions of the 

 fishes to a difference in the water in the ends of one of the experi- 

 mental tanks. Water differing as little as possible from that in 

 which the fishes usually live was introduced at both ends of the 

 other tank in most experiments (control) . ' Treated water from 

 the device already described (Shelford and Allee, '13, p. 214), was 



