393 



to the polluting substances. Fishes turn away from dangerous sub- 

 stances which are normally found in their usual environment, but 

 with strange and unusual substances such as are thrown into streams 

 by gas-works and other industrial plants, they frequently enter and 

 follow up to points where the concentrations are fatal, or fail to recog- 

 nize the dangerous substance at all and often stay in it until they are 

 intoxicated and finally die there. (Chart II, graphs 8-ii; Chart V, 

 graph 60. ) 



Conditions and Methods of Study. 



The experiments were performed in a gradient tank (N), figure 

 I. The tank used in these experiments was 122.3 cm. long, 15 cm. 

 wide, 13 cm. deep. The front wall was of plate glass and a plate 

 glass top was used at times. Water of two kinds, normal and polluted, 

 was used in the experiments. One kind was allowed to flow into one 

 end at a definite rate and another kind into the other end at the same 

 rate. It flowed out at the middle at the top and at the bottom so 

 that the two kinds of water met at the center. The outflow at the 

 center did not of course prevent the mixing of the two kinds of water 

 and thus the middle section, ecjual to one half or one third of the 

 tank was a gradient between two kinds of water. The water entered 

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

 the cross-bars of which contained a number of small holes. The 

 cross-bars of the tees were at the center of the ends of the tank behind 

 screens. The drain openings were located at the center near the top 

 and in the bottom. The outer openings of the drain tubes were at 

 the level of the water in the tank. We found no evidence that fishes 

 reacted to the slight current produced by the water flowing in at the 

 ends and drifting toward the center and out through the drains. 

 Since each half of the tank held about 9 liters, it required 15 minutes 

 to fill it or to replace all of the water in one of the halves. The tank 

 was enclosed under a dark hood. Two electric lights were fixed in 

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

 midway between the screen partition and the center drain. The light 

 was 15-20 cm. above the surface of the water which was 13 cm. deep. 

 The experiments were observed through openings in the hood above 

 the lights or through the glass side late at night. Fishes do not usual- 

 ly note objects separated from them by a light. 



Water differing as little as possible from that in which the fishes 

 usually live was used for control readings. Controls were observed 

 and conditions in the two ends of these were the same either because 

 the water introduced at the two ends was alike or because no water 



