222 FURTHER EXPERIMENTS. 



lumiDous intensity las been shown to be uniform on 

 each side of the line having the mean wave-length, i.e. 

 a little above the line D in the yellowish green of the 

 spectrum. 



I then took a long shallow trough in which were 

 a number of Daphnias, and placed it so that the 

 centre of the trough was at the brightest part of 

 the spectrum, a little, however, if anything, towards 

 the green end. After scattering the Daphnias equably 

 I left them for five minutes, and. then put a piece of 

 blackened cardboard over the brightest part. After 

 five minutes more, there were at the green end, 410; 

 in the dark, 14 ; at the red end, 76. Here the two 

 ends of the trough were equally illuminated ; but 

 the preference for the green over the red side was very 

 marked. 



I then took five porcelain vessels, seven and a half 

 inches long, two and a half broad, and one deep, and 

 in each I put water containing fifty Daphnias. One 

 half of the water I left uncovered ; the other half I 

 covered respectively with an opaque porcelain plate, a 

 solution of aurine (bright yellow), of chlorate of copper 

 (bright green), a piece of red glass, and a piece of blue 

 glass. Every half-hour I counted the Daphnias in 

 each half of every vessel, and then transposed the 

 coverings, so that the half which had been covered was 

 left exposed, and vice versa. I also changed the Daph- 

 nias from time to time. 



Here, then, in each case the Daphnias had a choice 

 betw^een two kinds of light. It seemed to me that this 

 would be a crucial test, because in every case the 

 colored media act by cutting o^T certain rays. Thus 

 the aurine owes its yellow color to the fact that it cuts 



