6O C. J. CONNOLLY. 



subject to the same light energy from source, but obviously we 

 cannot conclude from this, that quality of light and not brightness 

 of background, determined the effect. These tests of relatively 

 short duration, seemed rather to indicate that brightness and not 

 color or wave-length determined the shade. Further tests were 

 required to settle the point. 



In the next series of experiments, the effect of different mono- 

 chromatic lights on fishes placed in white and black lined boxes 

 was tried. For this purpose Wratten filters were used. These 

 were of the larger size, namely three inches square, the smaller 

 size which were first used not supplying sufficient light surface at 

 the selected distance to illuminate fully the boxes. The filters 

 selected were: Yellow (K, No. 6) Red (F, No. 29) Green (B. 

 No. 68) and Blue (No. 45). approximately corresponding to the 

 four colored backgrounds used in the preceding tests. The 

 Mazda lamps were placed in boxes covered with black paper ex- 

 cept the space occupied by the filter set in the middle of the 

 front side. In front of the filter about 60 cm. distant, two boxes 

 were set up, one being lined with dull white and the other with 

 dull black paper, each being so placed as to receive equal amounts 

 of light. All shadows were avoided. 



Screened off from this apparatus, a similar one was set up 

 in the same large dark room, so that the effect of each of two 

 monochromatic lights on two different backgrounds could be com- 

 pared under identical conditions. Each test was of 24-hours 

 duration and comparisons were made on a white ground in diffuse 

 daylight. 



When the specimens placed in the white lined boxes were 

 compared, they showed the same sequence in shade as in the previ- 

 ous set of experiments, namely from light to dark; yellow, red, 

 green, and blue, though due to less intensity of light by use of 

 filters, the effect was less pronounced than with the white light 

 on variously colored backgrounds. When the effect of the same 

 monochromatic lights on specimens in the white-lined box was 

 compared to the effect on those in the black-lined box. it was ob- 

 served that in each case the latter werq darker in shade than 

 the former. And moreover little or no difference could be ob- 

 served in the degree of darkness shown by the specimens in the 



