BUNSEN-ROSCOE LAW 



91 



a dark room and the lights A and B are enclosed in a box. At the 

 beginning of the experiments the pipette is filled with a dense suspension 

 of larvae in sea water and then put with its point touching the bottom 

 of the dish. The animals flow out in a fine stream which is narrow at 

 the opening of the pipette and widens slightly, owing probably to the 

 negative stereotropism of the animals. 

 A glass plate (Fig. 32) Mkl, which has 

 a strong red line no and a fine parallel 

 line pq (cut with a diamond), is then put 

 on the dish and so adjusted that pq is in 

 the middle of the stream fg of the ani- 

 mals. Then the angle a which pq makes 

 with the perpendicular from A on ab is 

 measured. This perpendicular is marked 

 in the form of a red line on the black 

 base on which the glass vessel abed 

 stands. The angle a is measured with a 

 goniometer. When the lights are equal 

 in intensity should be 45 ; if the two lights have different intensities 

 and if A be the stronger light a should become smaller with increas- 

 ing difference in intensity. The individual measurements vary com- 

 paratively little, as long as the difference in the intensity of the two 

 lights is not too great; for this reason our observations do not go 

 beyond a wider ratio of the two lights than 10 : 1, though 4 : 1 is 

 perhaps the limit for good results. Table VII gives the results. A 

 is always the stronger light. Each table is the average of from 40 to 60 

 individual observations, each being the average of the path of many 

 thousands of animals. 



TABLE VII 



Value of a for different ratios of intensities of 

 the two lights 



FIG. 32. 



In the next series of experiments an opaque rotating disk with one 

 sector cut out was placed before light B. In one set of experiments the 

 sector cut out was 90. The rate of rotation (by an electric motor) 

 was 1,500 to 2,500 revolutions per minute. The other light was constant 

 and its distance was chosen on the assumption of the validity of the 

 Bunsen-Roscoe law for these cases. Thus when the two lights without 

 sector were equal at a given distance of A, by putting 90 sector before 



