220 BULLETIN OF THE BUREAU OF FISHERIES. 



FUSION RATE OF IMAGES ON THE RETINA. 



It is well known that if a disk containing alternate black and white sectors is rotated 

 rapidly enough the sectors become invisible and the disk appears uniformly gray. This 

 is due to the fusion of the images on the retina. If the images on the retina in fishes fuse 

 somewhat as they do in man, a flounder over such a disk, at a certain rate of rotation, 

 ought to become uniformly gray. And if the fusion rate is the same for fishes as it is 

 for man the flounder ought to become, when the disk is rotated fast enough to appear 

 gray to the human eye, as uniformly gray as it does on a stationary gray background. 

 The results of the following experiments show that this actually occurs. 



In these experiments a disk, 25 cm. in diameter, was fastened to a vertical shaft 

 and so arranged in a wooden vessel that it rotated immediately under a stationary hori- 

 zontal glass plate. The shaft extended down through the bottom of the vessel, where 

 a pulley was attached to it. The disk was divided into 3.2 equal sectors, half of which 

 were painted with white enamel and the rest, alternating with these, with black engine 

 paint. The vessel contained flowing water extending to a point a few centimeters above 

 the glass plate. The specimens to be tested were put on this plate in a glass inclosure 

 of such a form and so arranged that the head was held continuously over the center of 

 the disk. All air bubbles were carefully removed from under the plate. The disk was 

 turned by means of a water motor, the speed of which could be altered as desired. 



Four individuals were tested, three P. albiguitus, 16, 20, and 27 cm. long, respect- 

 ively, and one P. dentatus, 14 cm. long. The specimen 20 cm. long was more thoroughly 

 studied than any of the others. This specimen had been changed frequently during 

 several weeks from a white to a black aquarium and vice versa until adaptation to these 

 shades occured moderately rapidly. It was kept over the disk continuously, with the 

 exception of a few short intervals, for five days. During this period the rate of rota- 

 tion of the disk was changed at intervals varying from a few moments to an hour or 

 more, and the effect on the pattern of the skin was noted and recorded. Without going 

 into details regarding these records the results may be summarized as follows : 



When the disk was not rotating the pattern in the skin became very conspicuous. 

 The three ocelli became black and numerous other black spots of about the same size 

 as the ocelli appeared. Numerous larger irregular light and dark patches also appeared, 

 giving the fish a contrastive mottled appearance. Rotations of the disk up to about 

 50 revolutions per minute did not cause, in strong diffused sunlight, any appreciable 

 alterations in the pattern. If anything, it became more conspicuous. As the rate of rev- 

 olution increased above this, the pattern became less and less conspicuous until at 200 

 revolutions, in strong, diffused sunlight, the fish became as nearly uniformly gray as it 

 did on a stationary gray background. Under both conditions the entire surface, with 

 the exception of the ocelli and spots mentioned above, became, in the same length of 

 time, almost uniformly gray. The ocelli and spots were in all cases considerably darker 

 than the rest of the surface but not nearly so dark as they were when the disk rotated 

 more slowly. 



Thus, in moderately strong light the images produced on the retina of the fish 

 fused completely when the disk made 200 revolutions a minute, or, since the disk con- 

 tained 32 sectors, when the images were superimposed at the rate of 6,400 per minute. 



