180 WILLIAM BRODBECK HERMS 



The surging out of the mass of larvae from the body of the fish, 

 might have been supposed to be due to the disturbance caused by 

 cutting open the carcass along the abdomen, but it should be borne 

 in mind that when they were disturbed, as was shown in Experi- 

 ment 2 (in that case by a mere shifting of the lamp), they invari- 

 ably moved away from the source of light as is also shown in the 

 experiments to follow. 



Experiment Jf.. August 14, 1906, 8 :30 p.m. This experiment was 

 made on larvae from twelve to fourteen hours old, and showed that 

 a mass of larvae about 4 cm. in diameter could shift its position 2.5 

 cm. in about one minute. In this case the larvae travelled away 

 from the fish and toward the light at a comparatively rapid rate, 

 The experiment was necessarily closed after five minutes trial. 



This experiment differs from the preceding one in that the larvae 

 wholly left the fish and moved toward the light, while in the former 

 (Experiment 3) the larvae remained more or less in touch with the 

 body of the fish and were merely massed in a strongly illuminated 

 region. 



The following experiments may be cited as further illustration 

 of the positive reaction of the fly-larvae. While colored glass was 

 used in some of these experiments the results are not to be con- 

 sidered as decisive for their color reactions, though this may be an 

 approximation to the actual conditions. The glass was not tested 

 for its efficiency as a light filter by any rigid method; ruby glass 

 however, may be considered reasonably safe for red when of uni- 

 form thickness; the blue glass on the other hand is pretty surely 

 unreliable. The rather remarkable results obtained by means of 

 the spectrum are interesting and uniform. 



For experimental purposes the animals were retained in a 

 glass box (10x12.5x17.5 cm.) constructed of clear photographic 

 plates. Usually from 200 to 500 larvae were used for each experi- 

 ment, and their food consisted in all cases of fish. As a rule only 

 a portion of the fish upon which the eggs had been deposited was 

 taken as food for the larvae. A "Model C" automatic acetylene 

 lantern was used, which gave a fairly satisfactory light in quality, 

 though somewhat variable in intensity. The prism was made by 

 carefully cementing together three plates of glass (12.5 cm.xl7.5 



