130 ELEVENTH REPORT. 



was produced, by having the aquarium darkened except at one end, 

 and a vertical strip (X Fig: 3) at" one side next the open end so as to 

 throw sunlight .across the open end diagonally. There was also a slit near 

 the top of the closed end of the aquarium for a band of sunlight or a mirror 

 beam. The medusae gathered farthest from the source of light when sun- 

 light was passed through the end of the aquarium. Now crossing this sun- 

 light by a beam from a small mirror, through the slit, made the medusae 

 leave the place where the light overlapped, where it was more intense. The 

 position of the sunbeam and mirror was reversed with about the same result. 

 In this case the mirror was held high enough so that its circumscribed area 

 of reflected light fell near the farther corner of the aquarium, leaving a place 

 of high intensity where the band of sunlight fell, but not as high as that 

 where the two overlapped. Other portions of the aquarium seemed to be 

 of too low intensity, as the medusae remained in the lighted portions. The 

 effect of relative intensities was then nicely shown by lowering the mirror 

 so that its beam was projected horizontally across the band of sunlight. 

 (Similar to Fig. 2) Again there were more medusae in the corner farthest 

 from the two sources of light. In a final test sunlight was admitted through 

 a slit near the bottom of the acjuarium and the mirror placed l^ack of the 

 aquarium in such a position as to throw reflected light obliciuely against 

 the sunlight, as it were. (Fig. 4 and 5). Now the medusae collected some 

 distance from the back of the aquarium in the region of lesser intensity. 

 Ray-directing, seems to me, out of the question. 



Some experiments made to determine the height to which Gonionemus 

 would swim after a single stimulation showed such interesting reactions of 

 sudden change of intensity that I give them here, especially since they lead 

 rather naturally to the next section. Eight medusae were placed in a hydro- 

 meter jar 72 cm. deep and 9 cm. in diameter. When they had come to rest 

 in the bottom of the jar they were stimulated by darkening the room until 

 the medusae could still be seen. They swam to the top at intervals some- 

 what irregularly. The shades were now quickly raised and thus the top 

 of the jar was illuminated with strong diffused daylight. Three medusae 

 swam downward so directly as not to touch the sides of the aquarium not 

 varying nine centimeters from their course. Five floated down part way 

 and then two turned and swam the last third of the way down. In another 

 trial, out of three at the top one swam down 31 cm. without touching the 

 sides of the jar. From this it seems that change of light intensity starts, and 

 gravity, also a stimulus, directs the downward course. 



While there is no doubt that the up-swimming of Gonionemus is directed 

 by gravity, as stated by Yerkes [(5) p. 281], nevertheless light seems to be a 

 more important factor than he says. Indeed I may say it is a necessary 

 condition as may be seen from what follows. That it is not directive Yerkes 

 [(5) p. 281] has shown by his experiment of bottom illumination. The 

 medusae move up to the surface and turn over normally. But casual observa- 

 tion of the upper surface of the water shows that it is sufficiently illuminated 

 from the bottom to allow the medusae to come to the top. So I substituted 

 lateral illumination through an opening near the bottom of the aquarium. 

 The aquarium was otherwise darkened and reflection reduced as much as 

 possible by filtering the water. Now the medusae swam near the bottom, 

 8 cm. being the greatest height reached. As a check, bottom illumination 

 was then used and immediately one of the animals swam to the surface and 

 one came near it. Three other experiments were made and the most striking 



5. American Jour. Physiol., 3 903, Vol. IX. 



