CRAWLING OF YOUNG TURTLES TOWARD THE SEA 329 



without, however, making very clear what is meant by these, 

 and finally concludes that, though the animals are not influenced 

 in their movement by the sun, they are nevertheless positively 

 phototropic ('11, p. 75). He compares them with certain posi- 

 tively phototropic animals studied by Cole ('07) in which responses 

 to the area of illumination rather than to the intensity of the light 

 was the determining factor in their locomotion. In my own 

 tests on the turtles I have never seen any evidence that they are, 

 strictly speaking, phototropic. Thus I have never been able to 

 get a turtle in a dark room to creep toward a light such as hap- 

 pened with all the animals tested by Cole. Hence I think it 

 improbable that the young loggerheads are correctly described as 

 positivelj' phototropic. To me they seem to be an example of a 

 much more complicated set of relations than those seen in photo- 

 tropism. Their retinal images are immensely complex as com- 

 pared with those in many of the lower forms, and they respond 

 more to the details of these images than to the images as wholes. 

 That part of the image which represents the region of the horizon 

 is much more important in determining the direction of locomotion 

 in the young turtle than any other part. If a portion of the 

 horizon is interrupted by many masses rich in detail, such as 

 trees, shrubbery, houses, etc., it may form a center from which 

 the turtle will move away. If a portion of the horizon is un- 

 interrupted and very uniform, as where sea and sky meet, that 

 part may form a center toward which the turtle will go. These 

 conditions indicate that the details of the retinal image in the 

 turtle are the significant features in determining the direction of 

 its creeping rather than the image as a whole and that conse- 

 quently the turtle possesses a kind of vision more like that in the 

 human eye than like that in the eye of purely phototropic animals, 

 even if we include among these the peculiar instances pointed out 

 by me ('03) and by Cole ('07) in which the size of the illuminated 

 area is as significant as the intensity of the light. 



Hooker's contention ('09; '11, p. 74), that turtles move toward 

 blue rather than toward other colors and thus reach the sea, may 

 perfectly well be correct. When I made my tests I unfortunately 

 had no adequate means of experimenting with colors, and con- 



