34 STUDIES IN GENERAL PHYSIOLOGY 



justified in taking as the effect of light the movements of 

 the animals to the top of the test-tube ; it might, indeed, be 

 a geotropic phenomenon. To decide this the animals were 

 placed in a test-tube which was lined with thick black paper 

 except for a strip 2 mm. wide. The uncovered strip was 

 turned downward, so that light could enter the vessel only 

 from below. Diffuse daylight was reflected through the slit 

 from below by means of a mirror. The animals collected in 

 the lower, lighted portion of the glass vessel. Their helio- 

 tropism is therefore more powerful than their geotropism, 

 even when only weak diffuse daylight is used. 



The geotropic experiments succeed only when the animals 

 have been in the light for some time and have not yet come 

 to rest. When the animals are kept in the dark for a long 

 time and the test- tube is not disturbed, they do not creep 

 upward. The orienting effect of the light always exceeds 

 that of gravity. The effects of gravity, like the effects of 

 light, usually appear only during certain periods in the life 

 of the animals; at any rate, they cannot always be demon- 

 strated with certainty. 



The contact-irritability of the caterpillars of Porthesia 

 chrysorrhcea shows itself by the way in which the animals 

 remain in the corners and convex sides of solid bodies. I 

 covered the boxes in which I cultivated my caterpillars with 

 large, square glass plates. These did not close the box 

 tightly, so that the animals could creep out and creep upon 

 the glass plates. Only rarely, however, were they found on 

 the free surface of the plates. The animals moved along the 

 rough edges of the plate until they reached the window side 

 of the dish. I confirmed this observation almost daily for 

 months. When I placed the animals upon the outside of a 

 cubical block, they collected by hundreds in one of the 

 upper corners. Of course, only a few have room in the 

 corner itself, but, as is generally the case with these ani- 



