48 PLANT-ANIMALS [CH. 



that when specimens of the two animals are put 

 together into a dish, the bottom of which is half 

 black and half white, they segregate rapidly and 

 completely ; C. roscoffensis takes up positions on the 

 white ground, C. paradoxa on the black ground 

 (cf. Figs. 10 and 11). The distribution is in accord 

 with reasonable expectation based on knowledge of 

 the natural habitats of the two species. C. ros- 

 coffensis, attuned to a high light intensity, with its 

 place in the sun, is evidently a bright background 

 animal ; C. paradoxa, lurking in the shadows of the 

 weeds, though it also needs light for its growth and 

 development, is unused to well-lit situations and 

 seeks in preference the darker background. 



But though the selection of ground seems bio- 

 logically reasonable the question remains, IIOAV is it 

 done ? The hypothesis may be hazarded that it is 

 a phenomenon of association. C. roscoffensis is, as 

 we know, positively phototropic. In effecting a 

 phototropic response, it is bound in ordinary cir- 

 cumstances to pass from a darker to a lighter 

 background. The performance of the phototropic 

 movement is associated with the darker ground, the 

 achievement or consummation that is, a state of 

 immobility is associated with the brighter back- 

 ground. If therefore we adopt the hypothesis, 

 proposed by Semon, (1904) that environmental con- 

 ditions, which are contemporaneous with a particular 



