i] PERIODICITY IN CONVOLUTA 33 



an arch exposed on all sides to the light it blooms 

 freely. These effects of illumination on flower- 

 formation have been investigated by Vbchting, whose 

 researches are summarised by Goebel (1900). In 

 order that plants may form flowers in a normal 

 way, the illumination must not sink below a certain 

 amount which is very unequal in different species. 

 If illumination is allowed to sink below the required 

 amount, the size of the whole flower or of its individual 

 parts is diminished and, with decreasing illumination, 

 a stage is reached at which the formation of flowers 

 ceases. 



Similar phenomena are doubtless common among 

 animals though they have not been investigated 

 systematically. Thus, though the phenomenon is 

 not one of reproduction in the strict sense, we may 

 cite Loeb's account (loc. cit.) of the effect of light in 

 inducing regeneration of the polyps of the Hydroid 

 Eudendrium racemosum. If a stem of this Hydroid, 

 covered with polyps, is put into an aquarium, the 

 polyps fall off very soon. If the aquarium is in 

 darkness, no regeneration of the polyps takes place 

 even after several weeks ; but, when they are exposed 

 to the light, new polyps form in the course of several 

 days. We may suppose that light favours the forma- 

 tion of definite substances which are the pre-requisites 

 for polyp formation. 



Similarly, Ave are bound to conclude from our 

 K. 3 



