428 STUDIES IN EVOLUTION 



common axis of the branch or colony, after which no gem- 

 mation commonly takes place. By the explanation here 

 advanced this lack of a tendency to gemmation in the distal 

 portions of the corallites in this genus accounts for the ab- 

 sence of mural pores when such portions are in contiguity. 

 The periods of gemmation in Romingeria are periodic. Sev- 

 eral buds, often forming a verticil, are given off from the 

 parent corallite. Considerable elongation of the tubes takes 

 place before other series of buds are produced. The budding 

 is prolific at these points, and here also occur the mural pores. 

 The latter are therefore developed when the period of gem- 

 mation is in force. If pores are formed elsewhere when the 

 corallites happen to come into juxtaposition, it may possibly 

 be explained as the result of a stimulus produced by the 

 contiguity of the animals. Further observations are neces- 

 sar}^ to show that pores exist at other places than the bases of 

 the verticils or points where numerous buds are given off and 

 where from crowding the corallites are in juxtaposition. 



It therefore seems that, primaiily, the development of 

 mural pores is identical or homologous with the process of 

 gemmation. Whether this cause is operative in such forms 

 as Columnopora or Alevopora yet remains for investigation. 

 The porous condition of the walls in these genera may be an 

 inherited character without an active exciting cause, or it 

 may be teleologically different. 



