are reproductive in function. This complicated type of colony 

 reappears, still more perfected, in the Polyzoa, and, though 

 these arise from a sexual instead of from an asexual individual, 

 it finds in the specialization of the units, in the restriction in 

 numbers of the sexual individuals, and in the construction of a 

 common domicile, a close parallel in the colonies of the ter- 

 mites and of the social Hymenoptera. 



Another phase of asexual reproduction is represented by 

 strobilization, consisting of the fission and division of the 

 anterior half of the body into a number of segments v^hich 

 continually separate off from the anterior end to the base, 

 progressively attaining independent existence as ephyr^e which 

 develop directly into adult medusa. It is not -difficult to see 

 in strobilization the origin of cestode segmentation ; but in 

 the tapev^orms the segments, instead of being independent 

 entities, possess a common excretory and nervous system ; 

 and from cestode segmentation it is an easy step to the 

 segmentation of the annelids in which all the segments are 

 finally incorporated into a definite body, though in some 

 forms groups of sexual segments, growing a new head and 

 tail, may be continually given off from an asexual anterior 

 group as the proglottides are given off from the cestode scolex. 

 From annelid segmentation we derive the closely compacted 

 body of the insects and related types in w^hich we sometimes 

 see, as in the centipedes and myriapods, distinct traces of 

 an annelidan origin. 



Colonies of undifferentiated individuals are frequent in 

 the coelenterates, and are well represented by the colonial 

 anemonies. Similar colonies occur in many polyzoans, in 

 Cephalodiscus^ in Rhabdopleiira^ and in the tunicates. 



Many coelenterates, as the Trachomedusœ, and the cteno- 

 phores are never colonial and are not known to bud. Compa- 

 rable unsegmented forms occur in all the higher groups, being 

 represented, for instance, by most planarians, the rotifers, 

 the sipunculids and priapulids, the molluscs, Phor'onis, man}^ 

 tunicates, etc. 



There is a noticeable tendency, however, for these primarily 

 unsegmented types gradually to become more or less divided 

 internally into not more than three segments (as in Phoronis^ 

 the brachiopods, the chaetognaths, the Enteropneusta, etc.) 



(400^ 



