468 ANNELIDA. 



altered, and the animal passes from a creeping to a free-swimming 

 existence. Further, there is a sexual dimorphism, inasmuch as the 

 sexually mature male is more altered from the immature form than 

 the female, having fewer unaltered anterior segments. 



In Nereis dumerilii Claparede has shown that there is a still 

 more remarkable phenomenon of the same kind. In this species 

 it appears that there are four distinct sexual forms, differing from 

 one another in size, form, mode of life, etc.; they are (1) a sexual, 

 dioecious Nereis form, distinguished by its small size; (2) a sexual 

 hermaphrodite Nereis form ; (3) a dioecious Heteronereis which is 

 small, and swims on the surface ; (4) a dioecious Heteronereis which 

 is larger, and creeps on the bottom. It is not known how these 

 forms are related to each other. 



In the Syllidae phenomena of a very similar kind have been 

 observed, and there is a Heterosyllid sexual condition in which 

 the posterior sexual segments acquire a dorsal bundle of specially 

 long setae ; but in this family the phenomenon is often accom- 

 panied by asexual reproduction, which we will now proceed to 

 consider. 



Asexual reproduction. The power of asexual reproduction is 

 closely associated with the power of reproducing lost parts, and with 

 the power of indefinite growth, i.e. of growth after the adult 

 condition has been attained. In many Polychaetes the number of 

 segments continues to increase throughout the life of the animal by 

 the addition of new segments between the penultimate and anal 

 segments ; and in most, if not in all of them, the power of repro- 

 ducing lost parts is very extensive. For instance, in many of them 

 it has been observed that if the body be cut into two parts, the 

 anterior part will produce a new hind end, and the posterior part 

 a new head. 



These two phenomena the reproduction of lost parts and the 

 protracted formation of new segments are both instances of the 

 phenomenon of budding, and it is not surprising to find reproductive 

 gemmation normally occurring in the group.' 



The simplest cases are those in which a zone of fission is formed 

 between two segments, which becomes differentiated into an anal 

 region for the part of the worm in front, and a cephalic region 

 for the part of the worm behind. This results in the division of 

 the worm into two (Salrnacina dysteri, Filigrana iinplexa). Sexual 

 reproduction does not appear to take place in a worm undergoing 

 fission. 



