OBSERVATIONS ON POLYZOA. 211 



REPRODUCTION. 



The Phylactolaemata have two modes of reproduction, 

 one by buds, and the other by eggs. The .former occurs 

 in two ways ; by statoblasts, either fixed or free, and by 

 regular buds, which grow out from the side of each poly- 

 pide. The first are the founders of new colonies. The 

 last merely increase the number of individuals in each es- 

 tablished community. The colonies are, however, some- 

 times multiplied by other processes, which cannot be classi- 

 fied under either of the above heads. In large specimens 

 of Plumatella Arethusa the polypides on the old trunk die 

 first and the remnants of the coenoecia are gradually swept 

 away, leaving the branches as so many independent colo- 

 nies (PI. 8, fig. 1). This, also, is not uncommon with Plu- 

 matella diffusa, and is, probably, peculiar to all the species 

 of this genus that distribute their branches over a large 

 surface. 



1 have directed, perhaps, more attention to the old age 

 than to any other period of the growth of the individual, 

 and among the many curious and novel facts, which this 

 comparatively untravelled path of investigation hashed me 

 to, there are few more interesting than the above. 



Specimens of Fredericella may be often observed at- 

 tached near the ends of their branches by the soft ectocysts 

 of their younger polypides, the ragged end of the branch 

 floating freely above. These may sometimes have been 

 torn by accident from the parent colony, but in the majori- 

 ty of cases they owe their liberation to the decay of the 

 original stock. In Pectinatella and Cristatella the march 

 of extinction is, also, from within outwards. But, in con- 

 sequence of the greater width and the common occupation 

 of the coenoecium by the polypides, the decay of those in 

 the interior does not effect the vitality of the trunk, and 

 their living coenoecia carry both the quick and the dead 

 (Pi. 9, fig. 11). 



Thus death, which is an active agent in multiplying the 

 number of independant colonies in Fredericella and Plu- 

 matella, is, probably in Lophopus, and certainly in Pec- 

 tinatella and Cristatella, of no avail ; the constrictive pow- 

 er of the endocyst being its functional substitute in the 



