DEVELOPMENT OF OBELIA 279 



forming a long fusiform contractile fibre resembling the 

 unstriped muscle-fibre described in the frog, but differing 

 from it in exhibiting a faint transverse striatipn. The rest 

 of the ectoderm cells form an epithelial layer lying over the 

 muscular layer. Further, the nervous system is better de- 

 veloped in the medusa than in the hydranth, consisting in 

 the former of a ring of branched ganglion cells running round 

 the margin of the umbrella near the bases of the tentacles 

 and a number of scattered ganglion cells lying close upon 

 the muscular layer of the sub-umbrellar surface, their branched 

 processes inosculating with one another and with processes 

 of the ganglion cells of the marginal ring. 



The free medusa is the sexual member of the hydroid 

 colony. At the time of its liberation it shows no trace of 

 reproductive organs, but after a while a finger-like down- 

 growth, involving both ectoderm and endoderm, is developed 

 on each of the four radial canals. The reproductive organs 

 are formed on the walls of these down-growths. In Obelia 

 geniculata the generative cells, oogonia or spermatogonia, 

 originate in the ectoderm of the manubrium, migrate into 

 the endoderm, pass along the radial canals to the down- 

 growths, and there go through the early stages of their matura- 

 tion whilst still in the endoderm, but in the later stages they 

 lie between the ectoderm and mesoglcea. The medusae are 

 dioecious, ova being developed in one individual and sperma- 

 tozoa in another. When the ova and spermatozoa have 

 arrived at maturity they are dehisced into the water by 

 rupture of the walls of the gonads, and, as medusae float 

 together in crowds near the surface of the sea, some of the 

 spermatozoa are sure to come in contact with the ova to 

 which, indeed, they are attracted as it were by a strong 

 chemical affinity (chemiotaxis). The ovum, after fertilisation, 

 divides and forms a hollow blastula like that of Hydra. The 

 further steps of development have not been followed out in 

 the case of Obelia geniculata, but are well known in an allied 

 species, Cfytia flavidula. In this species the blastoccele is 

 filled up, as in Hydra, by immigration of cells from the 

 blastula wall. But whereas in Hydra the immigrating cells 

 are derived from all parts of the blastula wall, it is stated 

 that in Chtiaflavidula they are derived from one pole only, 

 so that we have a process which very nearly resembles gastru- 



