74 MORPHOLOGY. 



lacunte, whicli, increasing in size, become confluent with each other, and the ectoderm thus 

 becomes split into two layers by a true ckorization. Tlie external layer remains in contact with 

 the chitinous capsule, while the internal layer, remaining adherent to the endodenn, becomes 

 more and more withdrawn towards the axis of the bud, where it now constitutes the external 

 or ectodermal layer of an axile column or blastostyle. The capsule thus becomes lined with a 

 thin layer of ectoderm, which is continuous with the ectoderm of the blastostyle only at its distal 

 and proximal extremities, these two membranes being in the whole of the intermediate region 

 separated from one another by a wide interval. This interval, which constitutes the cavity of the 

 developing gonangium, is thus nothing more than a large lacuna ; and it is in this lacuna that 

 the sporosac or blastocheme now begins to bud forth from the axile column. The excreting and 

 modelling of the chitinous gonangium would seem to devolve for some time still on the ectodermal 

 lining instead of being, as in the hydranth-bud, transferred at a very early period exclusively to the 

 disc-like summit of the axis. After a time, however, the lining membrane entirely disappears, and 

 henceforth the excreting and modelling of the gonangium seems to devolve on the terminal disc 

 of the blastostyle. While the gonangium is yet young, numerous irregular fleshy bands may be 

 seen stretching across the cavity from the blastostyle to the external wall. These bands are the 

 remains of the original union between the two layers into which the ectoderm has split. They 

 are generally torn, and disappear as the gonangium, increasing in size, has its walls more and 

 more widely separated from the blastostyle ; but they are also occasionally more or less visible 

 in the full-grown gonangium. 



A compai'ison between the developing hydranth and its liydrotheca, on the one hand, and 

 the developing blastostyle and its gonangium on the other, aflbrds a most instructive parallelism, 

 showing the close connection between the hydranth and the blastostyle. If in the hydranth-bud 

 the development were arrested at the point to which it arrives just before the terminal disc has 

 withdrawn itself from the roof of the young hydrotheca (woodcut, fig. 33 d), in order to develop 

 its tentacles, we should have in almost every particular a gonangium with its blastostyle (see 

 woodcut, fig. 18). The development of a mouth and tentacles, however, points towards a different 

 destination ; and now, instead of producing zooids destined for generation, the hydranth applies 

 itself solely to the nutrition of the colony. 



The gonangium does not always present the simple form which we find in Laomedea Jlexuosa, 

 and we have already seen the remarkable modification which it undergoes in the female colonies 

 of Sertularia rosacea, S./allax, and S. tamarisca, by the formation of a marsupial chamber for the 

 protection of an extra-capsular sac, in which the ova are retained during the earlier periods of 

 their development. 



c. Development of the Sjjorosac. 



The development of the sporosac or adelocodonic gonophore, in its simplest form, may be 

 easily studied in Hydractinia ecMnafa. In this hydroid the gonophores are borne on a blastostyle 

 (PL XV, and woodcut, fig. 4 bb, c), which here, just as in the blastostyle of the Sertidarians and 

 Camjianularians, is morphologically nothing more nor less than an arrested hydrantii, but in 

 Hydractinia never developing a gonangium. 



