38 GEMMATION. 



(PI. I, figs. 4 — 7), being generally more numerous and better developed on the more convex 

 side ; they thence pass outwards over the annulus, and project in short rays beyond the mar- 

 gin. The disc in all the species is of a deep brown colour, and would seem to be composed of 

 a single layer of hexagonal cells, whose external walls in most cases slightly project beyond 

 the surface of the disc, and thus give to the latter an elegantly mammillated condition. In 

 some cases, however, the cellular condition of one or both discs is very obscure. The annulus 

 is also composed of cells, which here occur in several layers ; these cells are also for the most 

 part larger than those of the disc and of a different colour ; they are filled with air, giving to 

 the annulus a light spongy texture, and act as a float by which the statoblast, when set free, 

 is kept near the surface of the surrounding w^ater. 



When the statoblasts are placed under circumstances favouring their development, they 

 open by the separation from one another of the two faces, and there then escapes from them 

 a young polyzoon already in an advanced stage of development, and in all essential points re- 

 sembhng the adult individual in whose cell the statoblasts were produced (PL XI, figs. 42, 43). 



The statoblasts have been always viewed and described as the eggs of the polyzoon in 

 whose cell they occur, a very natural mistake, and one the more excusable as the true ovary 

 had not yet been detected. Into this error I fell myself;* but I have now become convinced that 

 they are a peculiar form of bud, and must on no account be confounded with genuine ova. 



They are produced in the funiculus, from which they are evidently developed as buds, 

 and may generally be seen in various stages of growth, arranged upon this chord like beads 

 on a necklace, being younger as they approach the distal extremities of the funiculus (PI. Ill, 

 fig. 7; V, fig. 5,^; XI, fig. 36). 



In LopJiopus, I have succeeded in following them through various stages of their early 

 development. Their first appearance here is in the form of a little swelling upon the funi- 

 culus, consisting of a mass of minute cells, surrounded by a denser layer, which is continuous 

 with the surface of the funiculus. The swelling now increases in size, and assumes a more 

 regularly oval form, while its contents appear more uniformly granular, and are plainly seen to 

 be composed of two masses in close apposition with one another (fig. 37). We next find that the 

 two masses have lost their distinctness, and become fused together, and the whole contents 

 now appear to be composed of minute cells confined by an external, common, transparent 

 membrane, which is itself manifestly cellular (fig. 38). This cellular condition of the contents 

 must not be confounded with a true segmentation. The whole body now assumes a more 

 lenticular form, and within the external envelope two other investments begin to show them- 

 selves. One of these (the more internal) extends over the whole surface of the cellular con- 

 tents, but the other is confined to the margin of the lenticular mass which it embraces in the 

 foi-m of a ring (fig. 39). No manifest structure, beyond a simply granular one, can be as yet 

 detected in either of these last-formed envelopes ; but the ring is soon seen to be composed of 

 distinct cells (fig. 40), which present a bright central nucleus-like point, and a number of con- 

 centric layers, which remind us of the secondary deposits in certain vegetable cells. Up to this 

 point the investments are all colourless,'^and nearly transparent ; but we now find that the inter- 

 nal envelope and annulus become more and more opaque, while the former assumes a deep 

 brown colour, and the latter becomes yellow. They have both acquired a horny consistence, 



* ' Report on Fresh-water Polyzoa.' 



