378 W. K. BROOKS OX THE LIFE-HISTORY 



The two naturalists who first described the development of the Geryonidae, Fritz 

 Muller (55) and Haeckel (30) published their accounts at a time when embryological 

 knowledge was much less advanced than it is to-day and when comparatively little was 

 known of the histological structure and significance of the hydroid larva. They both fell 

 into the error of regarding the central capsule of cells as the sub-umbrella, and believed that 

 the digestive cavity, and its endodermal walls originated at a much later period; but our 

 present comparative knowledge of the embryology of other organisms would now lead 

 us, even in the absence of any record of its later history, to regard the central cells as 

 an endoderm, for the hypothesis that they are the ectoderm of the sub-umbrella implies 

 that the Geryonoid embryo is fundamentally different from all other known hydroid 

 embryos. 



At the present day the fact that the central cells of an ordinary planula become the 

 cells of the digestive cavity is, in itself, an evidence that the central cells of the Liriope 

 embryo are their homologue and equivalent, and the later history of the embryo fully 

 bears out this view of their nature, and puts out of question the acceptance of Fritz 

 Midler's and llaeckcl's interpretation. 



The next change which takes place, the formation of the mouth, is shown in tigs. 

 4 and 5. The cells of the oral area or peristome become ciliated and a depression appears 

 in the centre of the outer or ectodermal area, and a similar internal one is found in the 

 endoderm, as shown at e in fig. 4. These two depressions soon meet, and break 

 through to form the mouth, fig. 5, c, the edges of which become ciliated. Food is 

 now swept into the digestive cavity, although little growth takes place until the larva 

 is much older. If the gelatinous substance at the stage shown in tig. 5 were 

 absent, the larva would be identical in structure with a typical gastrula; but it is 

 quite clear, from the account of its origin which I have given, that it is essentially differ- 

 ent from the invaginate gastrulae of ordinary metazoa, and that the mouth is not an 

 orifice of invagination, but a younger structure than the digestive cavity. At this and 

 the following stages there is a noteworthy difference between our species and those 

 which were studied by Fol and Metschnikoff. In our species the endodermal cap- 

 sule, fig. 5, c and fig. 7, which is now a stomach, retains its rounded outline, and ulti- 

 mately becomes elongated along the principal axis, fig. 7, <1. Metschnikoff says that, 

 in Geryonia hastata, it becomes flattened so that its aboral wall is almost in contact 

 with its oral (52, PL 11, figs. 10, 11 and 14), while in Geryonia fungiformis, accord- 

 ing to Fol, the aboral side becomes pushed down into the oral half, so that it forms a 

 double cup, with a very thin cup-shaped cavity. The absence of this flattening, in the 

 American Liriope, shows that it has no important morphological significance. 



The tentacles,/', now begin to grow out around the edge of the peristome, as shown 

 in an oral view in fig. 6, and, in an oblique view in tig. 5. Two of them probably 

 appear before the others, and in the stage shown in fig. 6, there are three: two, which 

 are probably primary, opposite each other; and a third, 00° from these. A fourth soon 

 appeal's opposite' the third, and Fritz Midler's figures show that they are the primary 

 radial tentacles, figs, i), 10 and 11 /, of the medusa. They are solid, and consist of a 

 layer of ectoderm, continuous at the base of the tentacle with the ectoderm of the edge 

 of the peristome, and a solid endodermal axis, which may, in our species, bf clearly seen 



