DISCOVERY REPORTS 



35^ 



plicable. No wonder that the jelly-polyp was mistaken for the stem of a broken-off ' female medusoid '. 

 The jelly-polyp seems to be in fact a reduced nectophore. Both have the frenum-like longitudinal 

 muscular lamella referred to above as the ectodermal fold. 



One morphological feature of the gonodendra, common to those cystonects whose gonodendra are 

 well enough known to me for comparison, namely Physalia physalis, Rhizophysa eysenhardti, R. fili- 

 formis and Pterophysa conifera, is that in all of them the final sections of the branchlets consist of a 

 group of three kinds of bud: (i) a terminal gonopalpon without a palpacle (reduced tentacle), (2) a sub- 

 terminal nectophore (asexual medusoid bud) [(1) and (2) together forming what I have described, in 

 Physalia, as the sub-terminal section] and (3) male or female gonophores. But in Physalia alone there 

 is, proximal to these three structures, another palpon with a jelly-polyp (a reduced nectophore), these 

 two structures forming the terminal section, which is developed first and which gives rise by budding 

 to the sub-terminal section (carrying the nectophore and palpon). The sub-terminal section is budded 

 out (PL XXI, fig. 3) from the base of the palpon at a point opposite to its jelly-polyp. 



The difference between the gonodendra of Physalia and those of the other cystonects is that in 

 Physalia there are multiple-terminal as well as sub-terminal sections, whereas in the other cystonects 

 there is a single sac-like terminal section from which the multiple ones are budded. The sub-terminal 

 sections in all cystonects are clearly homologous. The arrangement in Physalia is the most complex 

 and so is probably the most highly evolved. These comparative morphological details form a key to 

 understanding the relationships between the various cystonects. 



Long ago, sketches of a terminal branchlet of a young gonodendron were published by Chun 

 (1897a, fig. 26b), and of an old as well as of a very young stage of the same by Richter (1907, figs. 7 

 and 8). Richter referred to four types of buds: (1) palpon, (2) male medusoid, (3) female medusoid 

 (i.e. nectophore), and (4) ' Gallertpolypoid ' (that is, jelly-polyp, interpreted by Haeckel as the ' pedicels ' 

 of detached female medusoids). Richter's figures, though incorrectly labelled, are basically sound and 

 may be compared with my photographs. Libbie Hyman (1940) in her indispensable book of reference 

 gave a figure of part of a gonodendron (fig. 155B) which is unfortunately misleading. She seems to 

 have mistaken palpons for jelly-polyps in both this and fig. 155 A. The latter figure in other respects 

 gives a good representation of an early stage of development of three gonodendra ; but the charac- 

 teristic arrangement of palpons, gonophores, jelly-polyps and nectophores is not correctly shown in 

 her fig. 155B. 



Nectophores 

 (PL XXIV) 

 The best criterion forjudging the degree of development of a gonodendron is the state of development 

 of the nectophores (PL XXIV, figs. 5-7). In early stages I find that the coelenteron of the nectophore 

 pedicel is spacious and the mesogloea thin. In the most advanced stages I have seen the mesogloea 

 is very thick and the coelenteron of the pedicel occluded (Text-fig. 28). Huxley (1859), Haeckel (1888) 

 and Chun (18976) all regarded the nectophores as female medusoids, which they thought probably 

 became detached and produced ova as free-swimming Anthomedusae. Chun even suggested that they 

 might develop into the Amalthaea amoebigera of Haeckel (1879) (a doubtful synonym of A. sarsi 

 Allman, fide Kramp). 



The nectophore of Physalia has no manubrium and no germ-cells have been seen at any stage of 

 its development. Moreover, the occlusion of the endodermal canal in the nectophore-stalk would 

 probably prevent later entry of ova in any way similar to that described by Brooks and Conklin (1891) 

 for Rhodalia. The secondary ectoderm of the sub-umbrella is peculiar in having folds of the meso- 

 gloea supporting it, figured by Goto (1897, figs. 8 and 9). What Goto called the manubrium is in my 



