270 BULLETIN OF THE 



formis, hut are stouter and covered with small regularly placed patches of pig- 

 luent, whose color was not noticed (tig. 4). The single specimen which we 

 iound liad four feeding-polyps. 



The tentacles (fig. 4) arise from the base of the polypite, and closely resemble 

 the tentacles of R. fdiformis. Along their upper side, near their proximal end, 

 they bear many simple buds,* which are undeveloped tentacular knobs (fig. 4). 

 The pendants, therefore, in this genus, seem to bud from the tentacles, and not, 

 as in Agalma, from the base of the polypites. These buds pass by successive 

 changes into well-developed tentacular pendants of two kinds. The adult 

 form of these two kinds of tentacular knobs does not difier as radically in their 

 anatomy as that of the pendants of R. filiformis. 



The first and more numerous tentacular pendant consists of a simple coiled 

 sacculus (fig. 5), without involucrum, mounted on a long, flexible peduncle. 

 The termination (fig. 5') of the sacculus is trifid, consisting of a prominent 

 median projection 'and smaller lateral protuberances. The tip of each bears a 

 black pigment-spot. The whole surface of the knob is abundantly supplied 

 with thread-cells. 



A second kind of tentacular pendant, similar in some respects to the former, 

 was also observed (fig. 6). It diifers from the former in having an undivided 

 tip which bears a small pigment-spot. The second form of tentacular pendant 

 is more slender than the first, and may be found to be simply the immature 

 condition of that described above. The " first form " of tentacular knob in 

 R. gracilis corresponds part for part with the "first form" of knob in R. filiformis. 

 Representatives of what are known as the "second" and "third"! kind of 

 pendant in the latter species seem not to exist in the present species, R. gracilis. 

 The whole outer surface of the tentacle is thickly covered with large lasso-cells. 



The sexual organs (s) differ in shape, size, and position from those of most 

 other Physophores. They resemble most closely homologous structures in the 

 genus Physalia. Instead of arising in clusters from the stem, each placed 

 midway between two polypites as in some genera, or from the base of a taster 

 as in others, they hang from the axis at a point immediately below the base of 

 attachment of the upper polypite. f In the single specimen studied they are 

 wanting on the other feeding-polyps. 



The cluster (fig. 3) has a botryoidal shape, and hangs from the axis by a 

 single slight peduncle. The pedicle of the sexual bodies opens directly into a 

 spherical body, as shown in fig. 3. This body varies in size and shape, and 

 contains a simple cavity. It is sometimes so contracted that it is indistinguish- 



* Huxley, op. cit.,l?\. VIII. fig. 18, represents similar buds near the j)roximal end 

 of the tentacle. 



t By the "second" and "third" form of tentacular pendants in R. filiformis, 

 I refer to those structures which Gegenbaur (Beit. z. n. Kennt. d. Schwimp., 

 Taf. XVni. figs. 8, 9) calls respectively the Ila-iulformiges Angelorgan and the 

 V'dgelkopfahnliches Organ. 



X They occupy a similar position as regards the polypites as the female bells of 

 Agalma. 



