REPORT ON THE SIPHONOPHOR^E. 195 



Gonostyles. — The stems of the gonodendra, which bear the clustered gonophores, and 

 which we call gonostyles, are sometimes themselves simple palpons, at other times 

 branches of sexual palpons,' or in a peculiar manner combined with sterile palpons. The 

 ramification is in most Physonectae not very rich (mainly in the males), and not to be 

 compared with that of the Cystonectas. There are, however, exceptions, as in the female 

 gonodendra of some Discolabidse (PL XX. figs. 11-16). The male gonostyles of the latter 

 exhibit a peculiar appearance, since their distal part, after the detachment of the ripe 

 androphores, is covered with tubercles (as the remaining pedicles of the latter), whilst the 

 proximal part produces vicarious gonophores. 



Gonophores. — The medusiform gonophores of the Physonectae are in general of small 

 size, especially the females. Their umbrella is sometimes well developed, with four 

 equidistant radial canals and a marginal ring-canal, whilst at other times it is more or 

 less reduced, and sometimes rudimentary. The manubrium is larger in the androphores, 

 where it is usually club-shaped or cylindrical, with a central spadix ; often coloured 

 white, yellow, or red ; it is often very prominent from the narrow mouth of the reduced 

 umbrella (PL XII. fig. 17 ; PL XVIII. fig. 17, &c). The manubrium of the gynophores 

 is much smaller, ovate or subspherical, and develops constantly a single large ovum only. 

 This is often surrounded by an irregular network of peculiar anastomosing canals 

 (PL XV. fig. 15; PL XVIII. fig. 16). These " spadicine canals" arise by a peculiar 

 process : the original central spadix in the axis of the young ovarium becomes excentric 

 by the unilateral development of a single large ovum ; it grows around the latter in the 

 form of a hemispherical cup, and envelops it finally like a capsule ; by the partial 

 irregular concrescence of its two walls arises the reticulum of canals, which is called 

 " netzformiges Canal-System" by German authors. 1 



Ontogeny. — The development of the fertilised egg is hitherto known in the case of 

 only very few Physonectae. The first observations on it were made by myself in the 

 Canary Island Lanzerote, in December 1866, and January and February 1867. I was 

 able to observe there the embryonic development and the metamorphosis of Physophora 

 magnified,, Crystcdlodes rigida, and Athorybia ocellata (84, Tafs. i.»-xiv.). Further 

 observations were published in 1874 by Metschnikoff, who illustrated the ontogeny of 

 Halistemma rubrum, Cupulita picta, and Agalmopsis sarsii (85, Tafs. viii.-xii.). 

 The embryology of Agalma elegans was afterwards described by Fewkes (89, pis. i.-iv.). 

 Judging from these few observations, it seems that the Pbysonectae are subject to a 

 rather complicated metamorphosis and produce medusiform larvae, the morphological 

 value of which is probably very great for their phylogeny. Usually these monogastric 

 larvae (Physomda) develop the pneumatophore from their exumbrella very early, and 

 around it a corona of provisional bracts. (Compare Family XL, Athoridae, p. 200, and PL 

 XXI. figs. 5-13.) 



1 Compare Weismann, Die Entstehung der Sexual-ZeUen bei den Hydromedusen, 1883, p. 206. 



