412 REPORT—1863. 
was produced upon the body of the polypite, and contained multitudes of ova, 
which had their origin in the walls of the manubrium. It is probable that 
the medusoid here described, though truly phanerocodonic, discharges its 
eggs without ever becoming free. 
Wright observed in a Coryne, to which he gives the name of C. gravata, 
fixed medusiform bodies of the phanerocodonic type, with the spermatic mass 
largely developed in the walls of the manubrium*. 
But the most complete observation of this class has been made by Agassiz‘, 
who has seen in his Coryne mirabilis fully developed medusz, referable to the 
type of Sarsia, originate as buds in the earlier months of the year, and then 
become free before any formation in them of ova or spermatozoa; while 
during the later months the same Coryne gives origin to medusa-buds which 
are arrested ata slightly less advanced stage of development, never become 
free, and produce in some cases ova, in others spermatozoa, in voluminous 
masses developed in the walls of the manubrium. 
In all the above instances the generative elements had advanced far 
towards maturity, and the gonophore had apparently reached its final form 
while still attached to the trophosome. It will, I think, be found that, when- 
ever this is the case, there is some degree of arrest in the complete medusal 
development, though not sufficient to reduce the gonophore to the condition 
of the adelocodonic forms. This shows itself especially in the tentacula, 
which usually remain in the state of mere tubercles, and, even when most 
developed, never attain the completeness and extensibility which characterize 
them in the free meduse. It is also probable that in these cases the manu- 
brium never developes a mouth at its extremity t. 
In other cases, however, the gonophores are detached, as free meduse, in 
a much less mature state, in order to undergo further development in the 
open sea, and while yet the generative elements are quite rudimental,- or 
even before any trace of them can be detected. Here, likewise, several cases 
have been recorded which prove, by direct and continuous observation, that 
these medusz also have, as their proper function, the perpetuation of the 
species by true sexual generation. 
The generative sacs which become developed upon the radiating canals of 
certain free meduse were seen by Van Beneden§ in the meduse thrown off 
by Laomedea dichotoma||. He mistook them, however, for nervous ganglia, 
an error which was afterwards rectified by Krohn §] ; while similar bodies were 
subsequently observed, and correctly interpreted, by Kélliker** in a hydroid 
which he named Campanularia dichotoma, but which is certainly not the 
true Campanularia (Laomedea) dichotoma, whose meduse are different from 
those described by Kélliker. Kolliker in this instance recognized the rudi- 
ments of the generative sacs even before the escape of the medusz from 
the gonangium. 
Gosset+ mentions and figures the generative sacs on the radiating canals of 
* Edin. New. Phil. Journ. 1858, vol. vii. p. 282. 
t Op. cit. vol. iv. p. 189. 
+ The gonophores of the StrrHoyoruora also afford an example of the same phenomenon, 
in the fact that in most cases, even though attaining the phanerocodonic form, they stop 
short of the completely developed medusa, while at a very early period they develope the 
generative elements within them. 
§ “Mém. sur les Campanulaires,” p. 26. pl. 2. fig. 15, in Nouv. Mém. de l’Acad. Roy. 
de Brux. tom. xvii. 1844. 
|| Named Campanularia gelatinosa in Van Beneden’s memoir. 
€ Wiegm. Arch. 1851, p. 267. ** Zeit. f. w. Z. vol. iv. p. 301. 
++ A Naturalist’s Rambles on the Devonshire Coast, 1853, p. 303. pl. 19. fig. 3. 
