136 
abruptly into a very extensile filament, which in its completely con- 
tracted state I have always found coiled into a regular spiral. The 
base of the tentacle presents a large cavity, which communicates 
freely with the marginal canal of the umbrella. 
Between each of the four tentacles are two lithocysts. They 
consist each of a spherical transparent colourless vesicle, containing 
a spherical highly refractile otolite, which is itself immediately in- 
vested by a very delicate vesicle, of which it looks like the nucleus. 
Between each lithocyst the margin of the umbrella is extended into 
a slight projection, having all the characters of a rudimental tentacle. 
There is a very wide velum, and in both it and the umbrella mus- 
cular bundles are largely developed. 
In no instance were generative sacs developed in the specimens 
which came under my examination. Gegenbaur, however, has 
found them well developed on the course of the radiating canals 
in certain species of his genus Eucope. a genus to which he 
would refer all such forms as those here described in Laomedea 
dichotoma ? and Campanularia Johnstoni , and which, he tells us, he 
has traced to gemmation from the Campanularidae. In the un- 
doubted medusae of Campanularia Johnstoni , they have been seen in 
a similar situation, but not fully determined, by Mr Gosse, while the 
detection in them of distinctly formed ova by Mr Hincks, and more 
especially by Dr S. Wright, who has figured and described them, 
has so far completed the observations needed on this point. 
General Conclusions. 
To the general conclusions contained in the previous paper, the 
following generalizations may now be added. 
Besides the ordinary development of the ova in intra-capsular 
sporosacs, we find, 
1. That in certain species bearing capsular gonophores ( Sertu - 
laria polyzonias , &c.) the ova, after attaining in the interior of the 
capsule a definite stage of development, are transmitted into an extra- 
capsular sac ( acrocyst ), where they undergo a further develop- 
ment previously to their liberation as free embryos, and that this 
sac is formed (at least in the species where its connections were most 
satisfactorily traced) by a hernial protrusion of the endotheque (and 
ectotheque ?) of an intra-capsular sporosac through the summit of 
the capsule. 
