the Geryonidae and ^Eginidse, 443 



between the two other groups of the Craspedota and Acraspeda. 

 Agassiz, on the other hand, has actually transferred them to the 

 Acraspeda (Phanerocarpce) . 



A very careful anatomical and histological investigation which 

 I made, after becoming acquainted with the genetic relations 

 just described, of two jEginidce {Cunina rhododactyla and C. aU 

 hescens) and two Geryonid(S [Carmarina hastata and Glossocodon 

 Eurybia), led me to the surprising result that these two families 

 present a far more essential agreement in internal structure 

 than could be supposed from the remarkably different external 

 form of the body. I shall only mention, in conclusion, the most 

 important agreements, in a few words. It is regarded as the 

 principal character of the ^ginidce that they have no annular 

 vessel like the other Craspedote Medusse, but merely blind sacs 

 which issue from the circumference of the stomach. But these 

 sacs are nothing but greatly dilated radial canals ; and, in fact, 

 they are united at the base by an annular vessel, which runs 

 along the margin of the mantle, and has hitherto escaped the 

 notice of observers merely on account of its very small dimen- 

 sions. The intimate structure of this annular vessel is exactly 

 the same as in Geryonia. As in that genus, so also in Cunina , 

 there is, immediately beneath the annular vessel, a slender cylin- 

 drical or semicylindrical cartilaginous band, from which issue a 

 number of centripetal and likewise cartilaginous bands, which rise 

 in a radial direction for some distance in the outer surface of the 

 margin of the mantle, and give support to it. Moreover, in 

 Cunina as in Geryonia, there is a fine nervous ring on the mar- 

 gin of the umbrella, contiguous internally to the insertion of 

 the velum, superiorly to the lower margin of the annular vessel, 

 externally to the gelatinous substance of the mantle, and infe- 

 riorly to the cartilaginous ring. The formation of the sexual 

 products in flat, leaf-like, saccular dilatations of the radial canals 

 is also exactly accordant in both the families of ^ginida and 

 GeryonidcB, and very different from that occurring in all other 

 Medusae. 



But the anatomical relation between Cunina and the larva of 

 Geryonia is far greater than between the mature animals of the 

 two genera. These two forms possess in common especially 

 the characteristic firm habit of the umbrella and the peculiar 

 structure of the solid, rigid tentacles, which are wanting in 

 the mature Geryonia, The primary mass of these tentacles is 

 formed by a cartilaginous cylinder, which is covered by a mus- 

 cular tube; over this is an epithelium, in which urticating 

 cells are here and there developed. The stomach of the young 

 larva of Geryonia is also a very shallow sac, as in Cunina. The 

 most essential anatomical difference between the Geryonidce ^nd 



