﻿120 proceedings of the [April, 1857. 



merous chymiferous tubes, a broad digestive cavity (?) and tenta- 

 cula arising above an undulating margin which hangs below 

 them. Also, a folded mouth. These characters, if I have con- 

 ceived them rightly, would indicate something approaching the 

 Aeginidae and Aequorea, but I would not venture to speak more 

 particularly. 



To sum up the matter, it appears to me that the Hydroid Medu- 

 sae form two natural groups or sub-orders, based on general form, 

 and an ensemble of structure characteristic of each, and also on a 

 difference in the method by which the locomotive disk is formed ; 

 which difference will probably prove constant. That this corres- 

 ponds in part to the distinction introduced by Ehrenberg between 

 the different forms of the fixed hydroid larvae of these Me- 

 dusae, by which he divided them into Tubularina and Sertularina. 

 But that these groups are not sufficiently inclusive, since there are 

 many hydroids which are never fixed — and that these should be 

 united with the two foregoing groups, viz: the Siphonophorae, 

 with the Tubularina; and the Aeginidae, and their allies with the 

 Sertularina. Should subsequent research prove these two meth- 

 ods of developing the disk which we observe in Tubularia and Cu- 

 nina to be constant for their groups respectively, to which I have 

 referred them, I propose to name the two sub-orders so distin- 

 guished, Endostomata, including the Tubularina and Siphono- 

 phorae and Exostomata, including the Sertularina and Aeginidae. 



We should next consider the Families which may naturally be 

 formed under these groups. The subdivision of the Gymnop- 

 thalmata into families by Forbes, has already been criticised by 

 Agassiz and Gegenbaur. His characters are the simple or branched 

 forms of the radiate tubes, their number, and the position of the 

 sexual organs ; thus constituting Willsia, a distinct family, on ac- 

 count of its branching tubes. The characters of his other five 

 families, it is difficult to sieze. They are besides Willsiadae — 

 Oceanidae, Aequoridae, Circeadae, Geryonidae, and Sarsiadae. Oce- 

 anidae and Sarsiadae are natural groups. Circe, which alone con- 

 stitutes the family of Circeadae, with the addition of a new genus, 

 Persa, described below, will probably also represent a natural 

 group. Tima, among his Geryonidae, is referable to the Euco- 

 pidae; and it is at least doubtful whether Stomobrachium should be 

 ranked with the Aequoridae, while Polyxenia Alderi is probably 

 the type of a new genus, whose position will hardly be assignable 

 until further research has made it better known.* 



*Is it a Hooded-eyed Medusa? 



