CCELENTERATA I HYDROZOA. 1 05 



angle, to four corresponding points on the under surface of the 

 umbrella, equidistant from its centre. Under the middle of the 

 umbrella, therefore, is a chamber, whose floor is formed by the 

 quadrate disc, whilst its roof is constituted by the under wall of 

 the central cavity of the umbrella, and its sides are open. The 

 reproductive elements are developed within radiating folded 

 diverticula of the roof of this genital cavity." 



It appears, finally, that amongst the old Pulmograde Aca- 

 lephse, or amongst what would commonly be called Jelly-fishes, 

 we have the following distinct sets of beings, which resemble 

 each other more or less closely in appearance, but differ in 

 their true nature : 



1. Free medusiform gonophores of various Cotymda, Sertu- 

 larida, Campanularida, and the Oceanic Hydrozoa. 



2. True Mcdusufo) entirely resembling the former in anato- 

 mical structure, but differing in the fact that their ova do 

 not give rise to a fixed zooid, but to free-swimming organ- 

 isms exactly like the parent hydrosoma (Trachynemida and 



3. Hydrozoa, which are provided with an " umbrella " (with 

 all the peculiarities belonging to this structure), but which re- 

 produce themselves without the intervention of free generative 

 zob'ids produced by fission (Pelagia). 



44. The free generative zooids of most of the Pelagidce, with 

 an umbrella and a single polypite, the primitive hydrosoma 

 being fixed and sexless (Aurelia, Cyanea, &c.) 



5. The free generative zooids of the Rhizostomidte, with an 

 umbrella and a complex central tree bearing many polypites 

 (Rhizostoma, Cephea, &c.) 



Of these five classes of organisms, Nos. i and 2 constitute 

 the Gymnophthalmate Medusa of Professor E. Forbes, whilst 

 Nos. 3, 4, and 5 are the Steganophthalmate Medusa of the 

 same naturalist. 



SUB-CLASS V. GRAPTOLITID^E. The organisms included at 

 present under this head are all extinct, and they are in many 

 respects so dissimilar, and their structure is so far from being 

 entirely understood, that it is doubtful if any definition can 

 be framed which will include all the supposed members of the 

 family. The following definition, however, will include all the 

 most typical Graptolites : 



Hydrosoma compound, occasionally branched, consisting of 

 numerous polypites united by a ccenosarc; the latter being 

 enclosed in a strong tubular polypary, whilst the former were 

 protected by hydrothecae. In the great majority of Graptolites 

 the hydrosoma was certainly unattached, but in some aberrant 



