SEXUAL OEGANS OF CCELENTEEATA. 121 



ings in the parts of the body where they are formed, but as they are 

 only present when the generative matters are being formed, they 

 may be regarded as "temporary organs." 



There are great peculiarities in the structural relations of the 

 parts which enclose the generative products, but they are connected 

 by a large number of intermediate stages. In those Hydroid 

 colonies that give rise to free Medusae (cf. § 74), the Medusae carry 

 the generative organs ; the MedusEe form the generative animals of 

 their proper Hydroid- Polyps, and elaborate the semen or ova in the 

 walls of their stomach, or in the radial canals, or, lastly, sometimes in 

 the circular canal. In some this production does not take place until 

 a long time after they have broken away from the Hydroid colony ; 

 in others it happens earlier. The latter bring us to those in which 

 the reproductive matters are formed while the Medusae are still 

 attached to the Hydroid stock. Next, then, comes that stage in 

 which the Medusa is not only not broken oif, but is not completely 

 developed. All the organs which are of use in the full and indepen- 

 dent mode of life — mouth, gastric cavity, tentacles, swimming-bell, 

 &c. — appear in a state of atrophy. We have in fact medusoid 

 buds, in which the sexual products arise. In others the medusoid 

 form is completely lost, and quite simple structures appear on 

 the Hydroid stock in the form of generative capsules, into which, at 

 the most, a process of the gastrovascular system still projects. Such 

 are the structures described above in Hydractinia. These genei'ative 

 buds arise, like the medusoid buds, and the Meduste themselves, 

 sometimes on the common stock, sometimes on the body of the 

 Polyps, and often only at certain parts of the latter, as, for example, 

 between the outer and inner circlet of tentacles in the Tubularia. 

 Where the proliferating Polyps are provided with a sheath, the 

 generative buds are always enclosed by the same test as the Polyps 

 themselves. Thus the phtenomenon of the budding of Medusae can 

 be followed back to a stage in which the bud has the appearance 

 of a mere generative organ of the hydroid stock. 



The Siphonophora resemble the Hydro id-Polyps. The formation 

 in them of sexual animals of the Medusa-type, with the simul- 

 taneous formation of other medusiform persons, helps to explain 

 the phaenomenon known in the Hydroid-Polyps as alternation of 

 genex'ation, as being a division of labour. In some of the Siphono- 

 phora the generative animals become free Medusae, in the walls of 

 whose stomachs the generative products are formed (Velella, Chry- 

 somitra). Most of them have only medusiform buds, which are 

 found in very various stages of degeneration (cf. Fig. 33, B g E). 

 The stomach of the Medusa becomes gradually represented by the 

 generative organs only, and the Medusa-bell degenerates into a 

 mere covering for them. Thus they occur arranged either singly 

 (Diphyidae), or grouped into racemose tufts (Physophoridae), which 

 are placed on the stem of the stock, or, it may be, on definite 

 persons belonging to it. 



Ed. van Beneden. De la distinction originelle du testicule et de I'ovaire. Bull. 

 Acad. Belg. 2™= Ser. T. xxxvii. 5.— G. Koch, Morph. Jahrb. Bd. ii. p. 83. 



