268 HYDROID^. Part IV. 



at right angles to this, so that the outUne of a transverse section appears broadly 

 ovate, with the broader end next the outer wall (fj). 



The cellular m.ass (ff*) terminates at the base of the globular terminal expansion 

 (PI. XXIV. Fi(/. 1", li) of the stem, and the channels here empty into the broad 

 open space. Within these channels a more or less brisk circulation is constantly 

 kept up, ajjjjarently by means of vibratile cilia, but these are so excessively fine 

 that Ave have not been able to detect them. The outer wall (PL XXIIP. Fir/. 

 8, It) is about as thick as one half the shorter diameter of the channels, has a 

 smooth exterior surface, and is composed of numerous irregularly-disposed cells 

 (Fiff. 9, b), of moderate size. The horny sheath is variable in thickness, according 

 to its height; at the toj) it thins out to a mere filmy, epidermoid covering, upon 

 the globular expansion, and, passing downward, it gradually thickens, till, at the 

 base, it is twice or thrice, and, in very old specimens, even four times as thick 

 as the outer wall midway up the stem.' 



Proles mcdusoidea. — A full-grown medusoid has a diameter about double that 

 of Parypha or Thamnocnidia, and possesses as comj^lete a sj'stem of chymiferous 

 circulation as that of Coryne ; it also closely resembles the latter in the disposition 

 of its circular and radiating (PI. XXIV. Fu/s. 18 and 19, e; PI. XXVI. Fig. 3, e) 

 tubes (c), of which latter there are four, and sometimes five. Here, however, the 

 resemblance ceases, for the medusoid of Tubularia has not the least trace of tentacles. 

 There are the same double walls of the disk, and the single wall of the proboscis 

 as in Parypha and Thamnocnidia; but, unlike the latter, the inner wall is chan- 

 nelled by the chjaniferous system of tubes. As the medusoid never becomes a free 

 animal, dependent upon its own exertions for subsistence, but, on the contrary, 

 receives all its nourishment from the hydroid, it is not at all remarkable that the 

 chymiferous system of tubes should be in some instances irregular in its persistence, 

 as we have observed it to be. These tubes always develop completely, but here 

 and there we find that the radiating tubes become obsolete, even before the medu- 

 soid has reached its maturity {Figs. 15, 16, and 17). Sometimes every trace of the 

 chymiferous system has vanished long before the medusoid begins to wither {Fig. 15), 

 or nothing but the areas of junction of the radiating and circular tubes are indi- 

 cated by red spots, from one to five in number {Figs. 16 and 17, c), near the 

 opening of the disk. By the time the hydroids are beginning to be set free, the 

 proboscis {Figs. 21 and 23, d) becomes remarkably elongated; but instead of pushing 

 itself out through the aperture of the disk, as does that of Parypha, it doubles upon 

 itself After all the young are developed and set free, the medusoids Avither, and 

 either drop off or are resorbed (PI. XXVI. Fig. 4). When undergoing this process, 



' For further details upon tlie structure of these walls, see the paragraph upon histology (|). 270). 



