80 MORPHOLOGY. 



orifice of the umbrella. In the CamjmnuJarians, on the other hand, he believes that the umbrella 

 grows up from below as a ring round the manuljrium, which is thus never included in a closed 

 cavity, but is from the first directly exposed to the surrounding medium. In accordance with 

 these views, M'Crady divides the gymnophthalmatous or hydroid medusae into the " endostomata" 

 and the "exostomata." My own observations, however, will not allow me to adopt this division 

 of M'Crady. In the medusa of Campanularia at all events the development is essentially the 

 same as that just described in the medusa of Corymorpha. 



The medusa has not necessarily attained its complete development at the time when it has 

 become fitted for an independent existence, and has detached itself from the trophosome in order 

 to spend its future life in the open sea. It is very common to find both tentacles and lithocysts 

 less numerous at the time of liberation than at a more advanced period ; while in some cases 

 [^quoria) the radiating canals continue to increase in number with the age of the free 

 medusa > (woodcut, fig. 35). 



In every case in which I have had an opportunity of observing the formation of new 

 radiating canals, these have been developed in a centrifugal direction. They commence as ofisets 

 from the base of the manubrium (woodcut, fig. 35,/), or from the previously existing canals, 

 and then becoming elongated in the gelatinous substance of the umbrella, they direct themselves 

 towards the umbrella margin until they meet the circular canal with which they inosculate. This 

 penetration of previously formed tissue by the nascent canals, their invariable maintenance in it 

 of a definite direction, and their inosculation with a canal already completed, are phenomena not 

 without their general significance in the formative forces of living beings. 



In some cases still more striking transformations have been witnessed in the free medusa. 

 Thus Gegenbaur observed that the Tracliynema ciliatam, Gegenb., a medusa not yet traced to 

 a polypoid trophosome, is in its young state a free-swimming flask-shaped body, with three or 

 four minute tentacles in a circle round the base of its contracted neck-like portion, and with a 

 clothing of vibratile cilia over its whole surface. It subsequently developes an umbrella 

 and gastrovascular canals, and becomes provided with numerous imperfectly contractile 

 tentacles." 



It is, however, in the family of the Geryonidm to which the JEginida, as follows from 

 Haeckel's observations, must now, notwithstanding their very diSerent form, be united, that we 

 meet with medusae which, during their free state in the open sea, undergo the most striking 

 change, passing through a series of metamorphoses which consist, not only in the development 

 of new parts, but in the loss of organs which, being destined to enjoy only a transitory existence, 

 disappear, as is described below, to make way for permanent ones of an entirely different form. 

 It is true that none of these medusae have as yet been traced to a hydraform trophosome ; 

 but they are not on that account of less importance in the general history of hydroid 

 development. 



^ Alexander Agassiz has shown that in their order of succession the marginal tentacles of the 

 Hydroid Medusse obey a law very similar to that which Milne-Edwards and Jules Haimes have 

 shownjto regulate the formation of the successive septa in the Actinozoa. A. Agassiz in ' Proc. Bost. 

 Soc. Nat. Hist.,' vol. IX, Aug. 1862. 



" ' Generationswechsel,' p. 51. 



