156 ZOOLOGY sect. 



medusa. Thus the actual formation of the medusa from the 

 hydrula of .^ginopsis corresponds precisely with the theoretical 

 derivation given above (p. 136), It will be seen that in the present 

 case there is no metagenesis or alternation of generations, but that 

 development is accompanied by a metamorphosis — that is, the egg 

 gives rise to a larval form differing in, a striking manner from the 

 adult, into which it becomes converted by a gradual series oi 

 changes. 



Metagenesis is, however, not quite unknown among the Trachy- 

 linse. In a parasitic Narcomedusa (Ctonina parasitica) the planula 



Fig. 115 Larva of .ffiginopsis. m. moutli; (. tentacle. (From Balfour, after Metsulinlkoff.) 



fixes itself to the manubrium of one of the Trachymedusse which 

 serves as its host, and develops into a hydrula. But the latter, in- 

 stead of itself becoming metamorphosed into a medusa, retains the 

 polype form and produces other hydrulse by budding, these last 

 becoming converted into medusae in the usual way. 



Order 3. — Hydrocorallina. 



The best-known genus of Hydroid Corals is Millepora, one species 

 of which is the beautiful Elk-horn Coral, M. alcicornis. The dried 

 colony (Fig. 116 A) consists of an irregular lobed or branched mass 

 of carbonate of lime, the whole surface beset with the numerous 

 minute pores to which the genus owes its name. The pores are 

 of two sizes : the larger are about 1 or 2 mm. apart, and are called 

 gastropores (B, gp) ; the smaller are arranged more or less 

 irregularly round the gastropores, and are called dadylopores (dp). 

 The whole surface of the coral between the pores has a pitted 

 appearance. Sections (C) show that the entire stony mass is 

 traversed by a complex system of branched canals, which com- 

 municate with the exterior through the pores. The wide vertical 



