132 THE ANATOMY OF IXVERTEBRATED ANIMALS. 



is provided with a pneumatophore. This is a dilatation, into 

 which the ectoderm is invaginated, so as to form a receptacle, 

 which becomes filled with air and sometimes has a terminal 

 opening, through which the air can be expelled (Fig. 13, 4). 

 Tt is sometimes small, relatively to the hydrosoma (Agalma, 

 PAysophora) ; sometimes so large (Athorybia, Fig. 24 ; Phy- 

 salia, PorpUa, Velella), that the whole hydrosoma becomes 

 the investment of the pyriform or discoidal air-sac ; while the 

 latter is sometimes converted into a sort of hard inner shell, 

 its cavity being subdivided by septa into numerous chambers 

 (Porpita, Velella). 



Nectocalyces may be present or absent in the PhysopJio- 

 ridoe. When present, their number varies, but they are con- 

 fined to the region of the hydrosoma which lies nearest to the 

 pneumatophore. 



In the great majority of the Jlydrozoa, the ovum under- 

 goes cleavage and conversion into a morula, and subsequently 

 into a planula, possessing a central cavity inclosed in a double 

 cellular wall, the inner layer of which constitutes the hypo- 

 blast, and the outer the epiblast. 



In most Hydrophora the ciliated, locomotive, planula be- 

 comes elongated and fixed by its aboral pole. At the oppo- 

 site end, the mouth appears and the embryo passes into the 

 gastrula stage. Tentacles next bud out round the mouth, 

 and to this larval condition, common to all the Hydrop>hora, 

 Allinan has 2*iven the name of Act inula. 



Generally, the embryo fixes itself by its aboral extremity 

 at the end of the planula stage ; but, in certain Tubularidce, 

 while the embryo is still free, a circlet of tentacles is devel- 

 oped close to the aboral end ; and this form of larva differs 

 but very slightly from that which is observed in the Disco- 

 phora. 



In the genus Pelagia, for example, the tentacles are de- 

 veloped from the circumference of the embryo, midway be- 

 tween the oral and aboral poles ; but it neither fixes itself 

 nor elongates into the ordinary actinula-form. On the con- 

 trary, it remains a free-swimming organism, and, by degrees, 

 that moiety of the body which lies on the aboral side of the 

 tentacular circlet widens and is converted into the umbrella, 

 the other moiety becoming the hydranth, or " stomach," of 

 the Medusa. 



In Luiemaria, it is probable that the larva fixes itself be- 

 fore or during the development of the umbrella, and passes 



