XCV1 INTRODUCTION. 



this stage (according to Joliet) it is fertilized by passing 

 spermatozoa, and immediately enters on its further deve- 

 lopment. At this time its behaviour is in many respects 

 that of a polypide. Thanks to the muscles which it has 

 borrowed, it now rises to the entrance of the cell and now 

 retreats to its recesses, until, having assumed its perfect 

 larval form, it passes through the sheath into the water. 

 The " auxiliary polypide " has discharged the functions 

 of an ooecium, and has both supplied the ovum with a 

 brood-chamber and a way of escape. In its leading par- 

 ticulars, this marvellous history rests on the authority of 

 three able observers. 



The true marsupium seems to be confined to the Chei- 

 lostomata, and even amongst them is by no means uni- 

 versal. In other sections of the class ova are frequently 

 developed in special receptacles, and not in the zocecia : 

 to these I have given the names yoncecium and gonocyst*. 

 Of the former we have a good example in the genus 

 Crisia f ; it also occurs amongst the Cheilostomata, and 

 in one family at least (Alcyonidiidce) of the Ctenostomata. 

 The gonocyst has only been noticed in certain Cyclosto- 

 matous genera; it is probably only a modification of 

 the goncecium. Of the history of these structures, 

 however, we know little ; they remain to stimulate and 

 reward further research. 



I do not propose to follow the ovum through the stages 

 of its development ; nor shall I enter minutely into the 

 numerous modifications of the larval form. To do so 



* See Terminology, page iii. 



t Smitt has investigated the goncecium of Crisia, and considers the pro- 

 duction of ova in it to be asexual. ' Om Hafsbryozoernas Utveckling,' &c. 

 (1865), p. 19. For a critical notice of Smitt's observations, see Barrels, 

 ' Embryologio ' (1877), p. 59. 



