DEVELOPMENT OF THE CTENOPHOEAc 91 



sacs appended to the central lacuna, and opening into it, and 

 are to become the deep longitudinal or gastric vessels of the 

 animal. 



Nearly at the same time two other masses, exactly similar to 

 the former, and also containing an extension of the central lacuna, 

 are found opposite to one another on the two remaining sides of 

 the alimentary cavity ; so that this cavity is novv^ surrounded by 

 four sacs. 



The two last-formed sacs now become divided, each into two, 

 the halves diverging more and more widely from one another, and 

 becoming more and more superficial, until we find them in contact 

 with the peripheral or ectodermal layer of the embryo, each sac 

 here corresponding to a pair of meridional swimming-bands and 

 to the interval between them. The final dichotomy of the sacs 

 now takes place, each of the last-mentioned dividing into two, from 

 the jDoripheral towards the central side, leaving each half imme- 

 diately under a meridional swimming-band ; so that every swim- 

 ming-band has now a sac to itself. The gradual conversion of the 

 wide sac-like vessels into the narrow canals of the adult is easily 

 understood. About this time the central cavity, into which the 

 vessels whose development has been just described open, and 

 which corresponds to the so-called funnel of the adult, has sent 

 off two branches, which run towards the aboral pole, and form 

 the canals which there open externally, one on each side of the 

 sense-body. 



Up to this point the embryo has been confined within the 

 coverings of the egg ; but in the stage next described the young 

 Bero'e was free. The meridional series of swimming-plates had 

 extended further towards the mouth. The mouth was surrounded 

 by a circular canal, into which the two deep longitudinal vessels 

 (gastric vessels) had opened. The circular canal was probably 

 developed from tlie oral extremities of these vessels as T-like 

 branches, by whose inosculation the ring became completed. Four 

 out of the eight meridional vessels (those, namely, which corre- 

 sponded to the narrow sides of the stomach) had also opened into 

 the circular canal ; but the other four meridional vessels, though 

 extending for some distance beyond the oral extremities of the 

 series of swimming-plates to which they belonged, still terminated 

 towards the mouth, each in a blind extremity. The eight meri- 

 dional vessels had begun to send out their caecal offsets, as yet few 

 in numbet and simple. In a still further stage, the animal now 



7* 



