304 Southern Cross. 



of food yolk. The enibiyos are exceedingly irregular in form and lie 

 abutting one with the other — the end of one fitting into a socket in the 

 adjacent one, and they appear moulded to the contour of the chamber 

 by pressure due to the contraction of the mother animal. The 

 cavities containing the larvie appear to be continuous one with the 

 other at this stage, and thus form a continuous chamber lying in 

 the coelenteron surrounding the oesophagus. The invagination 

 opening is still efiectually closed to the outside. One of the larvie 

 when sectionized showed in a longitudinal section the invagination 

 to form the oesophageal opening (fig. 16), and transverse sections 

 revealed the presence of mesenteries, in an early stage of develop- 

 ment, arising from the body-wall and projecting into the solid 

 food yolk mass (fig. 17). Near the oral end twelve mesenteries 

 representing the first cycle are apparent, and as the sections approach 

 the aboral end these gradually increase in size, and between each 

 pair there appear slight indications, also in pairs, of the second 

 cycle. These larvie are from 1 cm, to 1 * 5 cm. in length and from 

 0"5 to, 1 cm. in diameter. 



Much older embryos than these, measuring from 1 cm. to 

 1*5 cm. in diameter, were found in brood chambers of both 

 U. sulcata and U. carlgrcni. Several of these specimens have 

 the brood chambers opening freely to the outside through the 

 original invagination opening, the contained embryos evidently 

 approaching the stage when they are set free and begin an inde- 

 pendent existence. One specimen of U. carkjreni had been killed 

 with one of the embryos escaping from the brood chamber (figs. 18 

 and 19), others of the contained embryos being also visible. The 

 opening extended almost half way round the body and showed 

 every appearance of further extension. On sectionizing these 

 older embryos they are found to possess three cycles of mesenteries, 

 rig. 20 represents a sixth part of a transverse section through 

 an embryo of corresponding age of U. sulcata, including a pair of 

 directive mesenteries. As there shown the first cycle of mesenteries 

 is complete and the second cycle is well advanced. The longitudinal 

 muscles are well developed and already give indication of the 

 strength of the adult structures. The parieto-basilar muscle is also 

 plainly visible. The third cycle has as yet no indication of muscle- 

 structure. From a close examination of the consecutive series of 

 sections I can make out the existence of twenty-four tentacles, of 

 which six appear to be larger than the others and to be in 

 communication with the intra-mesenteric spaces of the first cycle. 

 In all the embryos the oral disc and tentacles are entirely retracted 



