Actiniae. 305 



and hidden, as is the case witii most of the adult specimens. This 

 denotes the presence of an ah^eady functional sphincter, a figure 

 (fig. 21) of which is shown taken from a longitudinal section as well 

 as functional mesenterial muscles. There is little or no food yolk 

 now left in the ccelenteron, another fact which denotes the early 

 approach of a separation from the parent. 



There are thus four distinct stages in the appearance of the brood 

 chamber and its contents, represented by the specimens of these two 

 Antarctic species. The earliest condition shows the developing 

 invagination of the body-wall of the parent, forming cavities from 

 2 to 3 mm. in depth, but absolutely void of contents, and freely open 

 to the outside through the invagination openings. Then comes the 

 second stage, with the chamber considerably increased in size 

 (6-11 mm. in depth), completely closed to the outside and now 

 containing developing embryos in the solid morula stage. Then 

 follows the condition shown by the specimen of U. sulcata (the only 

 specimen found illustrating this stage), with the embryos greatly 

 advanced in development, possessing a three-layered body-wall, 

 developing mesenteries, invagination to form oesophagus, and filled 

 with a dense mass of food yolk, but with the brood chamber still 

 completely closed to the outside. The fourth and last condition 

 shows the embryo just about to be set free, with three cycles of 

 mesenteries (one cycle complete), well-developed muscles, twenty -four 

 tentacles, all or almost all the food yolk absorbed, and the chamber 

 generally freely open to the outside through the original invagination 

 opening. From the fact that a specimen of U. carlgreni, containing 

 embryos of this last stage in brood chambers open to the outside, 

 possesses in addition, between the older brood chambers, a series of 

 new invaginations corresponding to the first stage described, there is 

 reason to believe that each brood chamber only serves for one 

 embryo, and that on its escape the walls of the chamber atrophy and 

 new chambers are formed for the next brood. 



Dr. Carlgren (1 p. 237) suggested that possibly the occurrence of 

 these special brood chambers may be connected with an external 

 fertilization of the ova, but, while not denying the possibility of such 

 in these species, there is no evidence in the appearance of either his 

 specimens or the ' Southern Cross ' specimens of any departure from 

 the usual internal impregnation. In all probability, after fertilization 

 in the body cavity, the eggs in early stages of cleavage are transferred 

 directly, by the aid of the tentacles, to the brood chambers, the 

 entrances to whicli are soon afterwards closed. 



From the fact tliat, so far as yet observed, the possession of these 



X 



