240 Dr. J. D. F. Gilchrist on 



atlvancod enihryos and larvrc — a good foi-tune wliich, liow- 

 cver, fis unexpectedly was not continued. The animals iu 

 this case were dead, hut by cutting longitudinal and trans- 

 veri^e sections of the branches numbers of embryos in an 

 ailvanced stage were seen, and one of these was actually 

 liatched out when being observed under the microscope. 

 Tills embryo was seen rotating actively within the egg- 

 capsule, and was so large that it was bent on itself, one end 

 sligiitly overlapping the other. 'ilie egg-capsule was 

 ruptured, and the embryo escaped, assuming the form of an 

 elongate ovoid larva, which ])rogressed actively over the 

 bottom of the containing glass by means of its ciliated 

 surface. Its shape and pigmentation were so characteristic 

 that no difficulty could subsequently arise as to the deter- 

 mination of free larva3 about this stage. Numerous eggs, 

 embryos, and a limited number of free larvaj were observed 

 in specimens procured subsequently, but no such fully deve- 

 lo))ed embr} in the egg as in this case. The identitication 

 ot the larval form does not, however, depend on this one 

 observation, as the characteristic pigment is assumed at a 

 much earlier stage in the embryo ; and free larvae similar to 

 the one observed were subsequently seen to come from 

 branches of the coenceciura under observation. It was found 

 ultimately that the best way to secure free larvee was to 

 siphon off tlie deposit at the bottom of the vessel in which 

 the specimen had been kept on board the trawler. On one 

 occasion many such larva3 were ibund, but, as it chanced to 

 be a very hot day, most of them died in a short time. 



The eggs occur in the tube-like cavity inhabited by the 

 animals, between which and the bottom of the tube there is 

 usually a l.irge ?])ace. One to six eggs occurred in this 

 position. They were each enclosed in a jjyriform egg-case, 

 drawn out at its narrower end into a short tubular part, 

 which ended abruptly in a somewhat flattened surface. This 

 flattened end was adherent to the side of the cavity. Some- 

 times the egg-capsules were found adherent to various parts 

 of the cavity, but fiequently were closely set in groups, their 

 bases being adherent to a somewhat restrict'^d area. No early 

 segmentation-stages were observed in the living ag^, though 

 subsequent examination of preserved material revealed 

 several, from the two-celled stage onwards. Some of the 

 earliest ciliated embryos were observed to be quite spherical 

 in shape and to rotate rapidly on themselves by means of 

 their cilia. Later stages were eloiigate and egg-shaped ; 

 one of these in the living condition mea.sured '55 mm. in 

 length and '24 mm. in greatest breadth. Numerous small 



