CTENOPHORES OF THE ATLANTIC COAST OF NORTH AMERICA 51 



begin to give rise to lateral branches, and the papillae flanking the apical 

 pole-plate have greatly increased in number. The steel-blue color of 

 the gelatinous substance and the pink of the meridional vessels begin 

 to appear at a later stage. 



When the young animal from Tortugas, Florida, is 26 mm. long, 

 and 18 mm. wide in the broad, and 7 wide in the narrow axis of the body, 

 the oral forks of each head side of the body remain separate, and do not 

 fuse across the narrow sides. The two broad sides are, however, con- 

 nected one with another by an anastomosing network from the 4 meri- 

 dional canals which flank the 2 narrow sides of the body. Hence the 

 network of the narrow sides anastomoses and forms a complete circum- 

 oral circuit before any anastomosing is seen on the broad sides of the 

 animal. In this young animal there are as yet no fusions between the 

 network and the 2 paragastric canals. The animal is now decidedly 

 pink in color. 



Beroe ovata is abundant among the West Indies and along the coast 

 of the United States as far north as Chesapeake Bay. This species is, 

 I believe, identical with the Mediterranean form, although the Florida 

 specimens are not so intensely pink as the Mediterranean ctenophore. 



Beroe shakespeari Benham, from New Zealand, is, I believe, identical 

 with B. ovata, for it bears so close a resemblance to the form of B. ovata 

 from Florida that were the two found together one could not separate 

 them specifically. 



Form, size, and color are practically worthless as specific criteria 

 in the genus Beroe. 



Beroe ovata has been the subject of some interesting experimental 

 work. Driesch and Morgan, 1895, found that if the egg be cut in two 

 at the 2 -cell stage, each blastomere gives rise to a half embryo, which, 

 however, has 3 entodermal cells instead of 2, as one would expect. Only 

 4 rows of cilia are developed. Hence each blastomere segments almost 

 as if still a part of the whole egg. The same statement is true of the one- 

 fourth blastomere obtained by separating the blastomeres of the 4-cell 

 stage, for each develops into a one-fourth larva with, however, 2 ento- 

 derm cells intead of i. It will be recalled that Chun, 1892 (Festschrift 

 fur Leuckart, pp. 77-108), showed that an isolated blastomere of the 

 2-cell stage in Bolinopsis produces a half-larva. 



Driesch and Morgan cut in two eggs of B. ovata which had been 

 fertilized but had not yet segmented, and the part containing the nucleus 

 divided in general like that of an isolated blastomere, giving rise to a 

 partial embryo. Hence the incomplete development is due to the proto- 

 plasm, not to the nucleus. 



Fischel, 1897 (Archiv Entwickelungsmechanik, Roux, Bd. 6, p. 109, 

 Taf. 6), has continued these studies of Driesch and Morgan. He finds 

 that if we isolate the blastomeres of B. ovata in the first, second, and 

 third cleavages, larvae develop with 4, 2, and i row of combs, respec- 

 tively. Fischel also fo\md that if we remove some of the blastomeres at 

 any segmentation stage, partial larvae result, which are the more nearly 

 complete the smaller the removed part. Hence the resiilt is the same 

 whether the nucleus has divided or not, the nucleus manifesting r.o 



