SCRUPOCELLARIA 361 



In the material from the Magellan region the marginal avicularia are all of the same, 

 rather small, size, and in the New Zealand specimen the larger ones are not so large as 

 usual. Both these specimens have the cryptocyst smooth, and are rather slender. 

 Ovicells are not present. 



S. ornithorhyncus is distinguished from British specimens of S. scrupea by several 

 small but definite differences. The marginal avicularia are farther from the distal end 

 of the zooecium in S. ornithorhyncus, often being entirely proximal to the spines and 

 always extending as far back as the proximal end of the opesia, or farther. The fertile 

 zooecia of S. scrupea have a smooth distal wall. Other differences are less definite. 

 S. scrupea has as a rule stouter branches, with less markedly serrated borders (because 

 of the smaller marginal avicularia), and more zooecia in the internodes. Its scutum is 

 blunter distally and I have seen no specimen with more than three outer spines. In both 

 species the joint may just touch the proximal border of the opesia of the outer zooecium, 

 but in S. scrupea it sometimes passes through the proximal part of the opesia, whereas 

 in S. ornithorhyncus it is commonly in the proximal gymnocyst at a distance from the 

 opesia. 



The supposed Australian specimens of S. scrupea in the British Museum, including 

 those in the Bracebridge Wilson collection (97 .5.1. 219, 220) which probably represent 

 the form called S. scrupea by MacGillivray, belong to S. ornithorhyncus, and it seems 

 possible that some other published records of S. scrupea may be based on this species. 

 For example, S. scrupea Marcus (1937, p. 56 ; 19386, p. 208), from Brazil and St Helena, 

 may be S. ornithorhyncus. Both specimens had four outer spines and the figure (1937, 

 pi. xi, fig. 27 A) shows most of the marginal avicularia in the more proximal position. 

 On the other hand, the character of the distal wall of the fertile zooecium does not 

 appear, and the branch is as stout as those of S. scrupea. No evidence is available about 

 the specific characters of S. scrupea Stach (1937, p. 379). Harmer (1926, p. 382) has 

 shown that S. scrupea Philipps is a synonym of S. spatulata (d'Orbigny). 



2. Scrupocellaria frondis Kirkpatrick. Fig. 18 A-D. 



Scrupocellaria frondis Kirkpatrick, 1890, p. 504, text-fig. 1. 



not Scrupocellaria frondis Thornely, 1912, p. 140. 



Station distribution. Ascension Island: St. 1. 



Geographical distribution. Fernando Noronha ; Pernambuco (Kirkpatrick) ; Ascension Island 

 Discovery) ; Tortugas (3 1 . 1 2 . 9 . 7). 



This species, originally described from a few fragments, may usefully be redescribed 

 here. 



Description. Zoarium branches rather straight and not widely divergent. 



Zooecia (Fig. 18 B) smaller than those of Scrupocellaria bertholletii. Opesia oval, 

 occupying about one-half length of zooecium, filled by scutum, no cryptocyst. Lumen 

 of scutum usually constricted or closed at a point near stalk, with three or four recurved, 

 undivided branches on each side. Spines three external, two internal, and, on non- 

 fertile zooecia, one distal. Proximal external spine curved across aperture, cervicorn 



