SCRUPOCELLARIA 363 



In the specimen from Pernambuco some zooecia agree very closely with the type, but 

 in others the branched spine has fewer tines and is less curved. 



Several species with similar scuta are known from the tropical region of the Atlantic. 

 Smitt (1872, pp. 13, 14) described three from Florida as S. cervicornis (Pourt.), 1 

 S. cornigera (Smitt) and S. pusilla (Smitt), and Waters described S. tridentata from the 

 Cape Verde Islands. Professor R. C. Osburn has found Smitt's three species at the 

 Tortugas, and has very kindly sent me specimens of S. cervicornis (3 5. 11. 26. 3) and 

 S. cornigera (3 5 . 1 1 . 26 . 2) and examined part of my material. S. pusilla 2 is represented 

 in the British Museum by some fragments from John Adams Bank, off Brazil (99 .7.1. 

 786, 787). Smitt's figures are accurate in all matters of importance, and there can be 

 no doubt that S.frondis is distinct from all three species. The bifid spine is a variable 

 feature and I have seen it in both S. cervicornis and S. cornigera (cf. Marcus, 1937, 

 pi. xi, fig. 26 A). S. frondis differs from both these species in the presence of the large 

 cervicorn spine, in the rounded scutum which never has the sharply pointed corners, 

 in the absence of lateral avicularia, and in the shape of the vibracula. The vibracular 

 chambers of S. cervicornis not only differ in shape, but are conspicuous in frontal view. 

 S. pusilla is a smaller, more delicate species. Its scutum is similar to that of S.frondis, 

 but it has no cervicorn spines, its vibracula are very different in shape, it has lateral 

 avicularia and its branches are more sinuous in outline. 



Professor Osburn tells me (in a letter, 1935) that his S. cervicornis (1914, p. 192) is 

 the true S. cervicornis Smitt, that its "cervicorn" spines would have been better de- 

 scribed as bifid, and that he has not before seen the species with the truly cervicorn 

 spine. 



S. cervicornis Verrill (1900, p. 593) from Bermuda may have been one of these 

 species, but in the absence of any mention of branched spines is probably not S.frondis. 



S. tridentata Waters possesses lateral avicularia, and enlarged trifurcate frontal 

 avicularia below the bifurcation, and has the spines arranged differently and all un- 

 branched. 



The vibraculum of S. frondis resembles that of S. bertholletii (Audouin), a species 

 whose rather larger zooecia have the same number of spines, but all unbranched, and 

 similar ovicells. The scuta of the two species are markedly different. The absence of 

 lateral avicularia is a distinction from specimens of S. bertholletii so far described, but 

 colonies from the Tortugas (Colman-Tandy Coll. 31. 12. 19.4; Professor Osburn, 

 35 . 1 1 . 26 . 1), and from Mozambique (1938 .5.2.4), which otherwise agree exactly with 

 that species, have none. 



Thornely has noted that the form recorded by her as S. frondis differs from Kirk- 

 patrick's species in the presence of marginal avicularia and the shape of the scutum. 

 These points are confirmed by examination of her specimen in the British Museum 

 (1936. 12.30. 173), but I have been unable to see any "tree-like" markings on the 



1 Not S. cervicornis Busk 1852a = .S. diadema (see Harmer, 1926, p. 375). 



2 Harmer (1926, p. 382) regarded S. pusilla as synonymous with S. spatulata (d'Orbigny), but as my 

 remarks are based entirely on specimens from the Western Atlantic, the locality studied by Smitt, I have 

 used the name S. pusilla in this discussion. 



