4 20 DISCOVERY REPORTS 



the Gorgona specimen, are less humped dorsally and have a slightly longer beak. The 

 figures show that the specimens from the Cape Verde Islands (Kluge), and St Helena 

 (Marcus) also differ from the typical form in these respects. The specimens recorded by 

 Osburn and Thornely were not figured and I have not seen them, but Osburn mentions 

 that his material from the Tortugas was without marginal spines. 



In B. cupulariensis (Osburn, 1914, p. 190; Harmer, 1926, p. 419), which shows some 

 resemblance to B. intermedia, there is a much more complete series of marginal spines, 

 and the last one on each side is usually directed outwards as in Osburn's fig. 7. The 

 spine beside the avicularium is absent, and the connecting tubes are shorter than in 

 B. intermedia. 



Klugella gen.n. 



Genotype. Flustra echinata Kluge, 19 14, p. 658, pi. xxxii, fig. 6, text-fig. 38. 



Definition. Colony flustrine consisting of broad, irregular lobes, with kenozooecia 

 filling angles between marginal zooecia. Zooecia not forked proximally, or with a mere 

 notch. Distal wall oblique with a group of uniporous septula. Lateral rosette-plates 

 1-4, in distal half of wall, multiporous. Opesia extensive. No cryptocyst. A small 

 proximal gymnocyst. Avicularium jug-shaped, attached rigidly to proximal gymnocyst, 

 with multiporous rosette plate at base. Rostrum directed transversely. Ovicells hyper- 

 stomial, entooecium calcified, ectooecium wholly or partly membranous. Avicularium 

 not fused to ovicell. 



Remarks. The definition of this genus has been framed to include Flustra crassa 

 Busk (1884, p. 53 ; see Klugella buski, p. 422) from Kerguelen, which resembles the geno- 

 type in its avicularia, and short, more or less spindle-shaped marginal kenozooecia, and 

 in the general structure of its ovicells. It differs in the bilaminar colony and in the 

 absence of any overlapping of the zooecia frontally. The indentation of the basal inser- 

 tion of the proximal wall (Kluge, text-fig. 38 c) which appears to be constant in Klugella 

 echinata and presumably represents the forked proximal end of the zooecium of many 

 other Bicellariellidae, is absent in K. buski. K. buski has less numerous rosette-plates, 

 there being only one or two 1 in the distal half of each lateral wall, as compared with 

 three or four in K. echinata. K. buski also differs in the absence of basal spines and (in 

 the limited material available) of rootlets. Marginal spines are less developed in 

 K. buski, but the two pairs of distal ones much resemble the two distal pairs of K. 

 echinata. The mandible of the avicularium is shorter in K. buski although the spout-like 

 beak is much alike in both species (see Fig. 36 C, Busk's figure being misleading). The 

 ovicells of both species are essentially alike in structure, having a small area of attach- 

 ment to the distal zooecium, a calcareous entooecium and an ectooecium of which the 

 greater part, at least, is membranous. Those of K. echinata are more shallow and, as far 

 as my experience goes, their only sculpture is a faint radial striation. 



The boat-shaped zooecia of K. echinata, with their row of marginal spines, resemble 

 those of various species of Beania and the basal spines are like those of B. pulchella (see 



1 Waters (1896, p. 284) says there are four (i.e. two in the distal half of the lateral wall) and figures one in 

 the distal half. Both numbers are found in the type. 



