4 oo DISCOVERY REPORTS 



avicularia are small and common in C. infundibulata, gigantic and rare in C. polymorpha. 

 In addition, C. polymorpha has the small curved frontal avicularia figured by Kluge. 

 Harmer's comparison (1926, p. 429) is misleading because, the stalked avicularia of 

 C. polymorpha not being known at the time, he contrasted the positions of the stalked 

 avicularia of C. infundibulata and the frontal ones of C. polymorpha. 



As can be seen from Plate VII, fig. 3, C. polymorpha grows luxuriantly. The photo- 

 graphed colony, which is the larger of two from the same station, is 13 cm. long, and 

 when held up by the roots forms a mass more or less circular in transverse section and 

 7 cm. in diameter. Both these colonies terminate basally in a tangle of rootlets, with 

 small stones and grit. 



A complete young colony of Cornucopina found among the material of C. polymorpha 

 from St. 175 (Bransfield Strait, 2 March 1927) has neither avicularia nor ovicells, but 

 agrees well with that species in the shape of its zooecia, and in the number of spines on 

 the more distal ones, those formed first having, as usual, more numerous spines. The 

 colony was attached by a bulbous base fitting into the axil of the forked Cyclostomatous 

 Polyzoan on which it was growing (Fig. 28 A). One end of the bulb is cut off by a 

 curved, transverse septum. A bifurcate rootlet springs from this end of the bulb, and 

 a short, rather thick-walled projection, originating just above the rootlet, gives rise to 

 a long tube, expanding to a zooecial chamber, and separated from the projection by a 

 joint like that separating the tubular and proximal parts of the other zooecia (cf. Harmer, 

 1926, p. 423). The tubular part of the ancestrula has a ring of thickening in its 

 wall suggesting the beginning of another transverse septum. The expanded distal part 

 resembles that of the other zooecia in shape, but has nine spines, evenly spaced in an 

 obliquely transverse row. Successive zooecia have fewer spines, though they continue 

 to be evenly spaced ; and although the colony consists of forty zooecia it is only near the 

 tips of the branches that the distinction between an outer group of spines and a single 

 one near the middle line appears. Five zooecia have produced rootlets which, running 

 down the basal surface and encircling the first zooecium, anchor the colony. A second 

 young colony from the same station consists of a bulb and one zooecium only. These 

 agree exactly with those already described, except that the zooecium has only eight 

 spines. 



The form described by Calvet from Biscoe Bay 1 as C. dubitata resembles C. poly- 

 morpha and C. infundibulata in having special ovicell-bearing zooecia, and in the 

 presence of both trumpet-shaped and stalked basal avicularia. In the large size of the 

 stalked ones, it resembles C. polymorpha, which is also the more probable geographi- 

 cally. Calvet suspected that his material had been decalcified in preservation, and one 

 finds that decalcified specimens of C. polymorpha very closely resemble his figure. This 

 might also account for the inflated appearance of the ovicell, which seems less reduced 

 than that of C. polymorpha, though smaller than the well-developed hyperstomial ovicell 

 of C. infundibulata. 



1 This must be in the neighbourhood of Biscoe Island, off Graham Land, which was the region visited 

 by the expedition, not Biscoe Bay, King Edward VII Land. 



