39 8 DISCOVERY REPORTS 



27) whose avicularia have the more typical wide range of size. This South Georgian 

 specimen has unusually numerous spines. In other colonies the maximum number is 

 six (two close together on the outer corner of the zooecium, then three spread along the 

 back, and one near the axis), but in the colony from St. WS 27 there are commonly six 

 along the back, making a total of nine, fairly evenly spaced, spines. Other specimens 

 from South Georgia have the typical number of spines, and zooecia of average 

 length. Although a certain amount of variability is thus to be observed, the variation 

 in the different structures does not seem to be correlated so as to present distinct 

 forms. 



Osburn (1940, p. 1) described C. antillea from 732 m. in the West Indian region. 

 He noted that it differs from C. pectogemma in the form of the trumpet-shaped avicu- 

 laria, in the presence of avicularia of a second type, in the number and distribution of 

 the spines, and in the ovicells. As no mention is made of the bulbous swelling on the 

 avicularium, this is presumably absent and constitutes a further difference. On the 

 other hand the variation in the spines and trumpet-shaped avicularia in my material of 

 C. pectogemma suggest that the differences in these structures may not be important. 

 In particular some of the avicularia of the material from Tristan appear to be very 

 similar to Osburn's. In the absence of figures, I am not clear what constitutes the 

 difference in the ovicells. C. antillea thus seems to be chiefly distinguished by the 

 presence of a second type of avicularium, by the absence of the bulbous swelling and 

 by some undefined difference in the ovicells. 



Two young colonies of C. pectogemma with ancestrulae were taken off Oates Land 

 (St. TN 194). There are no avicularia but the more distal zooecia are otherwise charac- 

 teristic. The proximal ones have more numerous spines, and differ from the corre- 

 sponding members of the young colonies of C. polymorphs more in size than in shape. 

 The ancestrulae have a small basal bulb which in the older specimen has formed a 

 rootlet. The rest of the ancestrula is separated from the bulb by a joint, and consists of 

 a very long and slender tube and an expanded part bearing the opesia and eleven or 

 twelve evenly spaced spines. One rootlet has run down the back of the older colony and 

 its swollen tip lies beside the bulb of the ancestrula which it somewhat resembles. 



These ancestrulae differ from those of C. polymorphs described on p. 400 in their very 

 much smaller bulbs and in the much longer tubular part (cf. Fig. 28 A and C). The 

 septum, noticed in C. polymorpha, is indistinct but probably present. The shape of the 

 bulb appears to be related to the substratum. The younger specimen (Fig. 28 C) is 

 moulded to the lateral wall of a zooecium of Himantozoum antarcticnm and is irregular 

 in shape. The older one was found unattached. Its shape is more regularly oval. 



2. Cornucopina conica Harmer. 



Cornucopina conica Harmer, 1926, p. 428, pi. xxix, figs. 1-3. 



Station distribution. Not represented in the Discovery collections. 



Geographical distribution. Malay Archipelago, 924 m. and 1158 m.; off Crozet Islands, 

 2928 m. (Harmer). 



