74 DISCOVERY REPORTS 



developing zooids. Regenerating heads on old stalks were taken in December, March and May, but 

 most often in June and July. We may conclude that loss of the head follows breeding and is in turn 

 followed by regeneration of a new head, the speed of regeneration perhaps depending on water 

 temperature, as in colonies of Aplidium in northern temperate waters. 



In the related genus Hypsistozoa, Brewin (1956a) has shown that, accompanying and following the 

 production of larvae, the distal part of the head of the colony is resorbed. Meanwhile, new zooids at 

 the junction of the stalk and the old head are forming a new head. This is a process very similar to 

 that occurring in Sycozoa. 



It seems curious that one species should show two habits so different as the single stalk and the 

 branched complex form. Perhaps some of the apparently simple forms are stalks broken off from 

 more complex colonies, but it is also possible that the simple and the complex forms represent 

 different phases of the life cycle. This is suggested by the growth of several young heads instead of 

 one, from the end of the old stalks, a process seen in several specimens (PI. II, fig. 7). Perhaps the 

 colonies which develop from settled larvae have single stalks, which, after breeding and loss of the 

 head, may develop a group of new heads from the original stalk, thus giving rise to a complex colony, 

 the original stalk becoming the stout basal stem of the new colony. 



Distribution. Owing to the confusion regarding species it is difficult to state exactly the distri- 

 bution of S. sigillinoides, but it is certainly widely spread in both western and eastern parts of southern 

 and Subantarctic waters (Falkland Islands, Patagonian Shelf, Tierra del Fuego, Kerguelen, Macquarie 

 Island, Tasmania, South Australia) and the Antarctic (South Georgia, Graham Land, Kaiser 

 Wilhelm II Land, Enderby Land, MacRobertson Land). 



Sycozoa georgiana (Michaelsen) (PI. Ill, fig. 5) 



Colella georgiana Michaelsen, 1907, p. 62, pi. 1, fig. 6; pi. 3, fig. 15. 

 For synonymy see van Name, 1945, pp. 154-5. 



Occurrence. St. 39: S. Georgia, 179-235 m. St. 140: S. Georgia, 122-136 m. St. 149: S. Georgia, 

 200-234 m. ?St. 363 : S. Sandwich Islands, 329-278 m. ?St. WS 33 : S. Georgia, 130 m. St. MS 14: 

 S. Georgia, 190-110 m. St. MS 68: S. Georgia, 220-247 m. 



Colony (PI. Ill, fig. 5). The form of the colony and the shape of the head are the main features which 

 distinguish this species from S. sigillinoides. Colonies are attached to some solid object over which 

 spreads a system of creeping branched stolons giving rise at intervals to single or grouped stalks. Each 

 stalk bears a characteristic head which is pear-shaped or almost globular, and never, in the specimens 

 examined, long as in S. sigillinoides. The heads in the ' Discovery ' material are not laterally compressed 

 as in Michaelsen's (1907) description. Usually the head is as wide as long, and sometimes is wider. 

 One of the largest specimens has a stalk of 4-7 cm. and a head 0-7 cm. long by 0-9 cm. wide. The stalk 

 and the head are of the same pale grey colour and the same smooth rather soft consistency, whereas in 

 S. sigillinoides the head and stalk differ from each other. Zooids are generally absent from the lower 

 part of the head, and the oral openings are confined to a wide zone round the single terminal common 

 cloacal opening. Sluiter (1932) found that the zooids were not arranged in rows, and Michaelsen 

 (1907) did not mention any regular arrangement. At first sight this also seems to be the case in the 

 'Discovery' specimens, but careful examination shows that the zooids are in short rows each with 

 a few zooids. 



Zoom. The zooid seems to present no characters by which it can be distinguished from S. sigil- 

 linoides. Certainly it cannot be separated by the number of branchial stigmata in a row, as Kott 

 (1954, p. 155) has done in her key. Sluiter (1932) stated that in young zooids the stigmata number ten 

 to twelve in each half row, but in fully developed zooids there are up to twenty, and in the ' Discovery ' 





