?2 DISCOVERY REPORTS 



164 m. St. WS 785 : Patagonian Shelf, 150-147 m. St. WS 799: Patagonian Shelf, m. St. WS 800: 

 Patagonian Shelf, 137-139 m. St. WS 814: Patagonian Shelf, 112-119 m. St. WS 816: Patagonian 

 Shelf, 150 m. St. WS 818: Patagonian Shelf, 272-278 m. St. WS 819: Patagonian Shelf, 313-329 m. 

 St. WS 833 : Patagonian Shelf, 38-31 m. St. WS 834: Patagonian Shelf, 27-38 m. St. WS839: 

 Patagonian Shelf, 503-534 m. St. WS 841 : Patagonian Shelf, 110-121 m. St. WS 847: Patagonian 

 Shelf, 51-57 m. 



Colony. The shape of the colony varies greatly but the causes of variation are only partly known. 

 In many specimens which are fully developed and more or less intact there is a stout main stem which 

 in life appears to have lain on the sandy bottom either free or loosely fixed. This stem fulfils the 

 function of a creeping stolon and gives rise to one or a few bunches of upright stalks, each stalk bearing 

 a single head (PI. II, figs. 3,4). A true rooting system does not seem to be developed, but in other 

 respects the habit is very like that of the Australian species S. tenuicaulis (Herdman) (see Herdman, 

 1899, pi. Dist. 1, fig. 2; Brewin, 1953. Text-fig. iB). 



Many specimens, however, have a single stalk bearing one head (e.g. St. WS 90) ; these are perhaps 

 broken pieces of complex colonies, for they have the same characters of stalk and head. 



A third type of colony consists of a simple or slightly branched stalk attached basally to a solid 

 object (e.g. St. WS 95). 



Many unattached heads were also taken in the collections, and these will be referred to later 

 (see p. 73). 



Finally there is the headless condition represented either by isolated hard orange-coloured stalks 

 or by whole complex colonies completely lacking zooids (e.g. St. WS 800) (PI. II, fig. 5). 



In well-developed complex colonies the stalks reach 6 cm. in length, occasionally more, and the 

 heads up to 2-5 cm. long and i-o cm. wide. 



Two groups of specimens from South Georgia (St. 45, St. 145) have a short fleshy stalk attached to 

 an alga, and a head at least twice as long as the stalk (PI. II, fig. 8). As the structure of the head, the 

 zooids and their arrangement, and the larvae all agree closely with the other examples of 5. sigillinoides 

 in the collection, I am including these specimens within that species, although not entirely satisfied 

 with this course. They are very like the specimens also from South Georgia figured by Arnback (1950, 

 fig. 26) as S. sigillinoides, and like Herdman's (1886) type of his species 5. quoyi from Kerguelen, and 

 also like Michaelsen's (1907) specimens from South Georgia, which he named S. quoyi var. zschaui. 

 I regard S. quoyi as a synonym of S. sigillinoides. 



In the ' Discovery ' colonies, which are preserved in alcohol, the head is soft and usually yellow-grey. 

 It is almost cylindrical, and rounded or tapered at each end. The stalk, although wide at the upper 

 end, is sometimes constricted where it joins the head, and is narrowest at the base. It is firm, buff 

 or dull orange in colour, and typically marked with fine annular ridges. The basal stem is similar to the 

 stalk but is stouter and usually shorter. 



As I have already remarked, I disagree with Brewin (1953) in her description of the common cloacal 

 openings in the genus. In well-preserved specimens of S. sigillinoides which I have examined, the 

 apex of the head has a single common cloacal opening, all the more clearly marked because the test 

 bears a series of teeth round its border (Text-fig. 19, t.t.). The opening is ring-shaped as viewed from 

 the apex of the colony owing to the projection of the central core of common test of the head. The 

 individual cloacal canals, each serving a double row of zooids, lead into the terminal opening, and not 

 directly to the surface of the colony. 



Zoom. The atrial languet has been described as 'of varying length ' (van Name, 1945), and 'some- 

 times fringed with a number of lobes ' (Kott, 1954). I have found that differences in the atrial languet 

 are related to the position of the zooid in the colony. All zooids except the terminal one in each row 



