Perag. (Peragallo, 1921: 84, pl. 4, fig. 17)
Coscinosira antarctica Mangin (1915: 55–57, fig. 39; pl. 1, fig. 8)Actinocyclus excentricus Perag. (Peragallo, 1921: 75–76, pl. 6, fig. 6)Melosira dougetii Manguin (1960: 236, pl. 1, fig. 4)
Fryxell (1977: 96–100, figs 1–12), Johansen & Fryxell (1985: 158, figs 19, 20, 46–48), Syvertsen (1985: 116–118, figs 15–27), Fryxell & Johansen (1990: 99, pl. 11.9, figs 5, 6)
Cells usually in short chains, closely linked by threads extruded from strutted process; resting cells common in the chain. Valves circular, 23–60 µm diameter, markedly convex (Fig. a, b). Areolation radial to fasciculate; areolae 11–13 in 10 µm in the centre, increasing to 14–16 in 10 µm near the margin, loculate, opening externally via a simple foramen, internally via an evenly pored velum (Fig. b). Strutted processes 4 or 5 on the mantle, in clusters in the centre, often in 3 or 4 pairs of clusters, 5 in 10 µm on the single marginal ring. Marginal ring of strutted processes ranging from 7–9 in 10 µm in resting spore valves to 3 in 10 µm for rudimentary valves. External openings of resting spore strutted processes long trumpet-shaped tubes (Fig. d), absent in rudimentary valves (Fig. f). Labiate process solitary, just inside marginal ring of strutted processes; internal opening almost sessile and orientated along the radius. Occluded processes present on vegetative valves (Fig. b) in a marginal ring inside the marginal ring of strutted processes, 2–3 in 10 µm; external tubes of occluded processes ±straight, 2–7 µm long. Girdle with areolate valvocopula, a largely unstructured copula with or without a line of poroids adjacent to the valvocopula, and 2–4 unstructured pleurae (Fig. f) with the ligulae rotating to the right. For a complete description, see Fryxell (1977).
Inshore areas of Prydz Bay, near Davis Station, East Antarctica (e.g. Plough I., 68°32’S 78°00’E and Ellis Fjord, 68°37’S 78°00’E); type locality, Petermann I., Antarctic Peninsula, 65°09.8’S 66°32.7’W (Peragallo, 1921; not southern Argentina as suggested by Fryxell, 1977); South Orkney Is. (Fryxell, 1977); Weddell Sea (Syvertsen, 1985); coastal waters off Syowa Station, East Antarctica (Ishikawa et al., 2001).
Slight differences were observed between material examined in this study and previous descriptions. Strutted processes seen by us were often in 3 or 4 pairs of central clusters; Syvertsen (1985) reported 1 or 2 cluster pairs, and Fryxell (1977) reported 3 or 4 simple clusters. Occluded processes present on vegetative valves in this study were 2-3 in 10 µm, with external tubes straight and 2-7 µm long. Syvertsen (1985, figs 15, 21-23) illustrated reflex-angled tubes 10-12 µm long.The resting spore is similar to the valve of T. gerloffii (as mentioned by Syvertson, 1985: 119), but it differs from the latter in not having the labiate process in the same line as the marginal ring of strutted processes; in having trumpet-shaped, external tubes to the strutted processes rather than irregularly linear ones; and a lower areolar density. The colony form illustrated by Rivera (1981: figs 168, 179) for T. gerloffii shows cells attached by a central strand two or more times as long as the pervalvar dimension of the frustule compared to one or less for the Weddell Sea material and no between-cell distance for the Prydz Bay material of the current study.