Diagnosis. Shell oblong-ovate, somewhat laterally com-pressed, apex recurved, approximately 0,20-0,25 of length from posterior end, moderate to high (h/l = 0,56-0,75), Anal foramen on anterior face level with apical whorls, lanceolate. Sculpture of predominantly radial ribs, of two alternating orders, numbering 35-45, of which approxi¬mately 20 are first order, microsculpture of radial rows of minute intritacalx pits. Interior with broad U-shaped septum underlying foramen, no lateral buttresses. White. Length up to 13,4 mm.
Description. (NM material) Shell with basal outline oblong-ovate, somewhat laterally compressed (w/1 = 0,62-0,77); apex recurved, approximately 0,20-0,25 of length from posterior end; protoconch and early teleoconch displaced to the right and tilted downwards; apex moderate to high (h/l = 0,56-0,75) (0,82 fide Thiele); anterior face pro-gressively more convex toward the apex, posterior face straight or slightly concave; base flat. Anal foramen on anterior face level with apical whorls, selenizone expanding and deepening rapidly, with low, curved growth-lines or weak lunulae; a single Iunule well developed where seleni-zone meets internal septum; foramen lanceolate, broadest above internal septum in apical view. Sculpture of predominantly radial ribs of varying strength, with a small number of weak concentric growth-lines; ribs smooth, of two generally alternating orders with additional finer inter-mediaries, 35-45 first and second order ribs (of which 18-20 are first order); a single anterior rib in line with foramen is frequently broader than the others, and may be shallowly bifid. Microsculpture of many rows of small, circular, intritacalx pits, most obvious in intervals, but also present on sides of ribs (rarely on crests). Ventral margin crenulate, size and position of crenules correlating with external ribbing, sometimes absent in large specimens. Interior glossy, with many rows of microscopic pores cor¬responding to intritacalx pits on exterior. Internal septum broadly U-shaped in section with distinct growth-lines when viewed through the foramen, no side buttresses; a shallow groove runs from the foramen toward, but rarely reaching, the anterior margin. White, slightly translucent. Protoconch puncturelliform, projecting to the right. Dimensions. Holotype: 5,5 x 3,5 mm, height 4,5 mm (fide Thiele), length 5,5 mm, height 5,0 mm (fide von Martens). Largest NM specimen 13,4 x 8,6 mm, height 7,8 mm. Radula and external anatomy. Unknown.

Although Thiele compared his PunctureUa capensis in some detail with P. fastigiata (A. Adams, 1853), with which it was confused by von Martens (1904), he gave little other descriptive data. Kilburn (1978) commented on the apparent discrepancy between Thiele's description and figure regarding rib strength. The present material, however, shows considerable variation in this respect and the above detail now seems of little importance. Alternation of first and second order ribs is obvious in some specimens and less so in others, although it is generally clearer in earlier stages. Likewise some specimens are more strongly ribbed with a correspondingly crenulate margin, while others possess only weak ribs and an almost smooth margin, Farfante (1947) noted similar variation in P. noachina (Linne, 1771).
Barnard (1963), unaware of Thiele's description of P. capensis, recorded the southern African species as P. noachina. The present species differs from P. noachina in possessing slightly fewer ribs, more numerous intritacalx pits and a broader internal septum, and in always lacking septal buttresses, even in adults. Furthermore, Farfante (1947) and Powell (1951) considered P. noachina to be restricted to the northern Atlantic, and the latter author synonymized the southern noachina-type taxa with P. conica (d'Orbigny, 1841). Thiele's figure of P. conica indicates it to be much taller than the present species (though this is not borne out by the measurements given by Powell). P. analoga von Martens, 1904, from Kerguelen Island, appears similar to P. capensis, but Thiele (1919) synonymized it with P. noachina and made no reference to it with regard to his P. capensis.
As pointed out by Kilburn (1978), the measurements given by Thiele and von Martens differ slightly from one another, but both indicate a higher apex than in the material here regarded as capensis; the larger series now available confirms the latter difference. This difference (h/l 0,82 or 0,91 as against 0,62-0,77 for our material) is no greater than that found in other Puncturella species, e.g. 0,60-0,85 for P. noachina (Farfante 1947), 0,75-1,00 [0,56 if the measure-ments given by Habe (1951) are accepted] for P. fastigiata. Perhaps differences in height in P. capensis are bathymetric, the type locality being at 70 m, while NM examples come from 400-510 m.

This species, never described in detail, has been overlooked by subsequent writers; the holotype was originally recorded by Von Martens as the Sino-Japanese P.fastigiata A. Adams 1853. Thiele concluded that the present species differed from fastigiata in its larger, more recurved apex, closer, more numerous ribs, a foramen that is situated relatively further from the apex, and an internal septum which lacks side buttresses. The dimensions given by him, 5,5 x 3,5 mm, height 4,5 mm, differ slightly from those cited by Von Martens.
The specimen recorded by Sowerby and Barnard as the boreal Puncturella noachina (Linne, 1771) is presumably referable here, although lower (height/length 0,69 instead of 0,82) and with radial ribs that are alternately more distinctly weaker and stronger than is shown in Thiele's figure of P. capensis. However, not only is the shell margin in this specimen damaged anteriorly, which has affected proportions somewhat, but some variation in height is to be expected (in P. fastigiata, for example, height/length proportions range from 0,75 to 1,00). Furthermore, Thiele's reference to 'erster und zweiter Ordnung' ribs indicates that his figure may not be quite accurate in showing ribs of uniform strength. The other alternative is the antarctic/subantarctic P. conica (d'Orbigny, 1841), which indeed agrees with P. noachina in shell characters, but differs anatomically (Dall 1914, Powell 1951). Obviously more South African material is required.

South-Africa. Transkei to Table Bay, 70-550 m.