resemble in many respects Ech. Darwini Clarke * from Brazil. Kozlowski† records the genus also from the Devonian of Bolivia.

Aspidosoma? sp.

A fragmentary asterid (2610 Alb. Mus.) in the Albany Museum may be compared with Aspidosoma pontis Clarke,‡ from Brazil, but the specimen is insufficient to decide its reference. Kozlowski § refers a Bolivian fossil to this genus with a query.

CRINOIDEA.

Ophiocrinus Stangeri, Salter.

1856. Ophiocrinus Stangeri, Salter, Trans. Geol. Soc., ser. 2, vol. vii, p. 223, pl. xxv, fig. 20.

The type of this species is in the British Museum. The exact locality is not known.

There is a good specimen of it (2608 Alb. Mus.) in the Albany Museum from Gydo Pass and one in the Pretoria Museum (169 Pret. Mus. type 2) from Montagu. It may be here mentioned that Salter's genus Ophiocrinus (1852) is not identical with Angelin's genus (1878) bearing the same name, as Wachsmuth and Springer || pointed out. Bather has given a precise synopsis of the characters of the genus which he places in the Rhodocrinidae.

Crinoid stems.

1856. Crinoidal remains, Salter, Trans. Geol. Soc., ser. 2, vol. vii, p. 223, pl. xxv, figs. 21-28.

Salter figured (but did not describe in detail) various stems and stemjoints of crinoids from the Bokkeveld Beds. Several distinct types can be recognised, and they probably belong to more than one genus. Such imperfect material is represented in all the collections of Bokkeveld fossils from various localities, but detailed descriptions of it are not of much value.

* Clarke, Foss. devon. Parana, 1913, p. 315, pl. xxviii, figs. 9-12. Kozlowski, Ann. Paléont., vol. xii, 1923, p. 100, pl. x, fig. 22. Clarke, Foss. devon. Parana, 1913, p. 312, pl. xxvii, figs. 1-8.

§ Kozlowski, Ann. Paléont., vol. xii, 1923, p. 99, pl. x, fig. 21.

|| Wachsmuth and Springer, "Revision of Palaeocrinoidea," Proc. Acad. Nat. Sc. Philadelphia, 1879, p. 320; ibid., 1881, p. 407.

¶ Bather in Lankester's Treatise on Zoology, pt. 3, 1900, p. 201.

VOL. XXII, PART 1.

3

BRYOZOA.

Fenestella sp.

Only one representative (74f S.A. Mus.) of this genus has been recognised in all the collections of Bokkeveld fossils, and this consists of the base of a zoavium and a few branches. The branches are small, parallel, and much thicker than the dissepiments which are set at equal distances apart and form subquadrate or rounded fenestrules longer than wide. Traces of cells can be seen on the branches, and in a length of 6 mm. there are 6-7 fenestrules.

Monotrypa? sp.

On the same piece of rock (11358) from Leo Hoek, in the British Museum, as that containing the impression of a coarsely rugate Grammysia (Gr. corrugata ?), there is a portion of a corallum of a small monticuliporoid consisting of small polygonal cells, a few of which seem to be larger and to form indefinite maculae. The only such fossil from the Lower Devonian of the southern hemisphere is Monotrypa raritabulata Knod,* described from Bolivia, but ours can hardly be considered identical with it.

BRACHIOPODA.

Lingula Keideli, Clarke.

1903. Lingula aff. densa Reed (pars), Ann. S. Afr. Mus., vol. iv, pt. 3, No. 7, p. 167, pl. xx, fig. 1 (120 S.A. Mus.) (non fig. 2). 1913. Lingula Keideli Clarke, "Foss. devon. Parana," p. 309, pl. xxvi, figs. 8-10.

One (120 S.A. Mus.) of the specimens which the author figured in 1903 as a form allied to Lingula densa Hall, may be attributed without any hesitation to Lingula Keideli Clarke, described from Brazil. In my figure the beak is made rather too obtuse, and should be more acute, as shown in one of Clarke's specimens, fig. 8. It agrees very closely with Kozlowski's † Bolivian shell which he referred to L. densa Hall, and probably is identical with it.

Lingula lepta, Clarke.

1903. Lingula aff. densa Reed (pars), Ann. S. Afr. Mus. vol. viii, pt. 3, No. 7, p. 167, pl. xx, fig. 2 (123 S.A. Mus.) (non fig. 1).

66

* Knod, Neues Jahrb. f. Miner., Beil. Bd. xxv, 1908, p. 560, t. xxxi, figs. 1, 2. † Kozlowski, Foss. devon. Brésil," Ann. Paléont., vol. viii, 1913, p. 3, pl. i, figs. 3-10.

1913. Lingula lepta, Clarke," Foss. devon. Parana," p. 308, pl. xxvi, figs. 3-7.

1913. Lingula cf. Manni, Kozlowski, "Foss. devon. Brésil," Ann.

Paléont., vol. viii, p. 2, pl. i, figs. 1, 2.

Clarke and Kozlowski have identified the South African shells, which the author in 1903 considered allied to L. densa Hall, with Brazilian species; the former gives one a new specific name, but the latter only compares it with the North American species L. manni Hall. Both, however, were misled in placing both my figured specimens in the same species, for the second figured one (123 S.A. Mus.) (E. 783 Stell. Mus.) has a much more truncate anterior end than the first one (No. 120 S.A. Mus.) (op. cit., fig. 1), which is referable to Clarke's other species, L. keideli. Through the kindness of Dr. du Toit I have been able to compare the specimens with examples which he collected in Brazil. There are specimens in the South African Museum (169a, 5415) from Uitvlugt, Montagu, and one in the Stellenbosch Museum (E. 783), which are referable to this species. Another (124) from Tunnel Siding is in the Sedgwick Museum.

Lingula scalprum, Clarke.

(Pl. IV, fig. 3.)

1913. Lingula scalprum Clarke, "Foss. devon. Parana," Mon. Serv. Geol. Miner. Brasil, vol. i, p. 310, pl. xxvi, figs. 1, 2. 1923. Lingula scalprum Kozlowski, "Foss. devon. Bolivie," Ann.

Paléont., vol. xii, p. 81, pl. ix, fig. 25.

There is one good specimen (H. 67 Stell. Mus.) from Hottentot's Kloof, Ceres, which is undoubtedly identical with Clarke's small species L. scalprum from Brazil. Kozlowski regards one of the specimens from Bolivia which Knod * included in his L. subpunctata as belonging to Clarke's L. scalprum.

Lingula (Glossina ?) sp.

1903. Lingula sp. Reed, Ann. S. Afr. Mus., vol. iv, pts. 3, 7, p. 167, pl. xx, fig. 3 (121 S.A. Mus.).

The small undetermined species of Lingula figured by the author from Hottentot's Kloof, Ceres, is of doubtful reference, and may belong to the subgenus or genus Glossina Phillips, for it suggests in

* Knod, Neues Jahrb. f. Miner., Beil. Bd. xxv, 1908, p. 536, t. xxvi, fig. 7 (non fig. 8).

shape G. acer Clarke * from the Grande Grève Limestone of Gaspé, and its allies L. perlata Hall, and L. spatiosa Hall, of the Helderbergian.

Orbiculoidea (Roemerella) cf. collis, Clarke.

(Pl. IV, figs. 9, 9a.)

There is a large high capuliform brachial valve (6716 S.A. Mus.) of a species of Orbiculoidea from the farm Driefontein, Clanwilliam, in the South African Museum, which seems to be comparable to the Brazilian species O. collis Clarke † and to O. montis Clarke,‡ from the Grande Grève Limestone of Gaspé. Our specimen is somewhat distorted, having been compressed in an antero-posterior direction, so that it measures about 45 mm. in height, about 55 mm. in its anteroposterior diameter, and about 73 mm. in its transverse diameter. The apex is blunt and eccentric, being somewhat nearer the posterior side, towards which it curves slightly back and descends steeply in a gentle concave curve; the anterior slope is gently arched and convex and also descends steeply, but less so than the posterior face. The shell is ornamented with regular equidistant sharp raised concentric lines which have fine concentric striae in the concave interspaces between them. The surface of the shell is also marked with 5-6 broad flat concentric bands of equal width at distances of every 9-12 concentric lines, indicating periods of growth. Closely placed low rounded equidistant radial lines are visible also on the anterior and posterior slopes of the valve, but they belong to an inner layer of the shell, as may be plainly seen below the apex on the posterior face.

Orbiculoidea (Roemerella) Baini (Sharpe) (emend.).

1846. Orbicula sp. Morris and Sharpe, Proc. Geol. Soc., pl. x, fig. 5, p. 277.

1856. Orbicula Baini Sharpe (pars), Trans. Geol. Soc., vol. v, pt. 7, p. 210, pl. xxvi, figs. 21-23 (non fig. 20, nec diagnosis) (11333, 11334, 11335 Brit. Mus.).

1903. Orbiculoidea Baini Reed, Ann. S. Afr. Mus., vol. iv, pt. 4, p. 168, pl. xx, figs. 4, 5, (158 S.A. Mus.).

1913. Orbiculoidea Baini Kozlowski, "Foss. devon. Parana," Ann. Paléont., vol. viii, p. 4, pl. i, figs. 11-13.

* Clarke, Mem. No. 9, New York State Mus., pt. 1, 1908, p. 214, pl. xlvii, figs. 7, 8.

Clarke, Foss. devon. Parana, 1913, p. 306, pl. xxv, figs. 23-26.

Clarke, Mem. 9, New York State Mus., pt. 1, 1908, p. 213, pl. xlvii, figs. 21-24.

1913. Orbiculoidea Baini Clarke, Foss. devon. Parana," p. 301, pl. xxv, figs. 5-15.

1916. Orbiculoidea (Roemerella) Baini, Williams and Breger, Prof. Paper 89 U.S. Geol. Surv., p. 24.

The shell figured by Sharpe in 1856 (fig. 20, op. cit.) as a small specimen," and showing the very fine radial striation which is not present in any of the others, is quite distinct from the shell from the Falkland Isles figured by him in 1846 as Orbicula sp., but not named or described at that time. The latter agrees with the shells usually termed O. Baini, and as such figured by Sharpe in 1856 (figs. 21-23, op. cit.). These have been regarded as typical of the species, and it seems, therefore, inadvisable to substitute a new name for this common form after it has been for so long in general use. But Sharpe's definition is poor and misleading, as he refers to radial striae which are different in nature to those on his "small specimen " (op. cit., fig. 20), as they are on the inner surface or layer of the shell and therefore only visible when the shell is more or less removed or flaked off. Clarke, in 1913 (op. cit.), remarked that the Parana shells are wholly devoid of the fine radial surface striation which Von Ammon's * Discina Baini, from Matto Grosso, possessed, this shell probably belonging to the same genus, if not species, as Palaeoscurria Sharpei (q.v.).

In the common Bokkeveld shell illustrated by Sharpe's other figures, 21 to 23 (op. cit.), the shell is circular, and the upper (brachial) valve forms a low flattened cone, with the apex subcentral, acute, and elevated to a height equal to almost one-fourth or one-third the diameter. The pedicle valve is slightly concave, and the pedicle foramen is large, oval, and sunken. The surface of the brachial valve, which is usually the one preserved, is covered with sharp equidistant narrow ridges, usually regular and of equal strength, while radial striae are completely absent on the exterior, as Clarke remarked. (op. cit.), only decorticated specimens showing fine, widely-spaced, radial markings, as may be seen in the British Museum specimen (No. 11333) figured by Sharpe (op. cit., fig. 21). Williams and Breger † compare this species with their O. (Roemerella) discus from the Chapman Sandstone of Maine, and probably Discina mediorhenana Fuchs,‡ from the Hunsruck States, and Orbiculoidea siegenensis (Kayser),§

* Von Ammon, Zeitschr. Gesell. Erdkunde, vol. xxviii, 1893, p. 359, fig. 4.

† Williams and Breger, Prof. Paper 89 U.S.A. Geol. Surv., 1916, p. 21, pl. vi, figs. 29-33 (? 28).

Fuchs, Abh. k. preuss. geol. Landesanst, 1915, Heft 75, p. 6, t. i, figs. 6-8. § Kayser, Jahrb. geol. Landesanst, 1890, p. 95, pl. xi, figs. 1, 2.