Occurrence. In clay between two hard calcareous bands in the railway cutting between milestones 241-24 on the Graaff-Reinet railway, about three miles from Uitenhage (276).

Remarks. The specimen here described was referred to by Messrs. Rogers and Schwarz as "an Olcostephanus . . . allied to O. atherstoni," and it was rightly stated to be distinguished from this by "being more compressed laterally, in having no umbilical ribs, and in the less overlapping of the whorls." H. uitenhagensis is considerably more discoidal in form, and the rate of increase in the breadth of the whorl-section is much less than in H. atherstoni; the umbilicus is also very much wider and relatively shallower than in Sharpe's type. The ribbing of these two forms is of closely similar character, and both agree also in the very inconspicuous development of constrictions.

In the lateral compression of the shell and the fine character of the ribbing, the wide umbilicus and the degree of involution, H. uitenhagensis is more reminiscent of the typical H. astierianus (d'Orb.) † than of some of the more coarsely ribbed and inflated Holcostephani that occur in the Uitenhage Series; but the points of distinction are so marked that a detailed comparison with H. astierianus is unnecessary. The same remarks apply when we attempt a comparison between H. uitenhagensis and H. sayni Kilian,‡ the finely ribbed form which was originally thought to be identical with H. astierianus. In general habit there is similarity to the finely ribbed H. filosa (Baumberger)§ from the Hauterivian of the Swiss Jura, but the two forms are far from being identical.

GENUS ACANTHODISCUS V. Uhlig.

ACANTHODISCus sp.

Description of a Single Specimen.-The specimen consists of a fragment of the chambered part of a large whorl. It measures about 70 mm. from end to end along the periphery, and the crosssection at the anterior end is 35 mm. in height and about the same in breadth. The specimen is considerably weathered, and has had the shell partially removed, showing here and there a portion of the lobe-line and displaying the siphuncle near the posterior end. The whorl is slightly flattened on the sides and evenly and broadly

* Rogers and Schwarz (1), p. 10.

Kilian and Leenhardt (1), p. 976;

p. 38, pl. iv., figs. 2, 3.

+ d'Orbigny (1), pl. 28, fig. 1 (1840). Sarasin and Schöndelmayer (1), part 1,

§ Baumberger (1), 4er Theil, p. 31, pl. xxiii., figs. 2a, 2b.

arched in the peripheral area. To judge by the channelling of the inner surface, the involution was relatively slight. Since the height of the cross-section at the posterior end is slightly over 30 mm., the rate of increase of the whorl was not rapid.

The umbilicus must have been shallow. The umbilical rim is not abruptly defined, though the surface of the flank slopes down steeply to the spiral suture. Primary ribs starting from the suture are radial in direction and are not of equal strength. At the rounded umbilical margin, some of the ribs (main ribs) swell into a rounded node, and a bifurcation of the rib may here take place, giving rise to ribs of unequal strength. Some ribs, however (intermediate ribs), do not bear umbilical nodes, and bifurcation at the nodes on the main ribs does not always take place. Some of the ribs arising from the umbilical nodes have a forward inclination as they pass towards the periphery. On the peripheral side of the middle of the flank, but near the middle, a second set of nodes is developed on the main ribs, stronger than the inner series. At these nodes a division of a rib into two or three peripheral ribs takes place, and these have a well-marked forward inclination. The intermediate ribs have a corresponding forward swing as they pass on to the periphery. The peripheral ribs are of approximately equal strength and pass across the periphery, though their development is less pronounced in the central area. This area is flat, and in better preservation might be weakly sulcate. On either side of it the ribs show traces of a swelling.

The inner surface of the fragment (that is, the concave surface) is deeply channelled, in such manner as to show that the previous whorl had a distinctly sulcate periphery, and that the central concave area was bounded on either side by a line of nodular swellings in the peripheral ribs.

Occurrence.-Collected at Brentford, Knysna Estuary (151h).

Remarks. It is unfortunate that only a single fragment of this well-characterised form is available for study. The specimen is so imperfect that its intimate relationships cannot be precisely ascertained, and the important evidence of young and early adult growthstages is wanting. There can scarcely be a doubt, however, that we are dealing with a representative of that group within Hoplites (sensu lato) to which Uhlig has given the generic name Acanthodiscus.* Whether this group of trituberculate forms constitutes a generic unit may be open to question, in view of the differences in the characters of the youthful stages shown by some of its members, though Uhlig (5), p. 607.

*

Professor Uhlig has suggested that the occurrence of cœnogenetic phenomena may account for these. So far as the characters of sculpture of this fragment are concerned, a very similar type is seen in the adult stage of Acanthodiscus hystricoides (Uhlig),* from the upper Teschener Schichten of Silesia (correlated by Uhlig with the Valanginian). To judge from a comparison of such scanty material with the figures of A. hystricoides, the African specimen differs in the more nearly circular section of the whorl, in the occurrence of groups of three peripheral ribs arising from the outer tubercles of the flank, and perhaps also in the greater number of ribs intermediate between the main ribs. It also shows a more strong forward inclination of the ribs as they pass from the flank to the periphery.

It is impossible to say with certainty whether this form bears any close relationship to Tate's Ammonites subanceps, which also came from the Uitenhage beds. Tate's figured specimen † is the only one of its kind known from these beds, and it is a small individual, representing in all probability an immature stage of growth. This probability, and the great disparity in size between Tate's type and the fragment of an adult Acanthodiscus above described, makes a comparison difficult and unsatisfactory. The fragment here dealt with formed part of an individual exceeding 110 mm. in diameter, with an umbilical diameter, measured from the inner tubercles, of probably 45 mm. Tate's specimen measures less than 20 mm. in greatest diameter. Allowing for this, however, there is some similarity in the type of sculpture in the two forms. If they are closely related the points of distinction that exist may well be due to the fact that different stages of growth are brought into comparison. In the specimen from Knysna the sulcation of the peripheral area is very weakly marked, but it is seen from a comparison of the outer and inner sides of the whorl that this is a modification accompanying advancing growth, and that the previous whorl was much more sulcate. It is also probable that the reduction of this character is to some extent only apparent, owing to the removal of the shell-substance on the outer side by weathering. Other points in which the specimen differs from Tate's type are the presence of well-marked nodes at the umbilical margin, the situation of the second series of nodes a little nearer to the periphery, the stronger forward inclination of the peripheral ribs, and the rather broader periphery and more nearly circular outline of the whorl in cross

section. It should be noted with regard to the first of these points that although Tate made no mention of any trace of umbilical marginal tubercles in his specimen, such traces exist in several of the ribs, which are swollen at that part which corresponds with the row of umbilical nodes in the Knysna specimen.

While bearing in mind that these forms are so far comparable, it is at the same time not improbable that Tate's Ammonites subanceps should be placed with those species of Hoplites (sensu lato) which Uhligh as united to form the narrower generic group Solgeria.* Great similarity is shown to figures of forms ascribed by Sayn to Hoplites arnoldi (Pict. and Camp.) and by Toucas to Hoplites botella Kilian. These have been considered by Pavlow to be wrongly determined, and to represent one species to which he has given the name Hoplites heteroptychus.§ It was even suggested by Neumayr that Ammonites subanceps might represent the young of Crioceras spinosissimum, and this does not seem excluded as a possibility. While Tate's specimen, which was very imperfectly figured, is certainly to be brought into the closest comparison with Neocomian forms of Hoplites, and not with the Jurassic Reineckia anceps as thought by Tate, only the collection of additional material will show with certainty its true narrow relationships, though I am inclined to think that its nearest allies are to be sought in representatives of Uhlig's genus Solgeria, of which Hoplites heteroptychus Pavlow is an example, rather than in members of the genus Acanthodiscus.

GENUS BELEMNITES Lamarck.
BELEMNITES sp.

Two specimens of belemnites are unfortunately only of a fragmentary character. One was collected by Mr. Rogers at Brentford, Knysna Estuary (153h), and the other was obtained by Miss M. Wilman at Coega.

1. The specimen from Brentford is a fragment of a rostrum, and neither the alveolar nor apical end is preserved. It is compressed in the dorso-ventral direction in such manner as to give a broadly ovate outline in cross-section. The greatest transverse diameter is situated at some distance-at least a third of the length of the fragment-away from the alveolar end of the specimen. From this

*

Uhlig (5), p. 624.

Sayn (1), p. 682, pl. xii., fig. 6.

Toucas (1), p. 606, pl. xviii., fig. 10.

§ Pavlow and Lamplugh (1), p. 467 (109 of authors' copy).

point the outline tapers very gently, almost imperceptibly, in the alveolar direction, more markedly (though still gradually) in the apical direction. The length of the fragment is 45 mm. At the end towards the alveolus the transverse diameter is 11 mm., the dorsoventral diameter 9.5 mm. At the other end the transverse diameter is 9 mm., the dorso-ventral measurement 8 mm. Situated on the side are two parallel, shallow longitudinal grooves, placed closely to one another. Towards the anterior end of the specimen the surface is becoming split off in the form of thin laminæ.

This specimen is evidently the representative of some subfusiform species. The shape of the rostrum, though this is represented by an imperfect fragment, is sufficiently suggestive, and in addition there is the presence of the shallow parallel grooves on the side and the foliaceous character of the surface towards the alveolus. These two latter points, apart from the shape of the guard, at once recall certain Hastati of the European Neocomian, and there can be no doubt concerning the broad relationships of the specimen. The greatest thickness of the guard is, however, nearer to the alveolus than in Belemnites jaculum Phill. (=B. subfusiformis Rasp.), B. pistilliformis Blainv. (=B. pistillirostris Pav.), or B. obtusirostris.* The outline in section is more oval and compressed than in B. jaculum, and although in that species there is some variation with regard to the form of the section, Mr. G. W. Lamplugh informs me that he has never observed quite such a degree of compression as is shown by the Brentford specimen.

2. The specimen from Coega is still less favourably preserved for a satisfactory comparison with known forms, but it also belongs to some subfusiform species, and one which has a rostrum of relatively slender figure. It is a fragment of a guard from which the alveolar and apical ends are missing, and is compressed in manner similar to that shown by the other specimen. It measures 46 mm. in length. The transverse diameter at the thicker end is 8 mm., the dorsoventral diameter 7 mm. The specimen tapers gradually from here towards the thin end, where the greatest diameter is 4 mm. This end is most probably approaching the alveolus, though it shows no sign of a ventral groove. Here also the surface shows no foliaceous character, but the specimen is so unfavourably preserved and so much weathered that these negative characters are not of much significance. It is possible, however, that this thin end of the specimen lies towards the apex, but if this be the case we are dealing with a hastate form remarkable for the slow tapering towards the * Pavlow and Lamplugh (1), pp. 77-82.