Chelonia-large number of scutes and plates of the carapace and skull-bones. Crocodilia-bones and limb-bones. Snake-vertebrae. Fish-vertebrae. Ampullariid-opercular. EXPLANATION OF PLATE 15. FIGS. 1, la.-Outer, and inner views of the operculum of the type-specimen of P. prisca, sp. nov. FIGS. 2, 2a, 2b.-Outer, inner and lateral views of another operculum of the same species. FIGS. 3, 3a, 3b.-Outer, inner and lateral views of the operculum of a specimen of P. globosa (Swainson) from Calcutta. All the figures are direct photographs of the opercula and with the exception of 2b are of the natural size of the specimens, Fig. 2b is enlarged twice natural size. ON A CALCAREOUS ALGA BELONGING TO THE Triplo porellea (Dasycladacea) FROM THE TERTIARY OF INDIA. BY JOHN WALTON, M.A.; Lecturer in Botany, Manchester University. (With Plate 16.) THE HE name Triploporelleæ was given by Oltmanns 1 to a group of fossil algæ, represented by the single genus Triploporella, which he considered to be a subdivision of the family Dasycladaceæ. The Siphonocladiales (which include the Dasycladaceae) differ principally from the other divisions of the Chlorophycea (Green Alga) in being infrequently septate; and when septation occurs it is quite independent of nuclear division, the segments containing several nuclei. Generally in this group the more frequently septation occurs the greater is the number of branches and the complexity of the thallus. In the Dasycladaceae there is a strict symmetry in the organisation of the thallus. There is a large axial segment, which bears closely placed branches arranged in whorls; these again may bear smaller branches. The thallus in many of the genera is encrusted with calcium carbonate. Among the subdivisions of the Dasycladacea the Dasycladeæ and Bornetelleæ are among living types the closest to the Triploporellea which Oltmanns regards as intermediate to these two groups. The material on which the following description is based was sent to Prof. Seward by the Director, Geological Survey of India, together with descriptive notes and illustrations by Mr. B. B. Gupta, Sub-Assistant in that Department; it was collected by the late Mr. Vredenburg from the Ranikot Beds in Sind, which are correlated with the lower Eocene of Europe. The material consists of several separate segments of a narrowly ovate shape (Pl. 16, figs. 1, 2, 3, 4). Those that are complete vary from 4-5 mm. in length and have a maximum diameter at the broadest end of 2.5 mm. 2 In shape the segments resemble those of Ovulites margaritula Lamarck sp., a branching calcareous alga from the Lower and Middle Eocene of France, Hungary, Belgium, and Italy which Munier-Chalmas considered to be generically identical with Penicillus (Coralliodendron, Kutz.). Whatever the relationships of 1 Oltmanns, 1904, p. 277. 2 Lamarck, 1816. (See Seward, 1898.) Ovulites may be, our fossil is certainly not allied to Penicillus, one of the Codiacea, but must be ascribed to the family Dasycladacea and to its subdivision the Triploporelleæ, (Oltmanns).1 Nevertheless the occurrence at both ends of the segment of depressions suggests a jointed habit such as is found in the recent genera Penicillus and Halimeda 2 of the Codiace and in Cymopolia one of the genera of the Bornetelleæ (Dasycladaceœ). Mr. Gupta pointed out that there are two depressions at the larger end of one of the segments (Pl. 16, fig. 5) and this may be taken as additional evidence of a dichotomous type of branching such as has been described for Ovulites. This segmented branching habit is not known to occur in the other two members of the Triploporelleæ. In both Steinmann, 4, 5 describes club-shaped plants with no indication of any continued proliferation of the axis. However in Triploporella Fraasi, Steinmann,6 (Upper Cenomanian) a very long segment is figured with slight constrictions. It may be that these constrictions are foreshadowings of the definite jointed structure found in Cymopolia and in our Tertiary form, for which the specific name ranikotensis is proposed. In longitudinal section each segment is seen to consist of a single siphon, the diameter of which is very nearly one-third the diameter of the segment as represented in the fossil. From this central siphon lateral branches are given off in whorls (Pl. 16, figs. 8, 9). The section shown in fig. 9, which is transverse to the axis of the segment, passes in a median plane through each of the 15 branches. It will be seen (figs. 6, 10, 12) that the members of successive whorls are usually alternate. The primary branches, the lower portions of which are which are shown in figs. 8, 9 are marked out by the different nature of the calcareous matter of the fossil between the branches of the first order. This calcareous matter may confidently be assumed to represent the secretion of calcium carbonate on the exterior surface of the cellwall of the axial siphon and its primary branches; the calcium carbonate deposit is white and opaque whereas the carbonate which fills the lumen of the siphon and its first-order branches is fairly transparent. The preservation is not good and it is doubtful whether one can distinguish the pores which must have existed between the axial siphon and its laterals. The amount of weathering to which the surface of the segment had been subjected determined the features seen on the surface. In fig. 12 the pattern on the surface of the two uppermost whorls (x) is due to the fact that the weathering has extended for some distance inwards from the outer ends of the branches of the first order which are therefore seen in transverse section (cf. fig. 6). In the next three whorls (fig. 12, y), three or more darker patches are seen superposed on the end of each first-order branch (cf. figs. 2, 3). These represent the lumen of the second-order branches, the lighter network representing the calcareous deposit on the extremity of the first-order branch round the bases of the second-order branches. Below this zone the hexagonal outlines of the first-order branches are lost to view, and the finer meshed network (2) which is seen represents the calcareous deposit between the second-order branches the lumena of which are presented by the slight depressions (cf. fig. 1). The organic calcium deposit is apparently more resistant to weathering than the filling material. The number of secondorder branches from each first-order branch seems to vary from 2 to 7 in the specimens. In fig. 11 is seen a first-order branch rather shorter than usual, with the basal calcified portions of two second-order branches surmounting it. It has not been possible to distinguish with any certainty the presence of sporangia in this Indian specimen; occasionally, however, there are suggestions that the first-order branches functioned as sporangia, as in Triploporella Fraasi and Remesi. In fig. 9 there is a gap at (a) where three branches have not been preserved. There are also indications of small spherical bodies at the base of one. It is possible that the end portion of the wall of the sporangium was not calcified and hence the sporangium was not preserved. Other examples of the absence of certain members of a whorl of first-order branches of the siphon have been observed (fig. 4, h). In general habit Triploporella ranikotensis must have resembled Coralliodendron (Ovulites) margaritula Lamarck sp. and it is to be remarked that they are both of Eocene Relationship of the Alga. age. However, on examination of the relation of the lateral branches to the axial siphon in the segments of Triploporella ranikotensis it is seen that it is different from any plant described under the name of Coralliodendron or с Ovulites and that the forms with which it most closely agrees are those found in the Dasycladaceae. On the whole the shapes of the first-order branches are very like those of Triploporella, and the arrangement of the second-order branches is also similar to what is found in that group. Among the large number of forms elegantly illustrated and described by Pia1 there are a few genera in other groups than the Triploporelleæ (Olt.) in which the first-order branches are of this shape; e.g., Goniolina geometrica Roem. sp., which Pia groups with Triploporella in the Triploporellineæ, and Petrascula. It must be noticed tooth at in Cymopolia (Bornetellea) the thallus had a jointed structure, so that in this respect we have another point of contact with that group. As regards the structure of the branches of the first-order and the number of the branches of the second-order our species certainly resembles Triploporella (T. Fraasi and T. Remesi) very closely. It resembles T. Remesi Stein. most closely in the number and arrangement of its second-order branches and T. Fraasi Stein. as regards the shape of its first-order branches. The dimensional relationships between the three species, and one variety, of Triploporella which have been described are tabulated below: |