Page images

specimens was far from satisfactory for a complete anatomical study, although G. S. I. Type No. 16502 was somewhat better than G. S. I. No. K40/485. Slides prepared from G. S. I. Type No. 16502 showed sufficient anatomical structure to ensure its identification, while those from G. S. I. No. K40/485 were far too brittle to be ground to the required thinness. In this connection it is interesting to note that microscopic examination of G. S. I. Type No. 16502 revealed fungal hyphæ in its woody tissues (Pl. 16, fig. 2), but not in G. S. I. No. K40 485.

Geologically the locality is known to be Tertiary but no definite information is at present available as to the exact age of the fossil bed. In the recent geological map it is coloured as Siwalik, which is equivalent to the Tipam stage of the Assam series. Occurrences of fossils in this particular area have already been reported. Pinfold (28) found some fossils at about 41 miles distance from Dalu, on the road from Dalu to Tura, and also at about a mile from Bagmara, on the foot-path from Bagmara to Dalu. Based on palæontological information alone he was not inclined at that time “to correlate succession of Garo Hills with that of Assam.” Vredenburg (38) carefully studied Pinfold's specimens and made some interesting observations. He compared these fossils with those from northwestern India, Burma and Java, and provisionally placed them along with the Upper Gaj of Sind and the Pyalo beds of Burma. He was of opinion that further investigation would be necessary before coming to a final conclusion. Later on Evans (12) and his co-workers reported additional fossil localities

Dalu and Bagmara. These workers were of opinion that the fossil-bearing beds could be regarded approximately as Upper Gaj or Burdigalian. Further information on this question will perhaps be available when Dr. Fox, who has been engaged for some time on the geological survey of this part of Assam, completes his work.


(A) SPECIMEN G. S. I. Type No. 16502.

(1) Anatomical description. Growth rings. The cross sections did not show any growth marks either to the naked eye or with a hand lens (Pl. 15, fig. 1), nor did careful search under a high power microscope reveal any structural difference indicating boundary of the growth ring (Pl. 15, figs. 2 and 3).

However, at one extreme con er of the fossil, some suggestion of gro:vth marks was noticed, which under a high power microscope turned out to be nothing but a group of gum ducts aligned tangentially.

Vessels.-Distribution of the vessels is more or less uniform throughout the wood, without any tendency to grouping at any particular place (Pl. 15, figs. 1, 2, 3). Individual vessels or vessel groups are distinctly visible to the eye, appearing as dots against the ground tissue of the wood. This timber also shows some comparatively small vessels, which are distributed rather sporadically (Pl. 15, fig. 2). At first it was thought that these might be solitary gum ducts, but careful examination revealed their tracheal nature. For the most part the vessels are medium-sized and they are either solitary, or in radial or obliquely radial groups of 2 to 4. In transverse sections the tyloses are not very distinctly visible but in longitudinal sections their presence cannot be overlooked (Pl. 15, fig. 6). Vessel cavities are often plugged with a darkish deposit of a gummy nature. Vessel lines are often prominent on the longitudinal surfaces, due to the coloured deposit that they contain. Perforation plates are simple, horizontal to slightly inclined. Intervessel pit pairs are fairly large, alternate ; outlines of the border are oval and widest horizontally; the aperture is lenticular and horizontal or slightly oblique (Text-figs. 1, 2). Vessel-tracheid pits are similar to those of the inter-vessel. Vessel-ray pits are distinctly of two types and show some similarity to inter-vesse! pits. They are one to several per ray cell ; when one, usually large (Text-fig. 3); when several, they are comparatively small. The outline of the border is round to oval. The aperture is oval to lenticular, horizontal or slightly oblique. Vessel-parenchyma pits are of the same type as those of vessel-ray but in this case the large type is less common than the small type.

Tracheids.-Although they are not distinguishable under low power, careful examination of some vessels under a high power microscope shows their vasicentric distribution, often intercepted by parenchyma cells. They are most!y flattened to conform to the wall of the vessel and are usually in a single row. As regards pits, they are of the same type as those found on the wall of the vessel, their size and shape depending on the cells that are contiguous to them.

Wood fibres.-Outline of the fibres can only be distinguished under high magnification. As a rule, they are irregularly arranged, although they may show somewhat radial alignment in small patches (Pl. 15, figs. 1, 2). The thickness of the fibre wall varies from thick

[merged small][ocr errors][merged small][ocr errors][ocr errors][merged small][merged small]

Figs. 1-3. Dipterocarpoxylon garoense sp. nov. (Semi-diagrammatic camera-lucida

drawings). Fig. 1. Showing pits between wood rays and longitudinal parenchyma cells ; note grouping of pits. Fig. 2. Portion of longitudinal parenchyma cells showing interparenchyma pits. Fig. 3. Portion of a wood fibre with pits.

to very thick. Longitudinal sections do not give the correct impression of fibre wall thickness, which can best be gauged either in transverse sections or in macerated material. No septate fibre has been

[ocr errors]

observed, although occasionally some fibres show resinous contents. Inter-fibre pits are simple or bordered; the outline is round to oval; the aperture is slit-like, vertical or slightly oblique (Text-fig. 3).

Parenchyma.-- The amount of parenchyma present is scanty to fairly abundant. Paratracheal parenchyma cells are in a sheath of 1 to 3 (mostly 2) cells often intercepted by tracheids (Pl. 15, figs. 1, 2 and 3). Metatracheal parenchyma cells are in narrow to wide bands. They are irregularly spaced and mostly discontinuous (Pl. 15, figs. 1, 2). When surrounding the gum ducts they are usually in a row of 3 to 6 cells. Diffuse parenchyma cells are also present but are rather scanty. Inter-parenchyma pits are round to oval. They are of two types; the small type is in clusters (Text-fig. 2) while the large type is usually single. Parenchyma-ray pits are similar to those of the inter-parenchyma (Text-fig. 1).

Rays.--On the cross surface of the fossil the rays are visible to the eye and under a hand lens (* 10) their outlines are fairly distinct. They are more or less uniformly distributed, moderately numerous and moderately broad (Pl. 15, figs. 1, 2, 3). On the radial surface of the fossil they do not show up conspicuously. Of the slides cut in three different planes the tangentials have given the best view of the diagnostic characters of the rays. They are mostly heterogeneous and can be classified more or less into three groups (Pl. 15, fig. 5). (1) The uniseriate type is composed of upright cells only. As a rule this type is not very high. (2) The biseriate type is also fairly common, and it shows only a few procumbent cells at the middle portion of the ray, the rest being composed of upright cells. (3) The majority of the rays are 5 to 8 (mostly 5 to 6) seriate. A row of upright cells forms the outer boundary of the rays and the inner portion consists of procumbent cells packed in an irregular fashion. The pits between ray cells are numerous. Their outlines are not so distinct in tangential sections but from some partially macerated material it has been possible to notice that they are small, oval or nearly so. As a rule, the rays are very low to low, but when vertically fused they run to a considerable height.

Gum ducts.-A casual observation of the cross sections might give an impression of a large number of vertical gum ducts, but careful microscopic examination proves it to be incorrect. Some thin-walled vessels with resinous contents do show up somewhat like the gum ducts, but the tracheal origin of their surrounding cells

be easily detected. The gum ducts show a few layers of


surrounding parenchyma cells which are often arranged in radial rows (Pl. 15, fig. 4), while the vessels containing gummy deposit have at least a few tracheids round about them (Pl. 15, fig. 3). By this method it has been possible to detect some vertical gum ducts in this fossil and these are classified under two groups : (1) diffuse or scattered in groups of 1 to 3 and (2) in a long row of tangential bands (Pl. 15, figs. 1, 2). But no horizontal gum ducts have been observed in the fossil.

(2) Comparison with living dicotyledons. In the description of the fossil given above, the presence of vertical gum ducts has been recorded. This is of great importance from an anatomical point of view, for according to Solereder (37) and Record (30) only four orders are known to possess the gum ducts of the normal type found in the fossil. They are Cornaceae, Simarubaceae, Leguminosae and Dipterocarpaceae. In

Cornaceae the genus Masticia (30) is said to have vertical gum ducts, but the timber specimens of this genus in the Dehra Dun collection did not reveal any gum duct. Moreover, the woods of Mastizia show very little similarity with this fossil either in general structure or in minute

In Simarubaceae, only one genus Simaruba (30), which is mostly confined to tropical America, has this special feature. Besides the gum ducts there is very little similarity between Simaruba and this fossil. In Simaruba the distribution and the shapes of both ray and parenchyma cells are entirely different from those of the fossil. There appears to be, therefore, no justification for imputing any affinity of the fossil G. S. I. Type No. 16502 with the genus Simaruba. Then comes the sub-order Caesalpinieae of the Leguminosae in which the genera Copaifera, Daniella, Eperua, Kingiodendron, Prioria and Sindora have normal vertical gum ducts. Now, taking them one by one, in a Copaifera of tropical America, Record and Mell (29) reported vertical intercellular canals, but Chalk and his co-workers (3) did not find any canals in two African species that they had studied. Apart from the gum ducts, the general features of the Copaiferas are entirely different from what has been found in the fossil ; for instance, the Copaiferas show distinct growth rings and have vasicentric and confluent parenchyma, and their rays are exclusively homogeneous. In view of these differences there does not appear to be any affinity between this fossil and the Copaiferas. Next comes Daniellu, a

comes Daniellu, a genus growing in tropical

« PreviousContinue »