Flakes of molybdenite in quartz are apparently earlier than the latter and the little evidence available would suggest that the molybdenite is later than wolfram. Its position with respect to pyrite is uncertain. Pyrite is clearly later than wolfram and cassiterite, but its position with respect to quartz is not quite clear. In most cases it shows euhedral outlines against quartz, and, when the pyrite and quartz occupy sections of the same veinlet in wolfram, the pyrite G usually shows a crystal outline against the quartz section of the veinlet. Some pyrite is, however, replaced by quartz. In general, then, pyrite was earlier than quartz. Examples of quartz veining or replacing cassiterite are rare (Pl. 13, fig. 4), but quartz quite commonly replaces and veins wolfram. It is noticeable in several specimens that prior to deposition. of quartz the wolfram was crushed and fractured and sometimes finely brecciated (Pl. 13, fig. 2). In the majority of specimens the position in the sequence of fluorite is not clear, but in one or two specimens it definitely replaces quartz (Pl. 14, fig. 1). The commencement of fluorite deposition remains unknown, but it certainly continued after quartz. This is in contrast to the Mawchi ores, in which fluorite was determined to be amongst the earliest minerals deposited, but its lower limits there were uncertain. Veins of galena and bismuthinite have been seen in the Hermyingyi fluorite. Sphalerite replaces pyrite and quartz. Chalcopyrite separated from sphalerite on further cooling and in addition small patches of chalcopyrite replaced and veined sphalerite (Pl. 14, fig. 2). Galena veins penetrated sphalerite and in addition galena replaces chalcopyrite. Bismuthinite is closely associated with the galena, and bismuth obviously separated from the bismuthinite with lowering of the temperature. Veins of galena and bismuthinite in wolfram are common (Pl. 14, fig. 3). SUPERGENE CHANGES. Alteration of wolfram to tungstite and of chalcopyrite to covellite is found throughout these ores, and at depths of 314 feet from the surface. The leached appearance of some of the ore indicates the activity of surface waters at all depths. THE ORE LIQUID. The emplacement of these ores in the veins would appear to have been a simple injection of the ore liquid along the vein fissures. The elongated habit of the wolfram, many of the slender crystals projecting from the walls into the coarse quartz towards the centre, at once indicates that there could not have been any important reopening of the fissures prior to deposition of later minerals; whatever crushing of wolfram that did take place was local and probably due to small adjustments during crystallisation. The whole aspect of these coarsely crystalline veins is that of a siliceous ore liquid suddenly injected, probably under pressure such as to force the walls apart, and slowly crystallising from the walls inwards. There is no evidence leading to the assumption that the Sn was injected other than as oxide or the tungsten other than as the wolfram molecule. Fluorite is present in only very small amounts. although it is more abundant than in the Mawchi ores, and, apart from prolonging crystallisation to a rather lower temperature, fluorine was a relatively unimportant agent in the ore liquid. The total absence of B2O, is remarkable for ores of this nature, and there is not a characteristically high temperature mineral in the sequence. The ores were deposited at very low temperatures from a liquid which was apparently not acidic, and it is not unlikely that the final state of the ore liquid prior to crystallisation was that of a colloid. 3 The comparison with the Mawchi ores may be completed, then, with the observation that the Mawchi ores crystallised within a rather wider range of temperature, the earlier minerals being characteristically high temperature minerals, and the later separating at a much lower temperature. The Hermyingyi ores commenced crystallisation at quite a low temperature, probably in consequence of the rather higher amount of fluorite present. At Mawchi, fluorine escaped from the ore liquid at an early stage, at Hermyingyi it was fixed in the fluorite molecule, calcium carbonate being absent. The absence of calcium carbonate is also the explanation of the absence of scheelite from these ores, the wolfram, apparently unstable in the presence of calcium carbonate, remaining unreplaced. The quartz-mica-greisen which is so is so common along the walls of these veins was probably formed by the residual waters, unable to escape fully along the lode channels, soaking into the walls and replacing the felspars by muscovite. It is illogical to presume that this greisen provides evidence of the very dilute nature of the ore liquid, for the greisen bands are relatively narrow and the amount of water which they represent is relatively small. In my opinion there is no evidence which would suggest that the ore liquids were of such a dilute nature as is usually pictured; deposition from a by no means highly dispersed colloid is a possibility that is in conformity with the little evidence which this study has afforded. EXPLANATION OF PLATES. PLATE 13, FIG. 1.-Wolfram veining muscovite (M) which appears also to be interstitial to coarser wolfram (W). Quartz (Q). P. S. 268. X 54. FIG. 2.-Wolfram (light grey) brecciated and veined by quartz (dark grey). P. S. 268. X 54. FIG. 3.-Cassiterite (dark grey) replaced by wolfram (grey). Quartz FIG. 4.-Cassiterite (white) veined and replaced by quartz (grey). PLATE 14, FIG. 1.-Fluorite (F) replacing quartz (Q). Wolfram (W) and tungs tite (T). Cracks in fluorite, infilled with bakelite, appear like quartz. P. S. 267. X 54. FIG. 2.-Pyrite (P) replaced by sphalerite (S), in turn veined by chalcopyrite (C) and galena (G). P. S. 278. X 54. FIG. 3.-Veins of galena and bismuthinite (white) in wolfram. P. S. 267. X 54. FIG. 4.-Ex-solution droplets of bismuth in bismuthinite. P. S. 277, Oil immersion. X 540. Two FOSSIL DICOTYLEDONOUS WOODS FROM THE GARO HILLS, ASSAM. BY K. AHMAD CHOWDHURY, Forest Research Institute, Dehra Dun. (With Plates 15 and 16.) The material on which the present study is based was collected by Dr. C. S. Fox of the Geological Survey of India in January, 1933. Two years later he kindly handed it over to the writer for investigation. The material consisted of two specimens, both collected in the Garo Hills district, Assam. The specimen G. S. I. Type No. 16502 was from Damalgiri, about eleven miles west of Tura, headquarters of the district. In size it was about 8 cm. long, 4 cm. broad and 2 cm. thick. G. S. I. No. K40/485 was from a place about a mile north-west of Garobadha, nineteen miles west of Tura. It was a smaller specimen than G. S. I. Type No. 16502, about 4 cm. long, 3 cm. broad and 2 cm. thick. The preservation of both the |