In the bay of Emmahaven near Padang (W. Sumatra) is a small coral reef. On the North side of this reef a small, sandy island occurs called Pasir Ketjil. It has lately been surrounded by brick walls, besides which a stone pier has been built in a S.W. direction (cf. fig. 1, a map of 1899 and fig. 2 and 3, the present condition). Since the building of that pier, coral sand has accumulated against it on the S-side. This reef lacks shingle ramparts. From this situation of the sandy island with regard to the reef, and from the placing of the younger coral sand accumulations (A in fig. 3), it may be concluded that the maximal wind-effect is strongest from the direction of the open sea, about S.W. Data about the wind frequency kindly supplied by the director of the Kon. Magn. Meteor. Observatorium at Batavia are given on p. 13. From these data have been derived the average wind frequencies per year (p. 15) recorded for 6, 9, 15 and 22 O'clock. The proportions of these data may be seen in the graphs fig. 4, 5, 6 and fig. 7. The monsoons do not occur in the coastal plain of Padang. The wind system in Padang is entirely dominated by land and sea wind but the sea wind is always stronger than the landwind so that it may safely be assumed that the direction of the wind of 9 and 15 O'clock predominates (maximal wind-effect). So it appears to have been the wind effect of the seawind, i. e. the product of the wind frequency given above and the velocity of the wind, about which we lack detailed data here, which we have recognised in the structure and situation of Pasir Ketjil (cf. fig. 2 and 3 with fig. 5 and 6). Although the frequency of the land wind is rather great (fig. 4 and fig. 7) its strength must be very little, for in the structure of Pasir Ketjil no influence, that can be assigned to the land wind, is apparent. We find here at the same time a confirmation of the conclusion we drew some time ago, namely that: data on the wind-frequency only (the prodominant direction of the wind) may sometimes give an indication as to the probable wind effect, but only when the wind in question is not obstructed by mountains. Finally the year-averages of the wind-frequencies of land- and seawind have been taken together in the graph fig. 8. So the part on the right side of the N.S. axis (landwind) can be practically ignored for geological considerations. The structure of the island Poeloe Pasir concurs with that of Pasir Ketjil. Along the coast of Emmahaven occur old and sligthly raised coral limestones. On plate 4 a section of this fossil reef is represented (equal scale for length and height) based on 15 seperate bore-holes; the data have been published by professor Sluiter. From this valuable section we may gather the following particulars: I. The reef has not grown on a rocky volcanic substratum or against the andesitic coastal lavas, but rests entirely on the muddy bottom of the bay, as is also the case with the reefs in he Bay of Batavia, the Thousand-Islands, and the Spermonde Archipelago. II. On the silty bottom rests a layer wherein mud and coral débris have been found. This shows that, in the innitial stage, only a few, branched coral species could grow while the vigorous sedimentation of silt was going on. III. These branched corals gradually attained to larger numbers and formed the basis for the actual, more cohesive reef, which was recognised in the drillings as „coral débris, branched type.” Just as with the reef near Krakatoa (bibl. 5) which grew under the unfavourable influence of strong sedimentation, and with the fossil corals of the Domaring and Menkrawit layers of E. Borneo, which grew under similar circumstances, we see here also that the branched types of growth (unfortunately we are not able to furnish a statement of genera in this case) are the pioneers. In the later, further stages of development of the reef, it is these types again that grow on the outside of the reef (see section of bore holes 9—12). IV. Only in the older stage of development of the reef, bigger globular coral growths appear by the side of the branched varieties. This is again shown by a different indication in the section. V. Later on, to all probability the reef was then already dead and „raised”, hill side waste has fallen over it (section of bore hole 3 and 4). VI. For the sake of clearness I have circumscribed the real, more compact reef with a thick line. Thus it is clearly shown that the basis of the reef, especially in the centre, lies deeper than the adjacent bottom of the bay. Sluiter has tried so explain this by supposing that the reef, when once it was growing into a compact mass has, to a certain extent, gradually sunk into the bottom of the bay. The same phenomenon was noticed while drilling on small islands in the bay of Batavia (bibl. 2). Incidentally I have already pointed out that: „this may be caused „for a large part by aggradation of the bottom as the rivers always „bring a large quantity of silt into the bay.” (bibl. 3, pag. 40). In reality sinking of the reef as well as aggradation of the bottom by a supply of silt in the bay, have probably taken place. Judging from the available data it does not seem possible to me to ascertain the extent of their respective influences. But, the configuration between bore-holes 12—14 and 3—5 of the section given here, as well as the structure and the situation of the island Edam in the bay of Batavia (see bibl. 2, fig. 3) seem to me to point to a real sinking of the reef into the soft bottom. The opportunity to pay a short visit by motor launch to the neighbouring island Poeloe Pisang (besar 1)) happened to present itself (see fig. 9). Against a nucleus of volcanic rock there is, at the South and East side, a plateau of coral limestone bordering on the sea. Part of this limestone has disappeared, probably, in consequence of the marine erosion. In this way a wave-cut rock bench has developed, which, however, is still situated above the normal low-water mark (see section fig. 10). At high tide sea covers this rock bench; the line to which the water then extends is marked by a „storm” rampart of fine white coral shingle. In this area there occur also erosion canals, simmilar to those of Tji Laoet Eureun (S. coast of Java) described at length by Dr. J. Cosijn and the present writer (bibl. 8). See Pl. 2, fig. 13 (breakers in the distance). If we suppose that the reef had originally grown to the level of normal low-water mark, then it follows from the present situation that the amount, either of the sinking of the level of the sea or of the rise of the level of the land, must have been between 3 and 5 metres. The exact determination of this amount was not possible because at the spot where it could have been ascertained best i. e. close to the nucleus of the island, observation was obstructed by débris that had fallen down. We are brought to a simmilar conclusion in Emmahaven itself where in some places (see Pl. 3, fig. 14) effects of abrasion are still to be seen above the present level of the sea. So these observations again support Daly's theory of a recent world wide sinking of the ocean-level.