and by four small flowers, which latter alone appear to be _repoussé_, the rest intaglio work. The button (No. 380) has no border; the whole space is filled by concentric circles in intaglio, with only two or three small ornaments in _repoussé_. On the other hand, on the button (No. 382) all the ornamentation is produced by _repoussé_ work; even to the border line of the interior lozenge, within which we see a circle filled with small ones, and above and below it a curious sign, which is very frequent on the Trojan whorls. On the large button (No. 383) we again see a border filled with twenty-eight crosses, and in the interior lozenge, in the middle, a double circle, a 卍 with curved arms, in each of which, as well as in the centre, is a point to mark the nails by which the two pieces of wood for the production of the sacred fire were fastened. The two acute angles are here again filled up with the same sign which we noticed on the preceding figure. The border with the crosses is of _repoussé_ work; the circle, with the 卍, of intaglio. In the figure No. 384, the border of the interior lozenge is ornamented with horizontal strokes; in the interior we see, in the middle, two spirals like Omegas, standing opposite each other, and in each acute angle a small ornament, perhaps a flower; the latter and the border are here the only work in _repoussé_, the remainder being intaglio. The button (No. 386) has an identical ornamentation of two spirals which stand opposite each other, and resemble Omegas. Finally, the large button (No. 385) has a broad border filled with twenty-eight small circles in _repoussé_ work, and of the same work is also the small encircled cross in each acute angle, whilst the large circle with the 卍 in the centre is of intaglio work. On the reverse side, the wood of all these twelve cross-like buttons is carved much like our shirt-studs, with the sole difference that the lower side is here of an oval form. Thus, there can be no doubt that all of them were used as ornaments on the clothes, but, of course, they can never have served as real buttons. All these buttons show unmistakeable marks of the funeral fire, but as the wood has been preserved, there can be no doubt that the fire was not intended to reduce the bodies to ashes, or to destroy the ornaments with which they were laden. There were further found with the five bodies of the fourth tomb 110 small golden flowers, in the shape of the four represented under Nos. 387-390, and 68 gold buttons without any ornamentation like Nos. 391 and 392; 134 round pieces of gold plate with a border, like Nos. 395 and 396; and 98 large shield-like pieces of gold plate in _repoussé_ work, with two rope-like borders, like No. 402 (p. 264). Not one of these 410 round pieces of gold plate shows any sign of having been fastened on wooden buttons, and we conclude from this that they must have been merely attached with glue to the clothes and drapery of the deceased.