painted in colours, and plain stelæ occur at Tikal and elsewhere, which probably once bore some painted design. The evident relation which Maya sculpture bears to wood-carving renders it probable that many wooden stelæ may once have existed, especially in the earliest times. It is worth noting that at Copan the stone foundations which support the stelæ enclose small cruciform vaults, reproducing in miniature the souterrains found in connection with some of the buildings at Mitla. [Illustration: _PLATE XXVI_ _Photo. Col. F. H. Ward_ 1 _After Maudslay_ 2 MAYA 1. SCULPTURED MONOLITH P., QUIRIGUA, GUATEMALA 2. PLAN OF RUINS AT COPAN, HONDURAS] [Illustration: FIG. 81.--Stone altar (Q) at Copan. (_After Maudslay_)] At present the attention of most archæologists has been confined to the ruins themselves, and little search has been made for the sources whence the building-material was obtained. Quarries however have been located in the neighbourhoods of Naranjo, Copan and Quirigua, and the presence at the last site of canal-like excavations suggests that the blocks were conveyed thither by water. A thorough investigation of such quarries as can be discovered would be most valuable in casting much light upon the process of working stone among the early Maya. In the east of the central area the most important site is that of Tikal; the ruins are extensive, and suggest that the place was long an important centre. The dates attached to the monuments are of great interest (for the comparative dating of the monuments, see Appendix III), since they range from an early period, and the extreme solidity of the buildings as compared with those of other sites seems to hint that they were the work of more primitive architects than those of Palenque for instance. The pyramids are of interest, since they present a far steeper pitch than those of any other locality (Pl. XXV, 2; p. 332). Certain of the stelæ and inscriptions are of a rather rude and archaic character, but the wood-carvings (Fig. 48; p. 225) are of particular merit, and are very similar in style to the stone lintels in low relief of Menché. The fact that the wooden lintels have been preserved here while they have for the most part decayed at other sites, is no argument against the antiquity of Tikal, for they are situated well within the buildings, which are, as remarked above, of very solid construction, and furnished with small single doors, besides being raised far above the ground on lofty foundation-mounds. On the whole I think that the architectural and artistic evidence supports the early dates which appear on the stelæ, and I should place Tikal as perhaps the earliest Maya site of which we have definite knowledge, though it is evident that it was inhabited well on into later times when the Maya had become far more expert in stone-carving than when the first buildings were erected there. The neighbouring site of Naranjo