In the margin and bed only of the river are the rocks inscribed. On the margin they extend in some instances to 15 or 20 yards. Except in the rainy season the stream is dry. The rock is a silicious schist of excessively hard and flinty texture. The marks have the appearance of having been made with a blunt, heavy tool, such as might be made with an almost worn-out mason’s hammer. The situation is about midway between Serra Grande or Ibiapaba and Serra Merioca, about 70 miles from the coast and 40 west of the town Sobral. The native population attribute all the “Letreiros” (inscriptions), as they do everything else of which they have no information, to the Dutch, as records of hidden wealth. The Dutch, however, only occupied the country for a few years in the early part of the seventeenth century. Along the coast numerous forts, the works of the Dutch, still remain; but there are no authentic records of their ever having established themselves in the interior of the country, and less probability still of their amusing themselves with inscribing puzzling hieroglyphics, which must have been a work of time, on the rocks of the far interior, for the admiration of wandering Indians. Mr. Franz Keller (_a_) narrates as follows regarding Fig. 110: [Illustration: FIG. 110.--Petroglyphs on the Cachoeira do Ribeirão, Brazil.] I found a “written rock” covered with spiral lines and concentric rings, evenly carved in the black gneiss-like material, and similar to those of the Caldeirão. Looking about for more, I discovered a perfect inscription, whose straight orderly lines can hardly be thought the result of lazy Indians’ “hours of idleness.” These characters were incised on a very hard smooth block 3 feet 4 inches in length, and 3-1/4 feet in height and breadth. It lay at an angle of 45°, only 8 feet above low water, and close to the water’s edge of the second smaller rapid, the Cachoeira do Ribeirão. The transverse section of the characters is not very deep, and their surface is as worn as that of the inscription farther down. In some places they are almost effaced by time and are to be seen distinctly only with a favorable light. A dark brown coat of glaze, found everywhere on the surface of the stones, laved at times by the water, covers the block so uniformly well on the concave glyphs as on the parts untouched by instrument, that many ages must have elapsed since some patient Indian spent long hours in cutting them out with his quartz chisel. As the lines of the inscription run almost perfectly horizontally, and as the figures near the Caldeirão and the Cachoeira and the Cachoeira das Lages are so little above low-water mark, the present position of the block seems to have been the original one. * * * On the rocky shores of the Araguaya, that