BRANNER: TUE STONE REEFS OF BRAZIL. 85 



sound. The surface of the Rio Formoso reef, however, is uot, on the 

 whole, as ragged and rough as those of many other of the stone 

 reefs. 



The broken stone generally shows quartz sand mixed with some 

 pebbles and shells. The pebbles are also of quartz, and the shells are 

 of bivalves, gasteropods, etc., such as live at present along the coast in 

 the vicinity. The shells most abundant are tlie prettily colored bivalves 

 Venus, popularly known here as mariscos. 



In many places, in fact, in almost all parts of the reef, cracks appear 

 to have been formed and then to have been filled up again with sand 

 that has hardened like the rest of tlie reef. These cracks have generally 

 the same direction as the reef, but they sometimes cross it at right 

 angles, and at other times diagonally. In two places the reef has small 

 breaks nearly or quite through it, crossing diagonally, and with the 

 appearance of being due largely to the presence of the cracks. 



From the land end on the south outward there is no break of impor- 

 tance until about halfway the whole length of the reef, where there is 

 a baretta or passage for canoes and jangadas. This break is known as 

 the harreta das jangadas. It is here that the reef receives the full 

 force of the ebbing tide as it comes down from the river. From the 

 harreta das jangadas to the outer end, the side facing the river is 

 either perpendicular or overhanging. In tlie harreta itself the rocks in 

 the bottom of the passage are hard and appear to be the same as those 

 on the surface of the reef at other places. I suppose the rocks in the 

 bottom of the passage to have formerly stood even with the surface of 

 the reef, but being undermined by the strong sand-laden currents, they 

 have lost their support and fiillen where they now lie. These blocks, or 

 more properly slabs, are all large and broad, and of the same average 

 thickness as the harder upper part of the general body of the reef. 

 From this place to the outer end, five hundred and seventy metres, 

 there are two other breaks, both of them smaller and of less importance 

 than this one. 



The whole of the riverward face of the reef is undercut. Blocks have 

 broken away where left without support, and in one place especially, 

 very large pieces have fallen, and, with one end held fast by the sand 

 and mud of the bottom of the channel, remain with the other end 

 tilted high in the air. One of these is so high that it is not covered 

 even by the highest tides; this one is popularly known hereabout as the 

 Pedra de Nossa Senhora, or the Kock of Our Lady. 



At still another place the reef is entirely undermined, while the sur- 



