GIGANTIC FLAT TABLELANDS 
tically to the west. At its foot the Rio Pedra Grande 
had its birth and then flowed westward into the 
Rio Roncador. Four gigantic flat tablelands stood im¬ 
pressively in a line. Three more, equally impressive, 
loomed in the southwest. Other minor ones, quite wall¬ 
like, rectangular in vertical section, appeared in the blue 
distance, while the horizon was barred by a long, flat 
plateau. 
Looking north as we descended from the tableland, 
we found on our left another extinct crater, semicircular 
in shape, with several superimposed strata of lava, each 
about one foot thick, capping its lip, which was broken 
up into three sections. The valley below that crater 
formed a cuvette, the bottom of which (elevation 2,200 
feet), showed deep erosion by water in one or two places. 
Sand covered the lava-flow, which had travelled north¬ 
ward. Quantities of heavy, spherical, bullet-like blocks 
of hard-baked rock were scattered all about, evidently 
shot out of the crater when active. 
We had travelled eighty kilometres from Cayambola 
in three days, and we had reached a spot of slight, well- 
rounded undulations where grazing was fair. I decided 
to halt early in the afternoon, more particularly as this 
spot appeared to me to have been at one time or other 
submerged; probably it had been a lake bottom. I had, 
since the beginning of my journey, been searching every¬ 
where for fossils, but in vain. I had not seen the vestiges 
of a single one. Personally, I was persuaded that Central 
Brazil could well be geologically classified in the archaic 
group, the most ancient of the terrestrial crust, and con¬ 
sisting (in Brazil) chiefly of gneiss, mica schists, and 
granite, solidified into their present form by intense 
eruptive phenomena and dissolved, not by immersion in 
ocean waters, as some suppose, but by deluges of such 
potentiality as the human mind can hardly conceive. 
It was quite enough to visit the central plateau of 
323 
