As I have before said, what we did was to put the cap on the pyramid that had been built by Colonel Rondon and his associates of the Telegraphic Commission during the six previous years. It was their scientific exploration of the chapadao, their mapping the basin of the Juruena, and their descent of the Gy- Parana that rendered it possible for us to solve the mystery of the River of Doubt.

From Vilhena we travelled in a generally northward direction. For a few leagues we went across the chapadao, the sands or clays of the nearly level upland plateau, grassy or covered with thin, stunted forest, the same type of country that had been predominant ever since we ascended the Parecis table-land on the morning of the third day after leaving the Sepotuba.

It is the kind of country known to the Brazilians as chapadao pronounced almost as if it were a French word and spelled shapadon. Our camp on the fourth night was in a beautiful spot, an open grassy space, beside a clear, cool, rushing little river. We ourselves reached this, and waded our beasts across the deep, narrow stream in the late afternoon; and we then enjoyed a bath and swim.

Then, at about the point where the trail dipped into a basin containing the head-springs of the Ananas, we left this type of country and began to march through thick forest, not very high. There was little feed for the animals on the Chapadao. There was less in the forest. Moreover, the continual heavy rains made the travelling difficult and laborious for them, and they weakened.

The country was like that around Campos Novos, and offered a striking contrast to the level, barren, sandy wastes of the chapadao, which is a healthy region, where great industrial centres can arise, but not suited for extensive agriculture as are the lowland flats. For these forty-eight hours the trail climbed into and out of steep valleys and broad basins and up and down hills.

With the chapadao we said good-by to the curious, gregarious, and crepuscular or nocturnal spiders which we found so abundant along the line of the telegraph wire. They have offered one of the small problems with which the commission has had to deal. They are not common in the dry season.

With the chapadao of the Parecis plateau we came to a land of sand and clay, dotted with lumps of sandstone and pieces of petrified wood; this, according to Oliveira, is of Mesozoic age, possibly cretaceous and similar to the South African formation. There are geologists who consider it as of Permian age.