ACROSS UNKNOWN SOUTH AMERICA 
2,500 feet, that of the bottom 2,450 feet. We continued 
our journey on the top of the plateau, with slight undula¬ 
tions varying in height from fifty to seventy feet. Snuff- 
coloured soil and red sand were invariably noticeable on 
the higher points, and grey ashes in the lower points, where 
erosion had caused depressions. 
Then, farther on, the plateau, with an elevation of 
2,450 feet, was absolutely flat for several kilometres, and 
showed sparse vegetation and miserable-looking ansemic 
trees, the thin soil over solid rock affording them 
inadequate nourishment. 
Eighteen kilometres from our last camp we came upon 
another oval basin (elevation 2,400 feet above the sea 
level), extending longitudinally from north-northeast to 
south-southwest. On its huge deposits of cinders grew 
deliciously green, fresh-looking, healthy grass, and a thick 
clump of burity palms, and birero trees of immense height 
and thick foliage. Those beautiful trees were called by 
the people of Goyaz " cutibd” and “ r pintahyba.” They 
were marvellous in their wonderful alignment among the 
surrounding circle of gorgeous palms. The latter were in 
their turn screened in their lower part by a belt of low 
scrub, so that upon looking at that oasis one could hardly 
realize that it had not been geometrically laid out by the 
hands of a skilful gardener. 
On the outer rim of the cuvette, away from the 
moisture, hundreds, in fact, thousands of cones, cylinders, 
and domes, from four to six feet high, the work of ants, 
could be seen, all constructed of bluish grey ashes. 
We had here a wonderful example, quite sufficient to 
persuade the most sceptical, of the influence of agglomera¬ 
tions of trees in the formation of clouds. The sky was 
perfectly clear everywhere except directly above the 
extensive cluster of trees in the large cuvette. Quite low 
down, only a hundred feet or so above the top of the trees, 
hung a heavy white cloud. It was a windless day. The 
290 
