lty of
finding it again in the tall grass and high vegetation, we were not able
to leave camp until the afternoon of June 18th. Soon after starting on
the march we went through a marvellous arch of thick foliage, creepers,
bamboos, and _akuri_ palms, previous to crossing a streamlet 9 metres
wide and 1 ft. deep--flowing towards the west. We had no end of trouble
near these streamlets, as they flowed between precipitous banks 50 to 70
ft. high. There was no trail. The animals frequently lost their footing
over the slippery, steep slope, and rolled down, baggage and all, until
they reached the bottom; or else they would sometimes stick half way down
against trees and liane, and we had the greatest difficulty in
extricating them again.

[Illustration: Quadrangular Rocky Mountain connected by Natural
Wall of Rock with the Vertical-sided Range in Background.]

There was a low range extending from north to south along the left bank
of the Rio Manso. From a hill 1,470 ft. high above the sea level on the
right bank of the river we saw a plateau in four terraces--the third of
the line of plateaux we had seen on our preceding march. Upon getting
higher we perceived to the south, beyond the four-terraced plateau,
another plateau with vertical walls, and to the south-west a high
double-humped dome--resembling Mount Vesuvius in Italy. Evidently one
more of the innumerable extinct volcanoes to be seen in that region.
The mountainous mass extended in a more confused form farther to the
south-west. On our side of the Rio Manso the country was gently
undulating--in fact, it formed many parallel ridges of low, well-rounded
hills with occasional deep hollows or basins between. One could not help
being particularly struck by the wonderful regularity and strong
similarity of the curves on the parallel hill ranges, as if all had been
turned out of the same mould. The hill-range we were on was 1,500 ft.
above the sea level. The others--excepting one or two--were lower.

There was an absolutely flat horizon line to the north, with no mountain
range in sight. The country opening up before us was from that point
almost entirely made up of campos, with _chapada_ or growths of trees
principally near streams in the valleys. We crossed a watercourse 30
metres wide and 1 ft. deep at an elevation of 1,350 ft. We called it the
Palmeira, owing to the many palms upon its banks. Here grew many great
_caja_ or _cajazeiro_ trees (of the genus Anacardiacea