country. Behind them was a flat-topped
plateau with a small cone rising above it. The sides of the latter
plateau formed a steep escarpment. To the south-east was a domed plateau,
red in its lower section, green on the top. Between this plateau and the
last wall-like mountain, several hundred feet in height, stood a conical
peak with a natural tower of rock upon it.

Beyond, to the south-east, could just be perceived two pyramidal
mountains, but they were very distant and scarcely visible. The valley
itself was greatly furrowed in deep, long channels. Due south were
dome-like mounds--each of these, mind you, standing out individually upon
an almost flat plain.

In the north-western corners of the great quadrangular Paredao rock I saw
a spot where it would have been quite easy to climb up to the summit, as
portions of the rock had crumbled down and had left an incline. But I had
no object in making the ascent on that side, especially as I had already
obtained the view I required from the south side. Also because I was
heavily laden, carrying cameras, aneroids, a large prismatic compass, and
three heavy bags of money slung to the belt round my waist, and did not
feel up to the extra and useless exertion. Great arches with a span of
over 80 metres were to be seen in the lower part of the western wall. To
the south there was a huge spur of lava with the geometrical pattern upon
its surface we had already observed elsewhere. In this particular case,
too, it appeared to me that the peculiar net of surface channels had been
formed in coming in contact with the air, and not underground in the
boiling cauldron of the volcano when the ebullition of the rock ceased.
They were only found at a lower elevation because they had gone down with
a great subsidence which had taken place, and in which neither the
quadrangular Paredao Grande, nor the peculiar isolated mountains we had
observed from its height, had been affected. They had remained standing
when all the rest sank for some six hundred feet and, in places, more.
That might perhaps account for the extraordinary shapes of all those
mountains, which could not otherwise be explained.

[Illustration: The Paredao Grande (Matto-Grosso).]

At the foot of the vertical giant block on the west many domes of lava,
channelled in a quadrangular network pattern, and ridges and cones, were
found, all with a slope to the west. I had a great struggle in my
research work that day, owing to the