ACROSS UNKNOWN SOUTH AMERICA 
all Central Brazil from west to east, roughly speaking, 
from the Madeira River to the Araguaya and beyond. 
After we had gone some five kilometres in a straight 
line from our camp to 10° bearings magnetic, we per¬ 
ceived a headland with a hill upon it 200 feet high. We 
had been greatly troubled in the afternoon for the last two 
days by heavy showers of rain and gusts of a north¬ 
westerly wind. Once or twice we became entangled in 
channels among the many islands, and had to retrace our 
course, but we went on until late in the evening, my men 
believing firmly that we had now reached civilization again 
and that the journey would be over in a few days. I did 
not care to disillusion them. 
Late at night we camped on a magnificent beach, 
1,000 metres long, at the end of Araujo Island, 1,200 
metres in length. 
We had gone that day, August nineteenth, 46 kilo¬ 
metres 500 metres. 
My men hung their hammocks on the edge of the 
forest. That camp was extremely damp and unhealthy. 
When we woke up the next morning all my followers 
were attacked by fever and were shivering with cold. 
We left at 7.30 a.m. under a limpid sky of gorgeous 
cobalt blue. We passed two islands — one 700 metres 
long (Leda Island), the other 2,000 metres (Leander 
Island). When we had gone but 11,500 metres we 
arrived at one of the most beautiful bits of river scenery 
I have ever gazed upon — the spot where the immense 
S. Manoel River or Tres Barras or Paranatinga met the 
Arinos-Juruena. The latter river at that spot described 
a sharp turn from 20° bearings magnetic to 320° bearings 
magnetic. We perceived a range of hills before us to 
the north. Close to the bank gradually appeared a large 
shed with a clearing near it on a high headland some 
200 feet above the level of the river where the stream 
turned. On the left bank, before we arrived at the 
228 
