Cuap. IX. DESCENT OF THE TAPAJOS. 253 
her third quarter, and we had many successive days and nights of clear, 
cloudless sky. I believe this wind to be closely connected with the 
easterly trade-wind of the main Amazons; indeed, to be the same, 
reflected from the west after the land-surface in that quarter has been 
cooled by it to a much lower point than the sun-heated surface of the 
stagnant Tapajos. The wind always arose in the morning after the air 
in the direction of the north-west had been further cooled by radiation 
of heat during the night; and it ceased in the evening, when the 
equilibrium of temperature between the Tapajos and the Amazons had 
become restored. The light land breeze from the east, which always 
began to blow soon after the strong north-wester ceased, is attributable 
in like manner to the wooded surface of the land being then cooler than 
the air on the river. The terral lasted generally from 7 until 11 p.m., 
but after midnight it usually veered gradually to the north-east, and 
blew rather freshly from that quarter towards sunrise. 
Point Jaguarari forms at this season of the year a high sand-bank, 
which is prolonged as a narrow spit, stretching about three miles 
towards the middle of the river. We rounded this with great difficulty 
in the night of the z9th; reaching before daylight a good shelter 
behind a similar sandbank at Point Acardatingari, a headland situated 
not more than five miles in a straight line from our last anchoring place. 
We remained here all day: the men beating “#6 in a quiet pool 
between the sandbank and the mainland, and obtaining a great quantity of 
fish, from which I selected six species new to my collection. We made 
rather better progress the two following nights, but the terral now 
always blew strongly from the north-north-east after midnight, and thus 
limited the hours during which we could navigate, forcing us to seek 
the nearest shelter to avoid being driven back faster than we came. 
On the znd of October we reached Point Cajettiba, and had a 
pleasant day ashore. The river scenery in this neighbourhood is of the 
greatest beauty. A few houses of settlers are seen at the bottom of 
the broad bay of Aramana-i at the foot of a range of richly-timbered 
hills, the high beach of snow-white sand stretching in a bold curve from 
point to point. The opposite shores of the river are ten or eleven 
miles distant, but towards the north is a clear horizon of water and sky. 
The country near Point Cajettiba is similar to the neighbourhood of 
Santarem : namely, campos with scattered trees. We gathered a large 
quantity of wild fruit: Caju, Umiri, and Aapirdnga. The Umiri berry 
(Humirium floribundum) is a black drupe similar in appearance to the 
Damascene plum, and not greatly unlike it in taste. The Adpirdnga is 
a bright vermilion-coloured berry, with a hard skin and a sweet viscid 
pulp enclosing the seeds. Between the point and Altar do Chad was 
along stretch of sandy beach with moderately deep water; our men 
therefore took a rope ashore, and towed the cuberta at merry speed 
until we reached the village. A long, deeply-laden canoe with miners 
from the interior provinces here passed us. It was manned by ten 
Indians, who propelled the boat by poles; the men, five on each side, 
trotting one after the other along a plank arranged for the purpose from 
stem to stern. 
It took us two nights to double Point Cururti, where, as already 
