206 MR. SPRUCE'S BOTANICAL EXCURSION 
it daily. We breakfasted upon it, and, unless I am mistaken, I 
never tasted such excellent arrowroot in England. I have lately 
learnt that the Arrowroot plant grows high up the Rio Tapajoz in 
great plenty. 
Leaving the Rio Caipurü we re-entered the Trombétas, and passed 
. through the Furo de Chiriry, with the object of coasting along the left 
bank of the river. The shores of the furo were lined on both sides 
with a beautiful grass, apparently some species of Luziola, growing 
deep into the water. The flowers were monecious, the male panicle 
uppermost, with very fugacious solitary flowers, each containing six 
yellow stamens, enclosed in glumes of the most delicate pink colour, 
with purple streaks. - In the axil of a lower leaf was the female panicle, 
much contracted, with slender bristly branches, the lower ones deflexed 
when in fruit. From the base of the dilated leaf-sheaths proceeded 
. pencils of capillary radicles, floating in the water or rooting in the mud. 
Emerging from the furo, we entered a long reach of the river, bearing 
north-west by north half west, apparently about a mile and a half wide ; 
but this, it must be recollected, is not the whole breadth of the river, 
the land on our right being an island. No sandy beach here ; skirting 
E: the forest is a strip of low entangled shrubs, consisting chiefly of a 
—  Myríacea, called Acara, with red edible fruit the size of sloes, and 
. another shrub called Juranduba, also with red fruit, but so acid as to 
- be uneatable. Five hours’ sailing in this reach with a light breeze— 
our rate probably two miles an hour—brought us to a narrow igarapé, 
. entering on the left from the Lago de Samaüma. At 64 p. m. (we had 
started not much before 10 in the morning) the river turned to about 
— N.N.W., still preserving the same width and with nearly parallel 
. straight banks. In an hour more the shore gradually rounded to north- 
= west by west. On the right bank enters a slender still igarapé, from 
a lake called Auapetüa. After passing this, we erossed to the left 
bank, and gained the point at the junction of the Rio Cuminá with 
the Trombétas, which was to be our halting-place for the night. 
— December 22.—1l was surprised to find the spot our pilot had se- 
lected for our night’s lodging already occupied by a detachment of 
five soldiers from Obidos, designed for the protection of the turtle-beds 
in the Trombétas and Aripecurü, and denominated an Estacamento, 
from the deposits of turtles’ eggs being marked out by stakes (estacas). 
I learnt from the officer in command that the Estacamento commenced 
