5o6 NOTES OF A BOTANIST 
undermining the high banks, yet still supporting them whilst the 
flood lasts ; afterwards, when the waters receding no longer sustain 
them from beneath, large portions of earth and forest fall simply 
by their own weight. 
The Ilhas de Caapi'm or Grass-islands I now find are chiefly 
floated out of lakes by the rising of the waters of the Amazon, 
which thus fulfil the double purpose of calling certain species into 
existence and afterwards bearing them away to the ocean. The 
chief, often the sole, constituents of the Grass -islands are the 
Canna-rana (wild sugar-cane) and Piri-membeca (Brittle-grass), 
Pa?iicu7n spectabile and Paspalum pyramidale. For the production 
of these plants white water is essential, as is proved by their 
absence from the whole Rio Negro and from the Rio Trombetas 
above the Furo, through which the waters of the Amazon are 
poured into the lower part of the Trombetas, where both these 
grasses are abundant. 
Lakes are usually of black water, but those into which white 
water enters in the rainy season invariably produce these two 
grasses, sometimes in such abundance that they become periodi- 
cally choked up, as I witnessed in two small lakes near Manaquir^, 
through which it is every year necessary to .cut a passage. 
In returning from Manaquiry, the Solimoes was so violently 
agitated by an easterly wind as to render it perilous crossing it in 
my small and deeply-laden montaria. We contrived, however, to 
reach a small Ilha de Caapim which was floating down at about 
three hundred yards from shore, put ourselves into the centre of 
it, so that the force of the waves was broken ere they reached us, 
and thus floated onward at our ease. Whilst my man composed 
himself to sleep, I amused myself with examining the composition 
of my novel berth. It consisted of but one species of grass, 
Paspalum pyramidale^ and after several futile attempts I 
succeeded in drawing out an entire stem, which measured 45 
feet in length and contained seventy-eight joints. It was simple, 
though others possessed two or three branches ; all the joints 
save the three or four uppermost ones gave out rootlets, and 
several of the lowest internodes were in a semi -putrid state. 
Floating on the water, and kept in by the grass-stems, were an 
Azolla, two Salvinise (one of them new to me and both in fruit), 
and a few barren plants of a small Hydrocharidea, and a small 
Pistia. It is to be noted that these Ilhas de Caapi'm are quite 
different from the floating rafts encountered by Humboldt on the 
Orinoco : similar rafts I have seen on the Amazon. 
[The following Notes were also written during 
the same period as the last ; and with a short account 
by myself of the present state of the rubber industry 
