BOTANICAL INFORMATION, 191 
cence of the stem, or of the upper surface of the fronds; the fruit too 
is elliptical, and the spathes foür or five in number. These are strik- 
ing differences, but I cannot yet pronounce on their value. 
A Palm much cultivated in the Barra and the adjacent sitios, and 
said to grow wild up the Rio Negro, is the Popúnha, which I suppose 
to be the same as the Pirijaó (Guilielmia speciosa, Mart.) mentioned by 
Humboldt as growing on the upper Orinoco. The fruit of this is per- 
haps more valuable as an edible than any other Palm-fruit; the sarco- 
carp contains a large quantity of starch, and it is sometimes developed 
to such a degree that the nucleus is quite obliterated. Eaten with 
salt, the boiled or roasted fruit much resembles a potato, but it is also 
very pleasant eating with molasses. A spadix of Popiinha, laden with 
ripe fruit, is one of the most beautiful sights the vegetable world can 
show: the fruits are of the clearest scarlet in the upper half, passing 
below into yellow, and at the very base to green. 
On a separate sheet I have written some account of a few of the 
edible fruits found wild in the forests adjacent to the Barra. "There 
are several others, which I have not yet obtained. Numerous Myrta- 
ceous and Melastomaceous fruits are eaten, but few of them possess 
any great excellence; perhaps the best are the Guayabas, which belong 
to various species of Psidium. A Melastomaceous tree sent from San- 
tarem, with fruit rather resembling a Guayaba externally, but twelve- 
celled, is very abundant here. The fruit is called Zapiri-guayaba, or 
the Tapir’s guayaba, but it is only insipid eating. The various species 
of Inga have the seeds enveloped in a sweet cottony pulp, which is 
very agreeable eating; the Zngd-sipé (of which I have already sent you 
the fruit) is the most esteemed. 
The ** Cow-tree" is represented on the Rio Negro by two dpocynea, 
the Cuma-í and the Cuma-aci, both species of Collophora, but only one 
of them known to Martius. The former is frequent near the Barra, 
and early in March was a great ornament to the forest, especially near : 
the river, being profusely clad with corymbose cymes of red flowers. 
It grows thirty or thirty-five feet high, with a diameter of about 
twelve inches, and the branches and leaves grow in threes. The milk 
flows out abundantly on a slight incision being made in the bark; it 
is of the consisteney of new milk, of the purest white, and very sweet 
to the taste. The Indian mode is to apply the mouth directly to the 
gash, and thus receive the milk as it oozes out: in this way I have 
many times partaken of it, without experiencing any ill effects. Its 
