IN HUMBOLDT'S COUNTRY 441 
shrubs were a Byrsonima, apparently a form of B. spicata, a Guat- 
teria, a Pagamea, etc. Under large stones grew the most delicate 
little fern I have ever gathered, looking at first glance like a 
miniature AUosorus crispus^ but in reality more allied to Schizjea, 
and along with it a small grass with broad truncate -cuneate 
leaves, which I had gathered abundantly in similar situations by 
the cataracts of the Rio Uaupes. Rooting into clefts of the rocks 
and twining on adjacent shrubs or over the rocks themselves, 
grew an Asclepiadea, with narrow leaves and minute white flowers, 
looking not unlike Galium saxatile. In moist, rocky places I found 
a shrub of about 4 feet high, with long pinnate branches, minute 
rigid leaves ending in an arista, and solitary axillary fruits the 
size and colour of haws. It is quite new to me and seems to me 
to be a capsular Myrtacea, but I have not examined it closely. 
There were also a few Melastom.acese and other things. 
The savannas near the pueblo (Esmeralda) were mostly dried 
up by the heat. The grasses showed only withered culms, but I 
recognised among them several species of Paspalum, Setaria, 
Andropogon, Trichopogon, etc. I crossed the two first savannas 
in the direction of Duida, but found scarcely anything in flower. 
It is curious that on the second of these the only tree, besides the 
Moriche palm, is a Qualea, which seems to me identical with one 
gathered on a low campo of quite similar character opposite the 
Barra, and which Mr. Bentham has called Q. retusa. The tree 
at Esmeralda had neither flower nor fruit, and if it was in the 
same state at the period of Humboldt's visit, most probably he 
did not gather specimens. 
On a savanna which extends towards the Guapo there were 
still some moist places left, and in them I gathered several inter- 
esting little plants. They include two Burmanniaceae (perhaps 
the true Burmannias), one of them with a violet flower far larger 
than I have seen in any other species of the tribe ; four Gentiane^e, 
of which two are Lysianthi, the one a small species with a bright 
blue flower, exactly resembling Cat7ipanula rotundifoUa^ the other 
a tall plant with green flowers ; the other two species are minute 
things allied to Schiibleria, three or four Xyride^e, two Ascle- 
piadeae, two minute Rubiacese with yellow flowers, species of 
Perama, one of them P. hirsuta (gathered also at Santarem), three 
Polygalse, in one of which I recognise P. subtilis, H. B. B., and 
several others. 
I gathered also all I could on the banks of the Orinoco, in- 
cluding the Palma Jagua, whose beauties are so highly and so 
justly eulogised by Humboldt in his Aspects of Nature. It is an 
undescribed Maximiliana, and I brought away with me specimens 
and notes on the living plant which will enable me to describe it. 
There were two splendid trees of it in the mouth of the Casi- 
quiari. I had one of them cut down and a frond and a spadix 
