IN THE CINCHONA FORESTS 289 
Asdepiadece^ 4. — All milky twiners. This order, like the pre- 
ceding, has its principal seat in the hot region, but is by no means 
confined to it, for two or three slender Cynoctona are frequent 
in the cooler parts of the Andes, trailing over the hedges of Cactus 
and Agave. 
So/anacecB, 5. — In this order, also, my collection contains a 
very small proportion of the species existing in the Red Bark 
woods. Shrubby Solana are almost endless, and two species rise 
to trees. Two or three species of Cestrum also occur as slender 
trees. 
Cordiacece^ i. — A Cordia, a stout sarmentose species, which 
threads about among the trees up to a considerable height, though 
it never actually twines. 
CofivolviilacecE. — ^This order see'ms confined to a couple of 
Ipomaese, both occurring very rarely. 
MyrsinecE^ 2 (or perhaps 3). — The most remarkable of all the 
plants I gathered is a Myrsinea, though, as it grows only at from 
5000 to 7000 feet, it barely touches the frontier of the Red Bark 
region. It is an arbuscle of 8 to 10 feet, bearing a coma of large, 
long, deep green coriaceous leaves, so that without flower it has 
quite the aspect of a Grias ; but above the leaves there is a mass, 
the size of the human head, of densely packed panicles and 
minute flowers, afl of the same deep red colour. I have not 
previously seen any Myrsinea at all resembling it in habit ; but 
I have examined it sufficiently to state with confidence that it 
belongs to this order, although probably to an undescribed genus. 
LabiatcE^ i. — Besides the solitary species gathered, there exist 
two species of Hyptis, one of them apparently H. Siiaveolejis ; 
but this order is always scantily represented in the forest. In 
cane-fields at San Antonio I saw a Stachys with small white 
flowers. 
Verbe7iace(E^ 2. — One of them a prickly suffruticose Lantana, 
threading among the bushes up to 18 feet in height; the other a 
woody twiner, with pretty waxy flowers, flesh-coloured externally, 
but the lim^b purple within ; it is probably a Citharexylon, allied 
to C. scande9is, Benth. (gathered on the Uaupes), though the 
habit is totally different from the arborescent Citharexyla which 
grow in the cooler parts of the Andes. A Duranta was noted at 
San Antonio. A Stachytarpheta, which I take to be S. Ja??iaice?isiSy 
and is known in Peru and Ecuador as "Verbena," seems to follow 
the steps of man in the Cordillera from near the plain up to 
10,000 feet. At Limon it exists sparingly as a weed. Another 
species of the same genus, with very slender spikes and small 
lilac flowers, abounds in open places. 
Gesneracece, 17. — The abundance of this family is one of the 
distinctive features of the Red Bark woods. One group, comprising 
several species, has a woody rhizome, creeping up the trees, and 
VOL. II U 
