37 
pressions, by means of printer’s ink, from these heterogeneous 
forms of leaves, in order to prove how unsafe all those distine- 
tions are, which have been derived from the leaves only. The 
dong known but yet very imperfect method of ectypa is parti- 
eularly advantageous for this and similar purposes, as it offers 
to much-occupied travellers the means of procuring in a few mi- 
nutes the most correct outlines. 
The more the Cinchona trees vary in the shape and smooth- 
ness of the leaves, according to the altitude in which they grow ; 
to the severity or mildness of the climate ; to the trees standing 
singly, or being closely surrounded by other plants ; to the luxu- 
riance of growth, and greater or less humidity of the soil; the 
more necessary is it, with regard to the diagnostic indications, to 
pay attention to the form of the flowers, particularly to the 
Yength of the anthers, to the proportion between the stamens 
and anthers, as also between the free ‘and the adherent par 
of the filaments. It is not sufficient to examine the species in 
such as have: a smooth or hirsute corolla, or the stamens exsert- 
ed, or inclosed in the tube of the corolla, An attentive observer 
fnds in almost every species a striking difference in the struc- 
ture of the corolla. Thus, the C. purviflora, Mut. has pubescent 
filaments,.and dilated at the base. C. macrocarpa, Vahl, anthers 
nearly sessile, placed in the upper part of the tube of the corolla. 
C. oblongifolia, Mut. filaments very short, anthers situated below 
the middle of the tube of the corolla.» The Cinchona o alifolia, 
Mut., or the white Cinchona, varies frequen’ ly wi ix t 
seven, the C. Condaminea with from three to four stamens only. In 
the first, the limb of the corolla is frequently found divided into 
six or seven, in that of the latter, mostly inte four segments. In 
the Cascarilla fina de la Provincia de Jaen, which M. Bonpland 
intends shortly to describe, t found the anthers always shorter than 
L 
