CONTRIBUTIONS TO BOTANY. 
299 
60 in racemo, hinc crebriter conglomeratis, pedicellis flore 
longioribus, cano-sericeis. — Chile, in Prov. Rancagua. — v. s. 
in herb. Mus. Paris. (Bertero, 188.). 
This species has a different aspect from the former : the 
branchlets are quite straight and sericeous, and the spines much 
thinner, not exceeding 4 lines in length, and foliiferous in 
the middle. Its leaves are much thinner in texture, neither 
sulcated above, nor costately nerved beneath, the nervures being 
very fine, and scarcely prominent ; they are remotely denticu- 
lated with extremely short teeth, sericeous on both sides, silvery 
below, 9-10 lines long, 4^ lines broad, on a petiole 1^ line long ; 
the stipules are comparatively large, concave, opposite, and 
meeting each other in the middle of the stem, which they thus 
completely embrace ; they are deeply bifid, 1^ line long, exter- 
nally sericeous, internally and on the margin red. The race- 
mose branchlets are 1 inch long, and bear from forty to sixty 
flowers, crowded in the manner above specified *. 
9. SCYPHARIA. 
Under this head are brought together a few spinose shrubs, 
or small trees, more or less fohaceous in habit, distinguished by 
their opposite leaves ; small flowers, with an urceolate 5-fid calyx •, 
small, deeply emargiuate, naviculate petals, enclosing as many 
stamens ; and a 2-celled ovary ; they are very different from 
Trevoa, to which they approach the nearest in their floral struc- 
ture. The fruit of most of them is unknown ; but Kunth de- 
scribes that of his Rhamnus Guayaquilensis as being an oval 
drupe, somewhat fleshy, smooth, and bilocular, or by abortion 
1-celled, as in Trevoa. The manuscript characters contributed 
by Bonpland are not to be relied on, as Kunth found them 
disproved by the analysis he made of the dried specimens ; 
the latter may therefore be accepted as the more correct. They 
correspond thus with the general features of the Colletiece, and 
the species here associated harmonize sufficiently together. 
Kunth describes the nut as containing two hard and crustaceous 
pyrense, one of which he found abortive, the other emptied by 
caries : this coating evidently corresponds with the hard crusta- 
ceous covering of the seed common to all the Colletiece. The 
genus, therefore, in the absence of perfect specimens, can only 
be characterized imperfectly, and its features will no doubt 
require some modifications whenever better evidence can be ob- 
tained j but in the mean while, the facts already established tend 
to confirm its validity. Its name is derived from o-/cu^o?, scy~ 
phus, in allusion to its cup-shaped disk, after the example of 
* This species is represented in Plate 41c. 
2q2 
