COOK AND DOYLE—-NEW GENERA OF PALMS FROM COLOMBIA. 2338 
Flowers from June to October. Puerto Cabello, Colonia Tovar, Colombia. 
Moritz no. 914. 
EXPLANATION OF PLATES 61, 62.—Plate 61, Catoblastus pubescens (Karst.) Wendl. ; at 
the right a seedling of two years. Fig. 1, a pinna; fig. 2, an inflorescence; fig. 3, a male 
flower ; fig. 4,an anther; fig.5,a pollen cell, Figs. 14, 15, fruit of Catoblastus praemorsus 
(Willd.) Wendl. Reduced. Reproduced from Karst. Fl. Columb. 1: pl. 81. 1858-1861. 
Plate 62, flowers and fruit from the same plate at natural size. Fig. 6, tip of male in- 
florescence ; fig. 7, tip of female inflorescence; fig. 8, section of fruit; figs. 9 and 10, 
views of the seed; fig. 11, female flower; fig. 12, same after removal of calyx and corolla 
segments; fig. 15, section of pistil. 
As this species is the type of the genus Catoblastus, its characters are of 
special interest, but some of the most important are still in doubt. Klotzsch, 
and more recently Jahn, have described the carpels as distinct and with separate 
stigmas, and neither of these writers gives any intimation that the carpels are 
unequal at the time of flowering. 
Karsten’s Iriartca pubescens,‘ which Wendland placed as a second species of 
Catoblastus, has the stigmas united into a short, cylindrical style, not half as 
long as the very large fertile carpel. The figure of Catoblastus pubescens 
given by Drude in Engler and Prantl’s Pflanzenfamilien shows a long, columnar 
style rising above the fertile carpel and is 
likely to mislead regarding the principal 
generic character. Karsten says that the 
staminodia are wanting or very small and 
hidden under the carpels, whereas Drude’s 
figure shows large staminodia with anthers. 
To avoid further confusion from this erro- 
neous figure in so prominent a work of ref- FiG:41.—Ovary of Catoblastus drudei 
erence it may be best formally to recognize with fertile and sterile carpel, a 
the fact that Drude’s Catoblastus is a dif-  %taminode below. At the right, same 
+ og + : . in section. After Engler and Prantl. 
ferent species.” Unfortunately its origin is 
not indicated unless by the fact that Drude includes Peru in the range of the 
genus, while Venezuela is omitted. 
It is also evident that the true characters of Catoblastus are not to be learned 
from Drude’s account of the genus nor from Karsten’s beautiful figures of C. 
pubescens. It is quite possible that one or both of these species will eventually 
be removed from the genus. For the present it may be suflicient to note that 
C. pubescens has an obvious alliance with our new genus Catostigma. The com- 
pletely sessile stigmas of Acrostigma and the presence of the rudimentary male 
flowers on the female inflorescences afford apparent distinctions, but if they do 
not prove to be adequate, C. pubescens will need to be transferred to Catostigma 
instead of Catostigma being united with Catoblastus. 
The true affinities of Catoblastus, as represented by C. praemorsus, may lie 
with Acrostigma rather than with Catostigma, but the beak-like connivent stig- 
mas indicated by Klotzsch do not suggest the condition found in Acrostigma. 
Wendland and Drude both describe the albumen of the seeds of Catoblastus 
as ruminate, perhaps relying upon Karsten’s drawing, but Klotzsch says that the 
albumen of praemorsus is uniform and describes the seed as marbled with 
numerous veins, which, however, may refer to the endocarp (écorce) rather than 
to the endosperm itself. 
?Binnaea 28: 262. 1856. 
* CATOBLASTUS DRUDEI. Figure 41. 
Oatoblastus pubescens Drude in Engl. & Prantl, Pflanzenfam. 2*: 61. f. GI, 
G2. 1887, not Triartea pubescens Karst. 1856, nor Catoblastus pubescens 
Wendl. 1860. 
