Oxymiii-a.] floka indica. l-iS 



ovali-oblongus. Carpella spalhulata, Btipite late compresso, sesquipoUicem longa, 

 mucronata. 



This species differs from A. Javanica in several important points. In that species 

 the ovaries are 8-10, globose, broader than and very distinct from their pedicel, and 

 terminated by a depressed, almost globose style. The abnormal inner antbei-s of 

 that species are also wanting in the A. Zeylanica, which, however, agrees in every 

 respect with Cuming's and Lobb's specimens from the Philippines. Cuming's No. 

 4i96 I, also from Luzon, is not certainly identical with our plant. We have not ex- 

 amined its flowers, but, according to a note by IVT. Planchon in Herb. Hook., the 

 anthers are different from those of the Java species. The leaves of A. Zeylanica 

 are somewhat like those of A. prinoides, St. Hil. ; but in that species they taper gra- 

 dually, and are not abruptly acuminated. The fruit is also very oblique, and the 

 dorsal margin is bent at a right angle, almost semispathulate, while in A. Zeylanica 

 it is nearly regular in outline, both margins being alike. 



20. OXYMITRA, Blume. 



Polyalthisc § Oxymitra, Bl. Ft. Jav. Anon. 



Sepala 3. Petala 6, biseriatim valvata ; exteriora multo majora, 

 elongata, plana, tenuia ; interiora ovata vel oblonga, conniventia, basi 

 interdum angustata. Stamina lineari-oblonga, apice tmncato-capitata, 

 antlierarum loculis linearibua dorsalibus discretis. Torus alte conicus 

 vel subtruncatus. Ovaria oblonga, dense strigosa ; oviila in sutura 

 ventral! pvope basin 1-3 funiculo clongato adscendentia. Stylus obo- 

 vatus, Carpella mouosperma. — Prutices scandentes, peduncnlis extra- 

 alaribtts unifloris, nervis foliorum oiliquis validis remotiusculis, nervidis 

 crebris subparallelis transversis connexis. 



This genus, one of those united by Blume under PolyaWiia, but distinguished as 

 a section under the name we have adopted, appears, so far as our limited materials 

 enable us to judge, a very natural one. The species are all very similar in habit and 

 general appearance, and are especially alike in the pecuHar nervation, the numerous 

 cross nerves between the distant principal ones being strongly marked. In the two 

 species of which we have been able to examine the ovaries, we find one ovule rising 

 from very near the base, and supported by a longish funiculus, with the nucleus pro- 

 jecting beyond the exostome ; but Elume describes and figures U. cuneiformis as two- 

 ovuled. We have, therefore, here, as well as in Phaanthus, one- and two-ovuled 

 species in the same genus ; and perhaps it wiU be found that the number varies in 

 the same species, but, from the extreme difiBculty of examining the ovaries of these 

 plants after drying, this can only be determined by investigating the recent plant. 

 The genus is in floral characters very close iu the natural series to Unona, differing 

 little from the section Dasywaschalon, except in number and position of ovules, and 

 in the presence of smaU inner petals. 



In addition to the four described below, Cuming's No. 1896, from the Philippines, 

 and 0. cuneiformis, Blume, are the only species referable to Oarymitra which have 

 come under our notice. 



1. O. latifolia (H.f. et T.) ; foliis ovalibus vel obovatia obtusis- 

 simis basi cordatis, pedunculis medio unibracteatis. 



Hab. In ins. Penang, PJiillips! — {v. s. in Herb. Ilooh?) 



Fruiex scandens. BainuK validi, elongati, cortice nigricante mguloso, pnuctis 

 minutis albis conspersi, juniores dense ferrugineo-tomentosi. Folia 6-12 poll, longa, 

 4-7 poll, lata, petiole j-poU., coriacea, rigida, superne glaberrima, nitida, subtus 

 dense pubescentia, sccus costam ncrvosque ferrugineo-tomentosa, nervulis seconda- 



U 



