192 I'.OTANY OF SOCOTRA. 



P. obtusa, BaJf. til. in Proc. Roy. Soc. Edin. xii. (1882), 83. Tab. LVII. 



Fruticosa scandens; foliis oblongo-obtusis ; sepalis sub fructu paulum auctis ; corolla; lobis 

 induplicato-valvatis ; stylis duobus. 



Frutex late scandens subtiliter pubescens. Folia petiolata altema oblonga obtusa stepe 

 apiculata basi subinrequilateralia 2J-3 poll, longa |-1 poll, lata margine subundulata 

 coriacea, oblique pennivenia supra glabrescentia subtus venulosa puberula; petiolo £-§ 

 poll, longo. Cymce racemoso-paniculatie axillares 3-4 poll, longte, ramis primariis I poll, 

 longis strictis, pedicellis £-£ poll, longis omnibus pubescentibus ; bracteobe miuutissimae. 

 Sepala obovata obtusa subsequalia extus pubescentia subsequalia persistentia i poll, 

 longa fructifera parum aucta subpatentia -fa poll, louga vix membranacea venulosa. 

 Corolla campanulata ad medium 5-fida calyce brevior, lobis induplicato-valvatis extus 

 villosis et glanduloso-punctulatis. Stamina fere ad basin tubi corollini late aftixa inclusa, 

 filamentis basi dilatatis glabris antberis oblongis sublongioiibus. Discus annularis. 

 Ovarium 2-loculare 4-ovulatum ; styli 2 requales a basi soluti filiformes, stigmatibus 

 capitatis. Fradus pyriformis v. ovoideus parvus glaber indebiscens \ poll, longus niger, 

 pericarpio glanduloso punctulato tenui coriaceo. Semina abortu 1 rarius 2 erecta late 

 inserta glabra ovoidea, testa mucosa ; albumen abundans mucosum ; embryo replicatus 

 cotyledonibus arete 2-lobis, radicula in f era, 



Socotra. At the west end of Kadhab plain. B.C.S. n. 355. 



Distrib. Endemic. 



A somewhat remarkable climber referred provisionally to this genus. At 

 one time I was inclined to take it as the type of a new genus. But it has so 

 many points of resemblance with species relegated to the polymorphous genus 

 Porana, that for the present I have placed it here. 



There are, however, several noteworthy points of divergence from the generic 

 character. In the calyx, which is very small in the flower, but in fruit enlarges 

 enormously, is found one of the chief characteristics of Porana. But in our 

 plants the sepals are large, and somewhat membranous in the flower, and only 

 increase to a small extent in fruit, never becoming so thoroughly scarious as in 

 typical Poranw. Again, in our plant Ave find two distinct styles, whereas in 

 Porana, although the style is often deeply cleft, the division never extends to 

 the very base. In Breweria as much variation exists, so that this is a character 

 upon which much stress need not be laid. Further, in the corolla we find the 

 aestivation of our plant is induplicato-valvate, in Porana it is plicate. Our plant 

 is of a more robust habit than most Poranw, and has obtuse, not pointed leaves, 

 nor are these cordate at the base. Although our plant presents these dif- 

 ferences, in the majority of technical characters it agrees well with the genus, 

 and I consider it more judicious to regard Porana as an incompletely defined 

 genus including our plant, than to create a new one for it. 



With the monotypic and little known Madagascar genus Bonamia, the 

 Socotran plant shows many points of resemblance, but the hairy bases of the 

 stamens, and the absence of albumen in that genus exclude our plant, 



