FOUK NEW GENERA OF AROIDE^. 197 



The material upon which I have founded this genus is so much 

 eaten by insects that I should have hesitated to publish it had it 

 not been that the habit, the structure of the ovary, and the ortho- 

 tropous ovules plainly show that it can belong to no genus at 

 present described, its nearest ally being Stmirostif/ma. 



There is no leaf with Burchell's specimen in the Kew Herba- 

 rium, but in his MSS. catalogue No. Bill is thus described, 

 " Spadix terrestris fere acaulis. Spatha alba, ante folia. Folia 

 pedata. In locis de^^ressis campestribus fluviis saepe inundatis." 

 Among the Aroids of Burchell's herbarium are four leaves (No. 

 8598) collected near Porto Eeal, which are pedate ; and as this 

 locality is also in the province of Goyaz, and the situation in which 

 they were found similar ("In humidis depressis. Petioli et folia 

 immaculata," Burchell's MSS.), I am inclined to think that, if 

 they do not belong to the same species as the flowering specimen 

 here described, they i^robably belong to the genus ; I therefore give 

 a short description of them : — Petiole about 6 inches long. Lamina 

 pedate, middle leaflet sessile, lateral branches with 4-5 sessile 

 narrowly decurrent leaflets along their upper edge, and about the 

 same number of short spur-like or falcate acute expansions of the 

 lamma along their lower edge ; leaflets narrow oblanceolate acumi- 

 nate, the middle one 4-5 inches long, ^— f inch broad, the others 

 gradually decreasing in size. These leaves are unlike those of any 

 other South American Aroid known to me, but of those hitherto 

 described they most nearly resemble the leaf of Chlorospatha Kolbii, 

 Engl. (Gartenfl. t. 933), but that plant produces flowers and leaf 

 together, and there are no laminal expansions on the lower edge of 

 the lateral branches of the leaf. 



Explanation of the Plates. 



Tab. 230. — Rhektophyllum mip.abile. — Fig 1. Leaf, one-fifth natural size. 

 2. Inflorescence, natural size. 3. Spadix, natural size. 4 & 5. Dorsal and side 

 views of one of the lowest ovaries. 6. Transverse section of the same. 7. Side 

 view of another ovary higher up the spadix, 8. Transverse section of the same. 



9. Longitudinal section of an ovary, radial to the axis. 10. Ditto, ditto, tangential 

 to the axis. 11 & 12. Two groups of anthers, seen from above. 13. Transverse 

 section through the top of an anther. 14. Front view of an anther (Figs. 4 to 10 

 magnified six diameters ; 11 & 12 magnified two diameters ; 13 & 14 magnified 

 four diameters.) 



Tab. 231. — I. Geaeum brasiliense, inflorescence natural size. Fig. 2. Male 

 flower, seen from above. 3. Ditto, side view. 4. Ditto, transverse section. 

 5. Ovary and neuter organ. G. Longitudinal section of an ovary. 7 & 8. Trans- 

 verse sections of two ovaries. U. Staminode (Figs. 2 to 9 variously magnified.) — 

 11. PsEUDODRACONTiUM ANOMALUM. Fig. 1. Spathe and spadix. 2. Spadix 

 with the spathe removed. 3. Side view of an ovary. ■!. Dorsal view of an ovary. 

 5. Longitudinal section of ovary radial to the axis. 6. Transverse section of 

 ovary. 7. Ovule. 8. Group of stamens, seen from above. 9. Ditto, side view. 



10. Stamen, inner face. 11. Stamen, outer face. 12. Transverse section of 

 anther, (Figs. 1 & 2 natural size ; 3 to 12 variously magnified.) 



