The leaves of my specimens are not serrulate, are 4 stamens, and the capsule is 4-seeded, hence 

 but they are as large as those of C. serrulata, ex- as regards the stamens and capsule, it is a Strobi- 



ceeding 9 inches in length and 4 in breadth, pu- 

 bescent on both sides, but scarcely scabrous: the 

 capsules are more than twice the length of the 

 calyx, though, certainly, not thrice the length. The 

 pubescence on the leaves, as shown on the drawing, 

 is rather too conspicuous. 



1496. Cryptophragmium axillare (Nees), cymes 

 axillary, regular, glandulose, about the length of the 

 petiol: leaves broadly ovate, acute at both ends, 

 glabrous, punctulate. 



Ceylon— March, 1836. 



Further consideration leads me to suspect that this 

 is not the true C. axillare^ though agreeing in so 

 many particulars. In Nees' plant the stems are said 

 to be repent, a foot and half long: here they are 

 evidently difftise, climbing, and probably many feet 

 long. In his, the stems are 4-sided, with decurrent 



lanthes^ with the corolla of Endopogon, while the 

 calyx and stamens are those of Stenosiphonium. 

 The two nearly allied species, E. capitatus and/oiia- 

 sjiSy have the stamens and capsule of Endopogon and 

 the calyx of Stenosiphonium, 



Ought in such a case an additional genus to be 

 constructed for the reception of these 3 plants, 

 which are all so closely allied in habit as to appear 

 inseparable ; or are we rather to stretch a point and 

 admit them into one of the already existing 3 gen- 

 era ? The latter has appeared to me the preferable 

 course, hence I have referred them all to Endopogon. 



They all coincide with Endopogon in the form of 

 the corolla and number of seed, and two of them 

 further coincide in the number of stamens, but they 

 all differ in having a 5-cleft, not 5-parted, calyx: but 

 to that I attach only secondary importance. The 

 inconvenience attending this course is tiie introduc- 



angles from the leaves, here they are terete, but tion of a didynamous species into a diandrous genus, 



angled: in his, the leaves with the petiol are 3| ^. 



inches long, and from 1| to 2 broad, here they are fessor 

 about half the size; but it is a branch and then didyn 



finding 

 ntroducing a diandrous species ii 



the leaves are said to be smaller: in his, the petiols 

 equal the peduncles, here the leaves are almost 



Influenced by these views, and attaching only speci- 

 fic, or at most, sectional value to the extent of ad- 

 sessile ; in other respects it seems to quadrate with hesion between the lobes of the calyx, I submit for 



the description and comes from the same country, 

 and is the only plant in my collection at all corres- 



the consideration of Botanists, the following diag- 

 nostic characters of the three genera just named ; 



ponding with the character of the species, a speci- the adoption of which will, I apprehend, obviate the 



men of which, from my collection, Nees quotes, 



Endopogon, Stenosiphonium. 



These two genera, as they stand in Nees' Mono- 

 raph, can scarcely be viewed as distinct though, at 

 rst sight, apparently, easily distinguished by the num- 

 ber of their stamens — 2 in the former, 4 in the latter. 

 But this distinction Nees has himself broken down 

 by his St diandrum, regarding which he remarks, 

 "ambigit inter Endopogones et Stenosiphonia sed 

 calyx vix usque ad medium divisus ;" thus making 

 the essential generic distinction rest on the greater 

 or less depth of the clefts of the calyx, and not on 

 the number of stamens, nor seed in the capsule, or 

 in other words assigning generic value to a circum- 



necessity of an additional one in an order, perhaps, 

 abeady overburthened with genera, some of them 

 resting on imperfect observation of the structure on 

 which they are founded. In saying that I think 

 fewer might serve, it can scarcely be necessary to 

 guard myself against being misunderstood in the 

 opinion already expressed, that some of the existing 

 genera may require sub-division, as that does not 

 imply that all the existing ones will be found worthy 

 of preservation. 



Endopogon. Flowers diandrous, rarely didyna- 

 mous. Corolla campanulato-infundibuliform, with a 

 long, slender tube. Capsule 4-seeded. 



Strobilanthes. Flowers didynamous. Corolla 



i'"'Z^^^^J^^'^t:^A^Z'±.,r.'3lTt infuDdibaUform, tube short. Capsile 4.seeked. 



To this high valuation I demur, and therefore in 



naming the following species, left the calyx com- 



paratively out of consideration, and in lieu thereof androus. 

 made use of the number of seed in the capsule 

 combined with the form of the corolla; viz. a cam- 

 panulate I^b, and long, slender tube, which is 

 common to both. 



Stenosiphonium. Flowers didynamous, rarely di- 



infundibuliform 



Endopogon, corolla campanulato-infundibuliform, 

 capsule 4-seeded. — Stamens usually two. 



Stenosiphoniumy corolla campanulato-infundibuli- 

 form, capsule 8-seeded. — Stamens usually four. 



Thus the number of stamens and seeds in the 

 capsule divides, into two genera, a group of species 

 which the form of the corolla unites. So far all is 

 easy. But the tetrandrous Stenosipkonium has at 

 least one diandrous species, and according to my 

 view, the diandrous genus Endopogon has a tetran- 

 drous species in my E. strobilanthes. 



Here a new difficulty arises, Endopogon differs 

 ftom Strobilanthes in the number of its stamens, and 

 to some extent in the form of its coroUa, the latter 

 wanting the long slender tube, the limb being near- 

 ly the same in both. In my £. Strobilanthes there 



a long, slender tube. Capsule 8-seeded. 



The first and last differ in the number of seed, the 

 second from both in the form of the corolla. 



1497. Endopogon versicolor (R. W.), bracts 

 lanceolate, subulato-attenuate at the apex, and, like 

 the calyx, densely glanduloso-hirsute : calyx 5-cleft, 

 segments lanceolate: leaves long, petioled, broadly 

 ovate, acuminate, crenate, glabrous above, white 

 beneath. 



Neilgherries, flowering ifl March and April 



This species is very nearly allied to E. hypoUucas, 

 from which it differs in the much larger size of all 

 the parts of the inflorescence, the long, lanceolate, 

 subulate bracts, and the abundant viscid, glandular 

 clothing of the ramuli and inflorescence, which is 

 nearly wanting on my authentic specimen of that 

 species. The under surface of the leaves is per- 

 fectly white, from a compact layer of very fine 



( 19 ) 



