" This species differs from the rest in having stami- 

 nodes, and the lobes of the perianth deciduous, by 

 which marks alone it agrees wiih Haasia mediae 

 Blunie. Perhaps it is the type of a distinct genus.'* 

 — Nees. 



The character of this genus is to have either her- 

 maphrodite, or unisexual flowers, 2-celIcd anthers, 

 and no starainodes : ray plant has staminodes and her- 

 maphrodite or bi-sexual flowers: the staminodes 

 are large and conspicuous, flattened cordate at the 

 base, perforated with pellucid points giving them 

 quite a foliaceous appeai'ance. 



Nees describes the species as dioicous, and speaks 

 of the ovary as rudimentary in the male flowers ; such 

 apparently is not the case in the flowers I examined- 



I have another species from Ceylon so exactly cor- 

 responding in appearance, that it seems impossible 

 to distinguish the two plants, but in it the staminodes 

 are wanting, hence it is a true Haasia which the 

 Continental one is not, in as much as it has parts not 

 present in the original species. There is another 

 plant in my collection, having much the habit of this 

 genus, and wanting staminodes, but in it the anthers 

 are 4-celled, showing that though it may belong to 

 the " Tribe," it can. scai'cely belong to the genus. 



Vy 



Tetranthera. 



1832. Sassafras Parthenoxyi^n (Nees, Lauras 

 porrecta^ Roxb.), leaves somewhat triplinerved, 

 opaque: young corymbs terminal, appearing about 

 the period of the expansion of the young leaves (cor- 

 ymbulis terminalibus subanthesi foliolosis). Nees. 



Native of Sumatra, Roxburgh. 



The appearance of the figure, which is copied from 

 Roxburgh's drawing in the Calcutta Botanic Garden, 

 does not quadrate with either the above specific char- 

 acter or with Roxburgh's description. I extract the 

 following from Roxburgh's description. " Leaves al- 

 ternate, petioled, veined, permanent, oblong, entire, 

 generally acuminate, firm, both sides smooth, the upper 

 polished, the under glaucous — 3-6 inches long from 

 2-3 broad. Panicles lateral, scattered round the base 

 of the young shoots, below their tender foliage, soli- 

 tary, long peduncled, expanding, small, composed of 

 a few diverging branchlets. Flowers numerous, pedi- 

 celled, pale yellow, calyx border divided into six, 

 alternately rather smaller, oblong, obtuse, expanding 

 segments, which are somewhat hairy on the inside/' 

 The di'awing differs in showing the floriferous branch 

 fully clothed with leaves, in other respects it con-es- 

 ponds with the description. 



1833. CriJCOBAPHNE WiGHTiANA (Nccs, TetrUU' 

 thera Wightiana^ Wall.), umbels racemose. 



Ncilgherries, Courtallum, &c. 



A common rather large tree on the Neilgherries, at 

 an elevation of from about 6000 feet to the top of the 

 hiUs. 



In this genus the flowers are dioicous. The male 

 flowers usually 6-cleft, with 12 stamens, the interior 

 six glandoliferous, extrorse aud no staminodes. The 

 female ones have 6 glanduliferous staminodes. The 

 imder surface of the leaves and racemes is clothed 

 with rusty-brown pubescence. Fruit glabrous, the 

 berry half immersed in the cup-shaped truncated tube 

 of the perianth. 



This genus seems to require revision, since, as re- 

 gards the variations of floral structure, found among 

 the species now ranged under it, it appears rather 

 complex and heteromorphous. When engaged in pre- 

 paring the series of drawings for the elucidation of 

 the genera of this order, I was, under the pressm'e 

 of then existing circumstances, prevented going so 

 fully into its examination as I could have wished, 

 and have since done, otherwise I might have shown 

 this more clearly than I have done, but still I think 

 an examination of the plates appertaining to the 

 " Tribe Tetranthereaa," will tend to lead others to the 

 same conclusion. 



Compare for example the plates 1834 and 1835 

 with 1838, all of , which are referred by Professor 

 Nees to the genus Tetranthera^ and the difference 

 between the two first and the last will be at once 

 obvious. Compai'e again 1837 and 1838, which I 

 have associated as species of the same genus, and the 

 exact similarity will, I think, be equally obvious. 

 According to my views, the two sets of forms can- 

 not be associated under the same generic character, 

 otherwise than by constructing it so loosely that al- 

 most all the tribe might be admitted into the genus. 



Contrast again this grouping with No. 1837, the 

 type of a distinct genus in which the real essential 

 character rests on the compressed or lamellar form 

 of the glandiflar appendages of the six interior sta- 

 mens, as contrasted with the thicker glandular form 

 of those of the other genera. " Lepidadenia est genus 



inter Dodecadeniam et Tetrardheram versans, flore pro 

 familia eximio, diversum ab utroque taminis petaloi* 

 deis planis obtusis subsessiUbus loco glandularum 

 terga s taminum interiorum ob vallantibus, ita, ut 

 seriem quasi exhibeant petalorum, stamina sex exte- 

 riora ab interioribus separantium." This, as con- 

 trasted with the other, is to my mind too narrow 

 a basis on which to establish a good genus. 



To show this more clearly, I shall quote Nees' essen- 

 tial generic character of Tetranthera^ under which he 

 ranges a series of 44 species, many of them depart- 

 ing widely from the character. " Tetranthera^ anthers 



4-celled, cup of the fruit discoid. Three interior sta- 

 mens biglandulose at the base. Leaves veined but 

 not coarsely reticulate (Folia venosa nee admodum 

 reticulata)." In his more extended character, he adds, 

 " six gland-like staminodes attached by pairs to the 

 three interior stamens, either sessile or stipitate." 



On turning to the species ranged under this generic 

 character, we find the four represented in plates 1834- 

 35-36 and 38, not one of which, curiously enough, 

 agrees with it. Then, as if to make the confusion 

 greater, we find at the head of the character of the 

 tribe, " Staminodia nulla." These discrepancies and 

 want of precision of language, in calling the stami- 

 nal appendages at one time glands, and at another 

 staminodes when no true staminodes are present, 

 make this a most difficult group of species to study, 

 though, when properly understood, I see no reason 

 why it should be more so than any other, since they 

 are susceptible of as easy distribution into several 

 well-defined smaller groups^ or genera, according to 

 the views of the monographist. 



The normal structure of the flowers of this order 



There is as yet only one other species of the genus, is not difficult to nnder stand, as the diagrams show, 

 a native of Java, which is distinguished by having and those of this tribe, with a few exceptions, do not 

 the umbels closer together, hence "umbeUelis spicatis," essentially depart from it. The exceptions are found 



constitutes its specific distinctive character. 



in plates 1834 and 35, and a few others in which the 



( 13 ) 



D 



