T. ^ PLANTS WRIGHTIAN^. ■ 57 



ly thickened, and the apex remains blunt after the falling of the style. — These 

 species, with H. trifoliata, Cav. (Port Desire, Middleton .'), differing in its foliage, 

 also H. gracilis. Hook. ^ Am. (Chili, Cuming, No. 880 !) and H. prostrata, Lag. 

 (Lima, Cuming, No. 1069 !), in both of which the petals are scarcely glandular, and 

 the claw of the vexillum long and narrow, would form the section Hoffmanseggia 

 proper. H. viscosa, IZooA'. (^ J.?-??. (W. Colombia, Cuming, No. 1175 ! and Peyta, 

 Gaudichaud), may also be referred to the same section, although the claws of the 

 petals are shorter and the pod less marginate and more readily dehiscent. The H. 

 glandulosa, Vog., from South America, which I have not seen, must be very near H. 



viscosa, but with a very long glandular raceme If we consider the section 



Pomaria as characterized by the nearly sessile petals, only differing from each oth- 

 er in the vexillum being rather broader, all with few or no stipitate glands (where 

 they exist they are generally dorsal on the vexillum), we may subdivide it into 

 three or four distinct groups. The first (§ Gladiatce), including H. gladiata * and 

 H. platycarpa, f Herh. Coll. Trin. Duhl., with the habit and foliage of H. stricta, 

 without black dots on the leaves, has the flowers nearly as in H. Drummondii, and 

 a straight or slightly curved pod, blunt at the apex, with the upper or seminal su- 

 ture more or less convex, and usually broadest below the middle. The second 

 group (§ Lunatce) would include H. oxycarpa, Drummondii, and caudata, with a 

 variable habit, the leaflets without black dots, and the pod broadly falcate, very sharp 

 by the persistent base of the style, more readily dehiscent than in any of the pre- 

 ceding, with the seminal suture straight or concave. The third group (§ Mela- 

 iiostictm), with the under side of the leaves sprinkled with black glandular dots, and 

 the pod nearly that of the Lunatce, is less known to me, as I have not myself exam- 

 ined any flowers. It would comprehend the H. melanosticta and H. Jamesii, if the 

 two are not one species. Pomaria humilis, Mart. Sf Gal, is probably allied to H. 

 platycarpa, and P. glandulosa, Cav., to H. melanosticta ; but both are unknown to 

 me. Of De Candolle's genus Melanosticta I have seen but a very imperfect speci- 

 men of M. Burchellii, with a pod shaped like that of a Csesalpinia, and still worse 

 specimens without flower or fruit of what appears to be a second African species. 

 Until they can be examined from better specimens, it cannot be determined whether 

 they are or are not generically distinct from Hoffmanseggia," Benth. in lift. — My 

 H. brachycarpa, which, having just reached me from the discoverer, has not yet been 

 seen by Mr. Bentham, would seem to be ambiguous between the sections Gladiata 

 and Melanostictce, approaching the former in the shape of the pod, but belonging 

 rather to the latter on account of the persistent style and the black glands on the 

 under surface of the leaves. I have still to add another species, entirely glandless, 

 which, with the sessile petals of a Pomaria, has the thin, tardily dehiscent and many- 

 seeded legume of Hoffmanseggia proper, viz. : — 



* " H. GLADIATA : stipulis ovato-acutis ; pinnis 3 - 6-jugis cum impari ; foliolis oblongis enervibus eglan- 

 dulosis ; calycibus acutis hirsutis gland ulosisque ; petalis oblongis brevissime stipitatis ; legumine lanceolate 

 subincurvo hirtello et parce glanduloso. (Zimapan, Mexico, Coulter.) " Benth. in litt. 



t " H. platycarpa: stipulis ovato-acutis; pinnis 4 -6-jugis cum impari; foliolis oblongis enervibus 

 eglandulosis ; calycibus acutis hirtis glandulosisque ; petalis ovatis vix stipitatis ; legumine late oblongo pu- 

 bescente vix glanduloso. (Mexico, Coulter.y Benth. in litt. 



