110 ON SOTJTH-AMEEICAN APOCTNACE^, 



3. Skytanthus Martiantjs, nob. : Neriandra Martiana, Miill. I. c. p. 62^ tab. 18 : Habsburghia comans, 



Mart. PI. Bras. Medic. In Brasilia, ad Rio San Francisco : non vidi. 

 TMs is said to resemble much the preceding species ; and it comes from near the same 

 locality. It differs only in the form and pubescence of its leaves ; the axils are f-1 in. 

 apart ; the leaves are oblong-elliptic, subacute at both ends, pubescent at first, afterwards 

 glabrous above, with close patent nerves, 1|-2| in. long, 6-7 lines broad, on petioles 

 5 lines long ; peduncle | in. long, twice trichotomously divided. 



4. Skytanthus Havanensis, nob. : Neriandra Havanensis, Miill. Linn. xxx. 401. In Havana (Sagra 



274) : non vidi. 

 A species with very slender branches, trichotomously divided, with axils 4 in. apart ; 

 leaves broadly ovate, inequilateral, and suddenly acute at the base, terminated by a 

 short, oblique, acute, often recurved acumen, glabrous, with many close, parallel, 

 straight nerves conjoined near the margin and f-1 line apart; they are 2-2f in. long, 

 l^lf in. broad, on petioles 6 lines long ; panicles trichotomously branched, shorter than 

 the leaves, bearing 12 to 18 flowers, on slender pedicels 2 to 4 times as long as the 

 calyx ; sepals broadly ovate, rounded and mucronulate at the apex, 1 J line long ; tube of 

 corolla 3j lines long, campanulate in the mouth ; segments 5f lines long, triangularly 

 inequilateral ; anthers small, ovoid, terminated by filiform appendages of nearly equal 

 length, twisted together ; ovaries short, glabrous. 



The Neriandra suberecta, A. DO., founded on the Uchites suberecta, Sw. et And. (non 

 EcUtes suberecta, Jacq.), belongs to my genus Qhariomma. It is referred by Bentham 

 and Hooker (Gen. ii. 705), to UrecUtes, here confounding the two plants figured by 

 Andrews and Jacquin under the same name, which are very distinct. 



In p. 28 I have alluded to a mistake imputed by Messrs. Bentham and Hooker to 

 Miiller, in regard to the supposed identity of Gameraria and Skytanthus ; but I find this 

 not to be Miiller's mistake, but one of the many blunders of Prof Grisebach, who in 

 his Oat. PI. Oub. n. 16, p. 170, confounds together Gameraria latifoUa and Neriandra 



Ohabiomma *. 



Most of the plants forming this group were placed, doubtfully, by De Candolle, in 

 Neriandra, a genus since abandoned ; and they certainly differ widely in several respects. 

 Neriandra (now Skytanthus, by right of priority) has a rather small hypocrateriform 

 corolla, with narrow segments; the stamens are seated below the mouth, and the 

 twisted appendages are exserted ; there is no disk. On the other hand, in Chariomma 

 the corolla is larger (often very large), with broad dolabriform segments, the stamens are 

 wholly included, being inserted' below the middle of the tube, and there is a conspicuous 

 disk. The type is the Echites suberecta of Swartz and Andrews (not of Jacquin). 



Chariomma, nob. Sepala 5, acute lanceolata, membranacea, erecta, valde imbricata, intus singulatim 

 squamulis 3 parvis acutis ad basin praedita. Corolla ssepius majnscula; tubus pro tertia parte imo 



* From x"!"" {gaudium), ojifjia {fades), from its large handsome flowers. 



