230 spicileCtIa flor.^ sinensis. 



Oct., 1866, leg. Sampson. This is certainly identical with the 

 Ceylon plant, the " capsnla rugosa" hy which Choisy sought to 

 distinguish Loureiro's H. inermis heing caused simply by the 

 impression made on the thin j)arietes by the seeds. I sent good 

 specimens of the Canton plant to Kew, but they must have been 

 lost or mislaid, from the uncertainty Mr. Bennett felt as toLom-eho's 

 name (' Journ. Linn. Soc.,' xi., 276.) 



36. Heliotrophim hrevifoHum, Wall. In collinis incultis circa 

 Amoy, Oct., 1857 ; in subulosis ad mare paeninsulae Macaiensis, 

 copiose, aiitumnoque 1861-5 ego ipse legi. A critical plant, whose 

 distribution is therefore doubtful. Dr. Anderson extends its area 

 to the whole of the tropics of the old world and Australia, and 

 unites mth it H. .strir/osum, Willd., together with H. tenue, Wall., 

 H. tenuifoUum, R. Br., H. fruticosum, Forsk., and H. parvifolium, 

 Edgw., but A. DeCandolle and Mr. Edgworth keep these species 

 distinct. (Cfr. 'Linn. Journ.,' vi., 205.) Mr. Bentham, by restrict- 

 ing H. strigosum, so far as this continent is concerned, to Western 

 Asia (El. Austral, iv., 397), appears to consider the Chinese species 

 distinct. And Boissier ('El. Orient.,' iv., 143) omits the usual 

 note of geographical area under H. strujosum, but refers to it 

 H. cordofanum , Hochst. & Steud ! and H. hicolor. Hochst. & Steud! 

 both regarded by A. DeCandolle and by Bunge ('Uber die Helio- 

 tropien d. mittelland-orient. Elor.,' 54) after examination, as good 

 species. 



37. Lettsomia Chalmersu, sp. nov. Caule volubili juniore 

 canescente, foliis ovatis acutis basi leviter cordatis supra glaberrimis 

 subtus cano-sericeis 2^-3^ poll, longis 1^-2^ poll, latis petiole 

 1-2-pollicari cymis axillaribus densis 7-10-floris pedunculo canes- 

 cente circ. 1^-pollicari fultis, bracteis oblongis obtusissimis cum 

 calyce adpresso-incanis 5 lin. longis, sex^alis ovato-oblongis obtusius- 

 culis 2 exterioribus majoribus flori adpressis in fructu accrescentibus 

 etintus rubescentibus glabris 4-6 lin. longis, corollae candidae hypo- 

 crateriformis extus adpresse sericese intus glaberrimae tubo 3-lmeali 

 limbo 9 lin. longo ad basin fere 5-partito lobis lanceolatis acutius- 

 culis rectis x^atentibus, staminibus declinatis lobis j)aulo brevioribus 

 filamentis glabris basi triangulari-dilatata dense glanduloso- 

 fimbriata, disco crasso annulari luteo, stylo glaberrimo stamina 

 parum superante, stigmate globoso-didymo, bacca globosa rubra 

 biloculari, seminibus 4 v. abortu 2 subcompresso-rotundatis nigris 

 glabris. Secus fl. West River, prov. Cantonensis, juxta Ting-ii-shan, 

 Nov., 1874, coll. Rev. J. Chalmers. (Herb, propr., n. 20203.) 



I have drawn up my character partly from the original dried 

 plant in fruit, and partly from living ones in flower, raised from 

 seed. Except for the exserted genitalia, the flowers are more like 

 those of a Jasmine than a Convolvulus. Allied to L. af/c/regata, 

 Wt., which, however, has a jDink corolla, with less elongated lobes. 

 Loureho's Argyreia acuta, which is probably near the present 

 species, is described as with lanceolate-ovate leaves, and a terminal 

 lax panicle. Choisy, who says he has seen Loiu'eiro's sj)ecimen, 

 refers to it L. (estiva, Wt., but the accuracy of this seems to be 

 questioned. Although I have followed Mr. Bentham, who in the 



