8 SPICILEGIA FLOR-flE SINENSIS. 



serratis utrinque sparsim hirsutis petiolis basi dilatatis amplexan- 

 tibus, peduuculis inferue glabris apicem versus hirsutis nudis 

 uiiifloris V. apice in ramos 2-4 1-2-floros bracteis tripartitis 

 laciiiiis oblongis v. linearibus iuciso-serratis stipatos divisis, calyce 

 caduco, petalis obloiigis intcgris albis venosis efoveatis extus medio 

 apiceque hirsutis 4^ Hu. longis, stamiuibus carpella sub-£equautibus, 

 carpelHs phirimis hiieari-oblougis levibus immargmatis in rostrum 

 tenue iis paulo brevius apice circinatim unciuatum abeuntibus. 



In locis humidis m. Po-hua-shau, Chinse borealis, alt. 5-6000 

 ped., d. 19 Aug., 1875, collegit am. Dr. 0. a Moellendorff. (Herb, 

 propr. 11. 20,351. ) 



It is with rehictauce that I characterise a new species in a 

 group where so many are still very unsatisfactorily defined, but 

 the present one seems beyond question different from any yet 

 described from northern Asia. Its nearest allies I take to be 

 without doubt li. aconitifolius, Linn. ! and B. platanifulius, Linn. ! 

 neither of which occurs in the Orient, in India, or perhaps in any 

 part of the Russian empne. It differs from these by its few- 

 tlowered inflorescence, efoveolate petals, and the far more numerous 

 differently-shaped nerveless carpels. 



2. Xasturtium (Clandestinaria) bemjalense, DC. In ditionis Can- 

 tonensis agris et cultis vulgaris. M. Boissier has, I know not 

 wherefore, placed his X. niloticum, which certainly belongs here, in 

 the section Cardaminum (' Fl. Orient.' i. 178). 



3. Capparis hastigcra, Hance. Circa Hoi-hau, ins. Hai-nan, 

 1877, Bullock. The specimens have leaves for the most part 

 elliptico-lanceolate, and without the hastate form which sug- 

 gested the specific name, but they undoubtedly belong to the same 

 species. 



4. lonidium siifruticosion, (ring. Cii'ca Hoi-hau, ins. Hai-nan, 

 a. 1877, Bullock. Recorded from China by DeCandoUe, but I 

 know of no recent instance of its being gathered in the country. 



5. Silene Oldhamiana, Miq. In arena litorea juxta oj^pidum 

 Pak-hoi, ora australiori prov. Kwang-tung, m. Martio 1878, coU. 

 T. L. Bullock. Here is another example of the southward 

 extension of a plant at first detected in North-eastern Asia, and 

 which until now had not been found south of Amoy. It sends 

 down a strong perpendicular root mto the sand. The late Dr. 

 Rohrbach placed this in the genus Melandriuin, an arrangement of 

 which few botanists will approve, and he himself, indeed, .was not 

 indisposed (* Linnaea,' xxxvi., 203) to refer all the species of his 

 section Klisanthe, to which the present i)lant belongs, back to Silene. 

 I think they will very naturally range there, together with 

 Hcliospertna, next after his series Compacttc. 



6. Hypericum {Xonjsca) chincnsCy Linn. In monticulo calcareo 

 Kun-yam-ngam, secus fl. North River, prov. Cantonensis, m. Julio, 

 18G4, leg. Sampson ; ad ripas fl. Lien-chau, Oct., 1876, coll. Rev. 

 J. C. Nevin. Apparently a rare plant, but undoubtedly wild. The 

 leaves are of rather a firmer texture than in cultivated specimens, 

 which are common in gardens. 



7. CalophijUum Lwphijlliun, Linn. Circa Kieng-chau-fu, ins. 



