336 WILSON EXPEDITION TO CHINA 



We have seen no specimens of this Rose from China, but it is safe to assume that 

 the Chinese plant belongs to R. Wichuraiana and not to R. Luciae Franchet & 

 Rochebrune which seems to be confined to central Japan, as Cr^pin identifies the 

 plant from Chifu with R, Luciae^ var. poteriifolia Franchet & Savatier, which be- 

 longs to R, Wichuraiana. 



3. Rosa anemoneflora Fortime apud Lindley in Jour. Hort. Soc. Lond. II. 316 

 (1847). — Herincq in Rev. Hort. ser. 3, III. 281 (1849). — Walpers, Ann. III. 845 

 (18o3). — Crepin in Bull. Soc, BoL Belg. XXll. pt. 2, 46 (1883). — Hemsley in 

 Jour. Linn. Soc. XXIII. 247 (1887). — Willmott, Gen. Rosa, I. 67, t. (1910). 



Rosa sempervir ens, P anemoniflora Regel, Tent. Ros. Monog. 83 (1877); in Act, 

 Hort. Petrop. V. 2, 367 (1878). 



Fokien: Dunn's Exped., April to June 1905 (Hongk. Herb. Nos. 2638, 2639). 



Dunn's specimens have single flowers, while the form originally described by 

 Lindley has the stamens converted into narrow petals quite distinct from the outer 

 whorl of normal petals. 



4. Rosa Bnmonii Lindley. See p. 306. 



5. Rosa Rubus L6veill6. See p. 308. 



6. Rosa glomerata Rehder & Wilson. See p. 309, 



7. Rosa Helenae Rehder & Wilson, See p. 310. 



8. Rosa filipes Rehder & Wilson. See p. 311. 



9. Rosa Gentiliana Leveill^. See p. 312. 



Rosa Gentiliana, var. australis Rehder & Wilson, n. var. 



A typo recedit foUis angustioribus plerumque oblongo-lanceolatis sensim acu- 

 minatis plus minusve falcatis, 2.5-6 cm. longis, corymbis pauci- v. plurifloris. 



Rosa Brunonis Hance in Jour. lAnn. Soc. XIII. 115 (non Wallich) (1873). 

 Rosa moschaia Dunn & Tutcher in Kew Bull. Misc, Inform, add. ser. X. 96 

 {Fl. Kwangtung <$: Hongk.) (non Miller) (1912). 



Fokien : without locality, Dunn's Exped., April to June 1905 (Herb. Bot. Gard. 

 Hongk. No. 2641). 



The narrower and smaller more or less curved leaflets and fewer-flowered corymbs 

 serve to distinguish this variety. Had we more material other differences might 

 be apparent. It is possible that the Rosa alha Loureiro (Fl. Cochin. 323 [1790]) 

 belongs here. 



10. Rosa longicuspis Bortoloni. See p. 313. 



Miss Willmott (Gen. Rosa, 1. 52 [1910]) says that Rosa Leschenaultiana Wight & 

 Arnott has been found in Yunnan by Henry, and Dunn (Jour. Linn. Soc. XXXIX. 

 486 [1911]) cites Henry No. 10693 as belonging to that species. In this herbarium 

 this number of Henry's from " Feng-chen-lin, mountain forests, south of Red River, 

 7000 ft., climber, white flowers," in our opinion does not belong to Rosa Leschenaul- 

 tiana Wight & Arnott, and is distinguished by its straight prickles, its long thread- 

 like petioiules and by its even serration with glandular-ciliolate teeth. It is 

 probably an undescribed species most closely allied to Rosa longicnsj/is Bertoloni, 

 but the material is too incomplete to determine this. Some of the flowers have 

 more than five petals and suggest a garden escape, but this, considering the remote 

 region in Yunnan where the specimen was collected, is unlikely. That a species 

 of Rose native of the Nilghiri and Pulney mountains of southern India should occur 

 in any part of China seems highly improbable. 



