THYMELAEACEAE. — DAPHNE 549 



rect, particularly as all the other species with the stamens inserted below the 

 middle have 5-merous flowers. 



22. Daphne rosmarinifolia Rehder, n. sp. 



Frutex 0.5-1.2 m. altus; ramuli homotini et annotini subangulati, dense strigoso- 

 pilosi, anno tertio glabrescentes brunnei. Folia persistentia chartacea, alterna, 

 approximata, lineari-oblonga, apice rotundata leviter mucronulata, basi cuneata, 

 margine revoluta, 1-2 cm. longa et 0.2-0.4 cm. lata, supra obscure viridia, subtus 

 pallidiora, utrinque glabra, margine initio pilis longis ciliata, apice villosa, costa 

 media subtus leviter elevata nervis secundariis obsoletis; petioli 1 mm. longi, mar- 

 gmati, glabri. Flores lutei, sessiles, folia paullo v. vix superantes, in spicis pauci- 

 floris sessilibus terminalibus bracteis deciduis spathulato-oblongis pubcscentibua 

 4-6 mm. longis suffultis; perianthii tubus supra basin leviter ventricosus, glaber, 

 10-11 mm. longus, lobis 5 ovatis v. ovato-oblongis patentibus, 3.5-4.5 mm. longis; 

 antherae 10, oblongae, 1.75 mm. longae, fere sessiles, biseriatae, infra medium tubi 

 affixae, filamentis brevissimis; discus fere annularis v. bipartitus irregulariter 

 smuato-lobatus, vix 1 mm. longus; pistillum 3.5 mm. longum, ovario ovoideo, 

 oblongo basin versus attenuato glabro, stjde 0.5 mm. longo, stigmate capitato 

 parvo. Fructus desideratur. 



Western Szech'uan: Tachien-lu, May 1904 (Veitch Exped. No. 4434). 



This is a well marked species easily distinguished from the other species which 

 have the stamens inserted below the middle of the tube by the narrow glabrous 

 leaves and by the glabrous bracteate flowers. Like D. ■penicillaia Rehder this has 

 a comparatively long style, though not as long as that of D. myrtilloides Nitache 

 and that of the following species. 



23. Daphne tenuiflora Bureau & Franchet in Jour, de Bot V. 151 (1891). — 



Keissler in Bof. Jakrh, XXV. 104 (1898). — Nitsche, Beitr. Kenntn. Daphne, 29 



(1907). 



Western Szech'uan: Tacliien-lu, June 21, 1890, Bonvalot (ex Bureau & 



Franchet). 



After having examined most of the Chinese species of Daphne and Wiksfroemia 

 which together number at present 46 species, I am inclined to doubt whether these 

 two genera are sufficiently distinct. The only obvious difference between them is 

 m the development and shape of the disk-hke excrescences at the base of the ovary, 

 or " receptacular eflSgurations " (Receptaculareflfigurationen) as they are named 

 by Gilg, which I shall call here disk or disk-scales. In Wikstroemia they are repre- 

 sented by 1-4 narrow and entire or sometimes broader and variously toothed or 

 incised scales, while in Daphne they are entirely wanting or represented either by a 

 minute ring-Kke or cupular disk or by a broad unilateral entire, truncate or rounded 

 scale. These broad unilateral scales seem to occur only in the species with 5- 

 merous flowers, and as there is sometimes no great difference between these entire 

 scales and the more or less lobed scales of Wikstroemia, I was almost inclined, even 

 ^ 1 kept the other species in Daphne, to refer the 5-merous species to Wikstroemia, 

 " they had not shown an apparently close affinity to the species of the sections Dapk- 

 nanthes and Daphnanthoides, and if in some cases as in D. rosmarinifolia Rehder and 

 jn D. tenuiflora Bureau & Franchet they had not had floral bracts which are en- 

 tirely absent in Wikstroemia. Between species of Wikstroemia like W. ligustrina 

 ™hder, W. Pampaninii Rehder and W. canescens Meisner with broad often not 

 ^ety deeply dentate scales and those of Daphne like D. gemmata Pritzel, D. modesta 

 Rehder and D. rosmarinifolia Rehder with broad sometimes slightly emarginate 

 or minutely dentate scales there is no conspicuous morphological difference in the 

 shape of these scales, particularly if one considers their great variability within 



