ROSACEAE. — SORBUS 461 
Sorbus Sargentiana Koehne, n. sp. 
Arbor 6-10-metralis trunco 10-25 cm. diam., rami 5-8 mm. diam., 
pallide olivaceo-cinerei; gemmae circiter 1.2 cm. longae, parcissime 
villosae. Stipulae herbaceae, magnae, 10-17 mm. latae, dentatae, 
saltem pro parte persistentes; folia cum petiolo (4-)5-7 cm. longo 
(18-)22-28 cm. longa, (3—)4—5-juga; rhachis exalata, pubescenti- 
tomentosa demum plus minus glabrata ac purpurascens, interstitiis 
2.5-3.2 cm. longis; foliola pari infimo supremoque minore excepto 
subaequalia v. aequalia, basi hine rotundata illinc supra basin con- 
tracta, oblongo-lanceolata, majora 8.5-13.5 cm. longa, 2.6-3.9 cm. lata, 
longe sensim acuminata acutissima, tertia v. quarta parte integra 
superne argute serrata v. serrulata dentibus utrinsecus circiter 
28-52, supra parce tenerrime villosa mox glabrata, subtus in costa 
nervisque densius ceterum parce lanata demum glabrata, costa subtus 
subochracea, nervis utrinsecus circiter 20-35 supra leviter demum 
argute impressis subtus prominentibus, reticulo supra obscuro v. 
parum impresso subtus demum pro parte prominente, chartacea v. 
demum subcoriacea; epidermis supra laevis, subtus reticulato- 
papillosa papillis brevibus v. brevissimis apice scaberrimis. In- 
florescentia verisimiliter ramum longiorem terminans, 15 cm. lata, 
8.5-10.5 em. longa, convexa, confertiflora, lanato-tomentosa albi- 
cans v. subochracea demum subglabra; flores mense Junio aperti; 
cupula villoso-tomentosa, sepala glabra; petala circiter 3.5—4 mm. 
longa, vix unguiculata, ovato-rotundata, medio supra pilis paucis 
obsita, alba, staminibus subaequilonga v. parum breviora; carpidia 3 
V. 4, raro 5, inde a placenta segregata, apice lanata. Fructus globosus, 
circiter 5-6 mm. diam., scarlatinus v. rubro-aurantiacus, sepalis erectis. 
Western Szech'uan: Pan-lan-shan, west of Kuan Hsien, wood- 
lands, alt. 2300-3200 m., June 1908 (No. 3011); Wa-wu-shan, 
Hung-ya Hsien, alt. 2600 m., September 14, 1908 (No. 887); Mupin, 
woodlands, alt. 2000-2600 m., October 1910 (No. 4207). 
A remarkable species which like S. insignis Hedlund (Pyrus insignis Hooker f.) is 
apparently one of the most beautiful of the whole genus. 
Though Sorbus insignis, of which I have seen no specimens, seems very similar 
to S. Sargentiana, it is not probable that the two are identical, because it is rare 
that China has a species in common with the Himalaya. A very large number of 
Chinese plants have been referred indeed to Himalayan species, but mostly with- 
“Sa cg rn am Pag o oa nk D m 
too, of salts Saget aa og um pe aer than has been 
generally supposed. For example, of the 27 Chinese species of Sorbus not a single 
one occurs in Japan. 
