

1917] SCHNEIDER— NEW CHINESE PLANTS ' 403 



its rather narrow and long leaves is different from that of all the other Asiatic 

 species of this group, and I cannot identify it with any species mentioned by 

 Rehder in his Conspectus specierum Asiae orientalis (Sargent, PL Wils. 

 2:272. 191 5), nor with any other form known to me. 



The name is given in honor of Count 1st van Ambrozy, a very successful 

 garden maker on his famous estate at Malonya, Hungary, as a slight return 

 for all his help in my dendrological studies. 



perta 



St Sorbus hupehensis Schn. , var . aperta, n. var. — S. < 

 in Sargent, PI. Wils. 1:465. 1913. — A typo praecipue recedit foliis 

 U-)5, non 6-8-jugis, foliolorum paribus in rhachide interstitiis pie- 

 risque 1 .8-2 .3 cm. longis separatis. 



See my remarks under the following variety. 



qfi Sorbus hupehensis, var. obtusa, n. var. — A typo praecipue 

 recedit foliis 4-5-jugis, foliolorum paribus in rhachide interstitiis 

 1.5-2.5 cm. longis separatis, foliolis apice distincte obtusis margine 



tantum triente superiori dentibus utrinsecus 3-9 serratis maximis 

 lateralium ad 5.5:2.2 cm. magnis subtus sub microscopio undique^ 

 satis dense papillosis. 



Yunnan boreali-occidentalis: prope Yung-ning, 19 Junii 1914, C. Schneider 

 (no. 1166; typus in Herb. Arb. Arn. et Hb. Schneider; arbor circ. 8 m. alta). 



In determining the Sorbus of the Auc upari a-groxxp collected by myself 

 in southern Szechuan and northwestern Yunnan, I cannot refer the above 

 form to any species or variety enumerated by KOEHNE in his Sorborum chitten- 

 slum conspectus analyticus (Sargent, PL Wils. 1:475. 1913)- ll seems to 

 me most nearly related to S. aperta Koeh., from the type of which it differs 

 by its 5-6 (instead of 4-5) pairs of leaflets which are distinctly obtuse at their 

 apex and also distinctly papillose beneath. As in S. aperta, the pairs of leaflets 

 are more distant on the rhachis, and the leaflets are somewhat larger than in 

 typical S. hupehensis. Otherwise, var. obtusa seems to connect the latter with 

 5. aperta, and I am unable to detect sufficient differences to keep S. aperta a 

 distinct species. I make it, therefore, a variety of S. hupehensis, of which it 

 represents the most northern form, chiefly distinguished by its fewer pairs of 

 leaflets. 



To the typical 5. hupehensis Schn. (in Bull. Herb. Boiss. II. 6:316. 1906; 



fig- 374* 



ifi 



(see later) if it is possible to keep this form even as a variety. 



Szechuan australis: inter pagos Wo-lo-ho et Hun-ka, in silvis apertis 

 montium, alt. circ. 3000-3400 m., 13 Junii 1914, C. Schneider (no. 3532; arbor 

 circ. 10 m. alta, trunco circ. 0.6 m. crasso; flores odore valde ingrato). 





