BETULACEAE. — BETULA 459 



more slender than in B. utilis var. Prattii Burkill. The size, shape and the wings 

 of the seeds are so variable in all these forms that it seems impossible to place any 

 dependence upon these characters. 



This Birch is common in the mixed forests on the hifi;h mountains of extreme 

 western Szech'uan between altitudes of 3000 and 3800 m., where it is a tree often 

 33 m. tall, with thick branches and a trunk from 4 to 5 m. in girth. The bark is 

 orange-brown or orange to j-ellowish orange or orange-gray and exfoliates in thin 

 sheets more or less persistent on the trunk and main branches. The bark is singu- 

 larly beautiful and makes the tree conspicuous in the forest. Pictures of this Birch 

 will be found under Nos. 173, 174 of the collection of my photographs and also 

 in my Vegetation of Western China, No. 134. E. II. W. 



Betula insignis Franchet in Jour, de Bot. XIII. 206 (1899). — Burkill 

 in Jour. Linn. Soc. XXVI. 498 (1899). — Diels in Bot. Jahrh. XXIX. 

 281 (1900). — Winkler in Engler, Pflanzenr. IV.-61, 59, fig. 19 A-C 

 (1904). 



Western Hupeh: Changyang Hsien, mountains, June 1901 

 (Veitch Exped. No. 1130; unripe fruits); same locality, August 1900 

 (Veitch Exped. No. 1130^; ripe fruits). Eastern Szech'uan: Tchen- 

 keou-tin, alt. 1400 m., P. Farges (No, 83, type, ex Franchet; fruits). 

 Western Szech'uan: southeast of Tachien-lu, 3000-3G00 m., 

 woodlands, June 1908 (No. 1427; tree 10-13 m. tall, girth 1.2-2.4 m.; 

 fruits); west of Kuan Hsien, ascent of Pan-lan-shan, alt. 2300-3700 

 m., thickets, June 1908 (No. 3365; thin tree 8-9 m. tall; fruits). 



1 have not seen the type of this species, but according to Franchet's description 

 and Winkler's figure it differs from Wilson's plants in its longer strobiles up to G cm. 

 long and 1.5 cm. thick, the strobiles of the form from Hupeh being about 3 cm. 

 long and 2-2.2 cm. thick. Those of No. 1427 from western Szech'uan are about 

 4 cm. long and 2.2 cm. thick, while in No. 3365 they are only 2-3 cm. long and 1.5 

 cm. thick. The last two numbers w^ere collected in June, and bear only the stroVJles 

 of the previous year. The lateral lobes of the bracts of No. 3365 are a little more 

 spreading than in the type and in the other numbers, and the serration of the leaves 

 is also a little different. Without more material it is difhcult to say whether these 

 numbers represent different varieties or not. The leaves of No. 1427 are up to 

 13.5 cm. long and 6.5 cm. broad. As a wliole B. insignis is a most distinct s]»ecies. 



In 1912 I was inclined to refer a specimen of Wilson's (Veitch Exped. No. 2799) 

 from Hujieh to what I called B. chincnsis, var. Deluvayi. Mr. Wilson tells me this is 

 a tree from 6 to 8 m. tall, and I now think the specimen, which was poor and im])er- 

 fect, should be referred to B. insignis Franchet. Unfortunately the number is not 

 represented in the herbarium of the Arnold Arboretum. 



In western Hupeh and in western Szech'uan this s])ecies is rare and I .«aw only a 

 few trees. It is a low tree with a bushj' head and smooth, firm dark bark. 1 suspect 

 that it has an aromatic bark like B. lenta Linnaeus, but 1 omitted to test it in the 

 field, and this character cannot be detected with certainty from dried material. 



E. H. W. 



Betula Potaninii Batalin in Act. Ilort. Petrop. XIII. 101 (1893). — 

 Burkill in Jour. Linn. Soc. XXVI. 498 (1899). 



