BETULACEAE. — CARPINUS 437 



I of its length along its line of connection with the bract, and is therefore connected 

 with the bract only at the upper and lower ends. Sometimes the lobe is grown 

 together on its lateral margins with both the margins of the bract forming a real 

 pocket, as 1 found in several bracts of Silvestri's No. 319. However, the most dis- 

 tinct variation I have found is sliown in Wilson's No. 7587, which has ovate-lance- 

 late bracts with entire or nearly entire margins. The lobes are large, connected 

 with the bract for only about h of their length near their base, the free part being 

 nearly obovate, acute and more or less toothed on its outer edge. 1 have never 

 found a special form of the bracts and their lobes combined with other distinct 

 characters of the plant. The difference in the shape, size, pubescence and in the 

 nervation and dentation of the leaves is not sufficient to found varieties on. 



A picture of this tree will be found under No. x260 of the collection of Wilson's 

 Japanese photographs. There is only the following distinct variety from central 

 China: 



Carpinus cordata, var. chinensis Franchet. See p. 425. 



Sect. 2. EUCARPINUS Sargent, Silva N. Am. IX. 40 (1896). — Winkler in 

 Engler, PJlanzenr. IV .-61, 28 (1904). — Schneider, III. Handb. Laubholzk. II. 893 

 (1912). 



Carpinus, subgen. Eucarpinus Nakai in Tokyo Bot. Mag. XXIX. 37 (1915). 



This section differs from the first in its looser fruiting catkins, which never 

 look like those of an Ostrya, and in its ovate or ovate-triangular more or less acute 

 nutlets. A natural arrangement of the species of this section according to their 

 relationship is a difficult undertaking. After a careful examination of the ripe fruits 

 of the species from our area I am inclined to consider the pubescence or glabrous- 

 ness of the nutlets a valuable taxonomic character. They are nearly always more or 

 less provided with silky hairs on the top of the ovary and on the lobes of the peri- 

 gone, which are more or less distinct; but the pubescence 1 refer to is a very fine 

 one and is spread over the whole surface, and sometimes is only visible with the aid 

 of a strong lens. The nutlets are nearly always somewhat glandular, but these 

 small yellowish brown glands are different from the minute hairs. Only further 

 investigation can show if the nutlets present some other characters in their 

 size, their ribs, and in the development of the perigone lobes. The shape, dentation, 

 venation, the reticulation and the size of the bracts and their lobes are very vari- 

 able in the same species, and it would need statistical investigations of hundreds 

 of the bracts of the same plant and of different plants of the same species to fix 

 the limits of this variation in certain species. It is useless to make new sections or 

 groups as Nakai did when deahng only with a few species of a very limited area 

 like Korea. The male flowers may afford good characters too, but 1 have not seen 

 them of all our species. 



3. Carpinus viminea Wallich,i PI. As. Rar. II. 4, t. 106 (1831). — A. De Can- 

 dolle, Frodr. XVI. pt. 2, 127 (18G4). — Stewart, Punjab PI. 200 (1869). — Brandis, 

 Forest Fl. Brit. Ind. 492 (1874); Ind. Trees, 623 (1906). — Kurz, For. Ft. Brit. 

 Burma, II. 477 (1877). — Hooker f., Fl. Brit. Ind. V. 626 (1888). — Gamble, Man. 

 Ind. Timb. ed. 2, 684 (1902). — Collett, Fl. Siml. 476 (1902). — Winkler in Engler, 

 PJlanzenr. IV.-61. 32, fig. 12 C-D (1904); in Bot. Jahrb. L. Suppl. 493, fig. 3 

 (1914). — Schneider, III. Handb. Laubholzk. II. 894, fig. 558 b, 559 i-k (1912). 



^ According to Wallich's Catalogue the name viminea was given by Wallich in 

 1824 in his herbarium. At this time Wallich united No. 2801 with it, which in 

 1831 was named C. faginea by Lindley (see p. 442), who gave to Wallich the 

 MSS. of the diagnoses which appeared in the Plantae Asiaticae. 



