SALIX 
25 
Smith’s. Ehrhart describes the ovaries of his 6". undulata as hairy ; but those of his own specimen are glabrous. Further, 
the description of S. alba x triandra by Wimmer {loc. cit .) also disagrees with Smith’s plant which cannot be said to have 
lax catkins and oblong-lanceolate laminae ; and the leaf-measurements given by Wimmer are also inconsistent with the view 
that he was describing Smith’s plant. There is no evidence to show that Wimmer ever saw an authentic specimen of S'. lanceolata 
Smith ; and it is clear that he never saw Smith’s figure, for this is cited as “ ex Hooker FI. Scot.” For all these reasons, 
we cannot accept the prevailing view that 6". undulata Ehrhart and .S', lanceolata are identical, apart from the matter of the 
pubescence or glabrousness of the capsules. 
Some modern authorities (e.g., v. Seemen in Ascherson und Graebner op. cit.) follow Wimmer in his treatment of S. 
lanceolata Smith ; but we think we have made it quite clear that, regarding this particular willow, Wimmer was not in posses- 
sion of first-hand knowledge. 
Owing to the confusion which prevails, some doubt attaches to many records of x S. lanceolata. 
Stream-sides, alluvial meadows and woods, and osier-beds, chiefly in the eastern and midland counties, from 
Surrey and Essex to Shropshire and the North Riding of Yorkshire; Perthshire (planted). 
Western and central Europe and Russia. 
x S. mollissima (= S. mollissima Ehrhart Beitr. vi, 101 (1791)) is another form of S. triandra x viminalis, nearer to 
S', viminalis , with rather larger leaves more hairy underneath, which does not appear to have been definitely recorded for 
this country : it possibly occurs here, however, x S. trevirani , which is sometimes separated as a special hybrid-form we 
include within the limits of x S. hippophaefolia , as well as some plants named x S. undulata Ehrhart. 
S. triandra x viminalis is recorded for southern Scandinavia and Denmark (doubtfully indigenous), Holland, 
Belgium, Germany, France, central Europe, western and central Russia. 
Section II. CHAMAETIA 
Chamaetia Du Mortier in Bijdr. Natuui'k. Wetensch. (15) (1825); Chamelyx Fries FI. Snec. Mant. i, 72 
(1832); Babington in Journ. Bot. i, 172 (1863) excluding Myrsinites ; Glaciales Koch Sal. Comment. 61 (1828). 
For characters, see page 13. 
British series of Chamaetia 
Series iv. Reticulatae (see below). Dwarf undershrubs of Arctic-Alpine distribution. Aerial 
branches prostrate to suberect. Laminae suborbicular, entire or subentire, strongly reticulated 
underneath, silky with hairs when young, usually glabrous and subglaucous when mature. Catkins 
on long leafless peduncles, narrow, cylindrical. Bracts greenish towards the base, reddish at 
the margin or towards the summit. Nectaries 2 — 4, free or united at the base and surrounding 
the base of the stamens or gynophore, with several (often 4) narrow erect dark green segments. 
Style short. Stigmas short, stout, reddish. Capsules sessile, broadly oval, covered with white hairs. 
Series v. Herbaceae (p. 27). Dwarf undershrubs of Arctic-Alpine distribution. Aerial 
branches short, a little ascending. Laminae broadly elliptical to suborbicular, smooth, thin, crenate, 
glabrous, flexible, markedly reticulate. Catkins on short leafless peduncles with 1 — 2 leaves at the 
base. Bracts concolorous, yellowish, rounded at the apex. Nectaries usually 2, sometimes more or less 
united at the base and surrounding the base of the stamens or gynophore, with two broad or narrow 
lobes. Style short. Stigmas divided. Capsules shortly stalked, narrowly conical, glabrous, often reddish. 
Series iv. Reticulatae 
Reticulatae [Borrer in Hooker Brit. FI. 422 (1830) nomen] v. Seemen in Ascherson und Graebner 
Syn. iv, 67 (1908); Chamitea A. Kerner in Verhandl. Z.-B. Gesellsch. Wien 275 (i860) as a genus; Chamiteae 
A. et G. Camus Classif. Sard. 129 (1904) as a section. 
For characters, see above. 
A. reticulata possesses so many remarkable characters, showing it to be, in spite of the great difference in habit, 
intermediate in several respects between Populus and species of Salix in general, that there is little wonder that Kerner (loc. cit.) 
suggested it should be placed in a new genus. However, the remarkable characters possessed by S. reticulata are so 
distributed among the other more primitive species of Salix that its generic separation from them cannot be maintained ; 
and indeed Kerner himself at a later date accepted this view. The characters by which 6". reticulata recalls Populus are 
the suckering habit, the long petioles, the broad laminae, and the perianthoid nature of the nectary. In its androecium, 
however, it has become a thorough Salix, more so even than S. pe?itandra, which has rather broad laminae, a double 
nectary, and, as a rule, 5 stamens at least. It seems to us that S. pentandra and S. reticulata diverged long ago from a 
primitive Salicalian stock, that each has retained a few of the Populus- like characters which this ancestral hypothetical 
group possessed, and that each of these species or their ancient allies have given rise to the other species of Salix, some 
of which (e.g., S. lanata and S. daphnoides, and A 1 , lapponum and A”, viminalis respectively) exhibit interesting features of 
convergent development. 
Sir J. E. Smith (Eng. FI. iv, p. 201) shrewdly remarked, so long ago as 1828, that “the spreading woody roots [of 
.S', reticulata ], dwarf stems, round veiny leaves, and terminal and long-stalked catkins, coming after the foliage, from the 
same bud and unattended by floral leaves, accord singularly with S. herbacea, to which the plant before us, however 
widely and essentially distinct as a species, is evidently akin.” On these grounds, we regard it as thoroughly justifiable to 
place the two series Reticulatae and Herbaceae in the same section. 
M. II. 
4 
