200 SALIX FBAGILIS, S. RUSSELLIANA, AND S. VIRIDIS. 



common British form of S. fragilis. With Smith's drawings in 

 the British Museum a male catkin is preserved, but it does not 

 throw very much light on the subject, though I am inclined to 

 think that it may be that of viridi*. The best British specimens of 

 male viridis which I have seen, were gathered at Malvern Link by 

 Mr. K. F. Towndrow, who has also sent me from the same place 

 almost typical female viridis. In Roxburghshire, Mr. A. Brotherston 

 gets another form (albescens And.) with leaves very like those of S. 

 alba, but with capsules nearer S. fragilis. In Boswell Syme's 

 Herbarium (now in the possession of Mr. F. J. Hanbury), are speci- 

 mens of this form collected at Duddingston, near Edinburgh, by 

 Boswell Syme, and labelled " 8; alba?" 



One point in connection with the Smithian species remains to 

 be noticed. Of his fragilis Smith says, " A tall bushy-headed tree, 

 whose branches are set on obliquely, somewhat crossing each other, 

 not continued in a straight line"; while his Riisselliana is stated to 

 have branches " slender and straight, not angular at their insertion 

 like S. fragilis.*' What this exactly means I do not know. Another 

 point which Smith considers important is, that in "fragilis" 

 the branches are more brittle at their base than in u Riisselliana" 

 Andersson, following Fries, says thai fragilis has branches springing 

 nearly at a right angle, while those of alba make an angle of 85°, 

 and .those of viridis one of 60°. The latter may be the case in 

 typical viridis, but in forms nearer fragilis or nearer alba, the angle 

 may reasonably be expected to be different. 



For various reasons " RnsselUana" — the Bedford willow — is, 

 Smith states, of greater commercial value than "fragilis" Be this 

 as it may, we can now, from a botanical stand-point, discuss the 

 Smithian species only from the descriptions and drawings of their 

 essential organs. The specimens figured came, in the case of 

 "fragilis," from Cambridgeshire and Bedfordshire; and in the case 

 of " Riisselliana" from Crowe's garden; and hence there is no proof 

 that either one or the other of them was the true Bedford willow. 

 At the same time it is by no means improbable that, as in the case 

 of S. alba var. ceendea (the Huntingdon willow), where we have a 

 " strain," showing great differences in commercial value, with 

 scarcely any morphological distinctions, there may be two M strains " 



of S. fragilis. 



Of the true S. fragilis there are in Britain two forms — at least of 

 the male plant, 'in one the male catkin is rather dense-flowered, 

 with the stamens much longer than the scales ; in the other the 

 catkins are lax-flowered , withthe stamens not much longer than the 

 scales. In the first the stair ens, in the second the scales, form the 

 conspicuous feature of tbe catkin. The first seems to^ be very 

 scarce in Britain, but is the only form which I have seen in i large 

 series of continental European specimens, and in any of the con- 

 tinental figures examined by me. The second is the common 

 British form, and I have seen no specimens from outside Britain. 

 It may be distinguished as var. britannica. All the British female 

 fragilis which I have seen seem to belong to the var. britannir"- 

 The continental female frumlis differs from ours chiefly in having 



