Salix l 7^3 



outside, the wood has been found so fresh at the end of that period as to be fit for 

 boat-building." Boards of willow were laid for floors in 1700. 1 (H. J. E.) 



SALIX CCERULEA, Cricket-Bat Willow 



Salix ccerulea, Smith, Eng. Bot. t. 2431 (1812), and in Rees, Cycl. xxxi. 141 (1819); Aiton, Hort. 



Kew. v. 365 (181 3) (not Forbes). 

 Salix alba, var. ccerulea, Smith, Eng. Flora, iv. 231, 232 (1828); Loudon, Arb. et Frut. Brit. iii. 



1523 (1838); Bean, in Kew Bulletin, 1907, p. 312, plate, and 1912, p. 205. 

 Salix viridis, Pratt, 2 in Journ. Roy. Agric. Soc. lxvi. 22 (1905), and Quart. Journ. Forestry, i. 325, 



fig. No. 02 (1707) (not Fries). 



A tree, attaining 100 ft. in height, with ascending branches, making a narrow 

 angle with the stem, and forming a pyramidal crown. 3 Terminal branchlets erect, 

 not spreading or drooping. Bark smoother than in 5. alba. Young branchlets 

 appressed pubescent, becoming reddish brown in winter and the following year. 

 Leaves similar to those of 5. alba, but thinner in texture, more translucent, and less 

 densely pubescent ; lower surface not white, but bluish grey ; margin ciliate, with 

 minute glandular serrations. 



Pistillate catkins, differing from those of S. alba as follows : ovary 4 slightly 

 stalked, more tapering at the apex, about \ in. in length ; each of the two style- 

 arms distinctly bilobed ; scale more concave, about two-thirds the length of the 

 ovary. Fruiting capsule, nearly \ in. long, on a distinct short pedicel. 



This remarkable tree, which is best distinguished by its pyramidal habit, with 

 stiff ascending branches and branchlets, is par excellence the true cricket-bat willow, 

 as it exceeds the other kinds in rapidity of growth. Bean calls it the "best close- 

 bark willow." In addition to its different habit, the leaves are readily distinguishable 

 from those of S. alba by their different colour, and also by their translucency, as when 

 viewed against the light with a lens, the tertiary venation is always plainly visible. 

 In the forms 6 of S. viridis, which approach S. alba in foliage, the leaves simulate 

 those of S. ccerulea in colour and translucency, but are considerably larger and are 

 also more coarsely serrate in margin ; and no form of S. viridis, known to me, has 

 either the peculiar habit or rapid growth of 5". ccerulea. 



S. ccerulea is considered by most botanists to be a variety of 5". alba ; but it 

 differs from the latter in the flowers, which approximate in their size and shape and 

 stalked ovaries to those of S. fragilis; and it is possible that S. ccerulea may be the 

 first cross between these two species, most of the characters of 5". alba, if this hypo- 

 thesis is correct, being dominant. 



S. ccerulea was first distinguished by Smith, who was unable to find any botanical 



1 Ellis, White Woods, quoted by Mitchell, Dendrologia, 56 (1827). 



1 Mr. E. R. Pratt at first accepted, on Mr. Linton's identification of certain specimens, the name S. viridis for the true 

 cricket-bat willow, though in Journ. Roy. Agric. Soc. lxvi. 22 (1905), he agrees that the East Anglia cricket-bat willow has 

 leaves indistinguishable from those of 5. ccerulea. 



' Owing to the fastness of growth of S. ccerulea, the nodes are at greater intervals, and the crown of foliage in consequence 

 is remarkably sparse. The bark is darker in colour than that of S. alba. 



4 Bean, Kew Bull. 1907, p. 313, states that the ovaries of 5. ccerulea are identical with those of S. alba. A careful 

 examination shows that the pyramidal tree with bluish foliage has always the distinct flowers described above. 



6 These forms of S. viridis are often referred to S. alba, var. carulea, especially on the Continent ; but I restrict the name 

 S. ccerulea to the pistillate pyramidal tree here described. 



