378 BRITISH FLOWERING PLANTS chap. 



that each whorl occupies the whole circiimference. For 

 instance, *S'. herhacea has a whorl of 3 nearly circular 

 leaves ; S. Caprea, with broad leaves, 5 in a whorl ; S. 

 pentandra, with lanceolate foliage, and 8, and S. incana, 

 with 13 linear leaves arranged in 5 spirals. In several 

 species the ascent of ants and other unwelcome visitors 

 is prevented by a waxlike, smooth, and slippery secretion 

 which they cannot scale. The seeds have a tuft of long 

 white silky hairs which enable the wind or water to carry 

 them about, and when at length they fall on damp earth 

 serve to fasten them to it. The nectary bears in many 

 cases a single large water -gland at the end, through 

 which the nectar exudes. 



I have described the buds of the Willow and Poplar 

 in my Buds and Stipides. It is curious that in some 

 species the young leaves are convolute, while in others 

 they are " equitant." The more the genus is studied 

 the more difficult, I might say the more impossible, does 

 it become to divide it into well-marked species. The 

 difficulty is increased by the facts that the male and 

 female flowers are on difl"erent trees, which are often 

 dissimilar, and that the leaves and flowers are out at 

 difli'erent seasons of the year, and also by the readiness 

 with which the species hybridise. 



S. viminalis. — The stomata on the under side of the 

 leaf are protected by silky hairs. The young stem is 

 hairy, afterwards glabrous. 



S. Caprea. — In this case the stomata are protected 

 by short crisp, but not silky, down. The leaves are 

 glabrous and shining above, tomentose below. The 

 stems are at first tomentose, afterwards glabrous. 



S. pentandra. — In this species stipules are sometimes 

 present, sometimes absent. 



S. herbacea. — The leaves of this species, which, like 

 the two following, is an Alpine and Arctic form, are 

 typical "snow-leaves" (see Viola palustris, p. 28)." 



S. reticulata. — This species also has rounded leaves, 

 but the edges are entire. It is possible that this difi'er- 

 ence is dae to the prevalence of rain rather than of 



