ARRANGEMENTS FOR RETAINING THE POSITION ASSUMED. 429 



green bands. At fii-st sight these two gi-een bands appear to bo Hat, Imt thoy are 

 not really so; each is convolute, and thus in the crocus-leaf there are actuallj' two 

 green tubes united to the white central stripe, which is destitute of ciilorophyll. 

 This leaf may be distinguished by its erect position from the rolled leaves which 

 have been described in detail, and which are similar in some respects, although they 

 dirt'er from them in significance. 



The spiral leaf furnishes another contrivance of this kind. It i.s frequently 

 seen in the leaves of bulbous plants, bur-reeds, and grasses, principally in young 

 plants — as, for example, the first green leaves of barley and rye. Leaves spii'ally 

 twisted like this ai-e alwaj's long, narrow, and erect, yometime.s but a single spiral 

 revolution is found, or even only a half revolution; sometimes two, three, often even 

 four circuits are described. The leaves of the New Zealand Flax (Phormium 

 tenax), and those of the Asphodel (Asphodelus albus), of narcissus, of iiiaii\' irises, 

 and of some pines, exhibit only a half, or at most only a single spiral twist: those 

 of the Lesser Bulrush (TyjAa anrjusti folia), and numerous species of Garlic {Allium 

 senescens, rotundum, obliquum) present two to three, those of Sternbergia Clusiana 

 three to four, and the Persian Sternbergia stipitata five to six revolutions. Leaves 

 of this kind have, consequently, a curled appearance. That such a spiral leaf 

 resembles the fistular leaves in its mechanical significance, and that it po.ssesses a 

 greater resistance to flexion than a flat leaf, is beyond question. 



It may also be noticed in the Reed-mace, that in a strong wind the leaves are 

 not only bent, but are also somewhat elongated, i.e. the spiral becomes somewhat 

 looser in the bent leaf. But as soon as the wind subsides, and the leaf returns 

 to its vei'tical position, the previous foi-m of torsion is resumed. The advantage 

 possessed by an upright spirally-twisted leaf over an erect flat one, with regard 

 to wind, becomes quite obvious when one imagines the two forms exposed side 

 by side to the same wind-pressure. When the gust strikes an erect flat and rigid 

 leaf, the whole of its surface is encountered at right angles, and the leaf undergoes 

 a large amount of bending, and possibly fracture; but when it strikes a spirally- 

 twisted erect leaf, the various portions of the blade are met at difl'erent angles; the 

 air current becomes, as it wei-e, diffused into innumerable streams, wliich, passing 

 along the revolutions of the spiral, effect only a comparatively small curvature, and 

 scarcely ever cause the leaf to be broken. When these spiral leaves are swayed by 

 the wind, from a distance the movement has a very peculiar look, much more like 

 trembling, tossing, and twisting than like bending. 



The arched form of leaf is closely allied to the spiral. It, too, is found in long 

 ribbon-shaped leaves. At the connuencement of development the arched leaf is 

 erect and lies in one plane, but when fully develoj^ed it takes the form of a bow, 

 with the convex side directed upwards. It may spring frcmi the sides of erect 

 lofty stems, or may originate close to the soil. Arched leaves appear very 

 noticeably in those grasses whose habitat is on the ground in, and at the margins 

 of, woods and on steep mountain slopes, e.g. in Milium effusum, Melica aldssivia, 

 Calamagrostis Halleriana. Brachypodium, silvaticurn, A vena fiavescens, and 



