November 1, 1897.] 



KNOWLEDGE. 



255 



Bombax, EriodenJron, etc., have the same peculiarity. 

 More frequently, however, the hairs are confined to one 

 end of the seed and form a tuft, as in the willow and 

 poplar, where they grow from the stalk at the base of the 

 seed and constitute an arillus. Asclepia and Epilohium 

 have an arillode ; the hairs spring from the summit of the 

 seed where the micropyle or little opening ia situated. 

 Hibiscus has seeds tufted at both ends. In species of 

 Anemone, Clfinatis, and Gium, and in the plume nutmeg 

 (^Atlierosperma), hairs are developed from the wall of the 

 fruit ; they arise from the flower-stalk in the cotton-grass 

 (Eriuphorum), so abundant on all our moors. The 

 pappus or hairs attached to the seed-like achenes of 

 Composita and Valerianacca probably belong to the calyx. 

 Pappus exhibits considerable variety : in the thistle and 

 hawk-weed it is feathery ; in the hawk's-beard and sow- 

 thistle the hairs are simple and undivided ; in the goat's- 

 beard and dandelion the pappus is stalked, whUe that of 

 the hawk- weed is sessile. These and other minor 

 difl'erences are probably not without physiological signifi- 

 cance ; indeed the thistle, dandelion, and many other 

 composites have most exquisite contrivances for securing 

 the proper development and exposure of the pappus 

 hairs. For the present, however, we shall leave out of 

 account pappus and other appendages developed on fruits, 

 and confine our attention exclusively to cases where the 

 hairs arise on the seed itself. 



Where seeds are not provided with hairs or other appen- 

 dages, their connection with the placenta on which they 

 grow is usually severed before the seed-vessel opens ; lying 

 loose and detached in the interior, they are scattered by 

 the bursting of the ripened capsule. This arrangement is 

 quite unsuitable in the case of plumed seeds. The plumes 

 have no room to develop and unfold in the limited space 

 within the seed-vessel, and were they at once scattered by 

 the sudden opening of the capsule, their damp matted 

 hairs would be of little or no use for the purpose for which 

 they are intended. The hairs require to be dried and 

 disentangled ; for this purpose it is essential that the air 

 have free access to them. Between the opening of the 

 capsule and the detaching and discharge of the seeds, some 

 little time must elapse to admit of the proper expansion 

 and desiccation of the plumes ; the seeds must be retained 

 in the nest, so to speak, until they are fully fledged. It is 



clearly an advantage as re- 

 gards dissemination that the 

 seed should remain attached 

 to the parent plant until it 

 has become fully equipped 

 for flight. This is usually 

 attained by the seed remaining 

 attached to the placenta by 

 its funiculus, which does not 

 give way until the aril is 

 perfectly expanded ; but in 

 the willow-herb the retention 

 of the seed while its pinions 

 are maturing is efifected in a 

 still more curious way. A 

 common weed in gardens and 

 hedgerows, the willow-herb {Ejiilubiiim montanum), ^OBSessea 

 as remarkable an arrangement for dispersion as is to be 

 found in the British flora. To the manner in which the 

 fruit of this plant, in conjunction with its arillate seed, has 

 become specialized for the promotion of dispersion we wish 

 now more particularly to ask the reader's attention. The 

 small pink flowers of the willow-herb are too well known 

 to require description ; they have four parts in each whorl, 

 as the rule is in the fuchsia order to which it belongs. 



Via. I. — Transverse Section 

 of Opening Capsule, showing 

 how the Seeds are attaclied 

 to the Valves by their Hairs. 

 Much magnified. 



The enchanter's nightshade {Circaa Vutetiana) and the 

 evening primrose ('A'nof/wra) also belong to the Onaip-ucea. 

 Other allied species of Epilohium are the French willow or 

 rosebay, and the great hairy willow-herb ; these have 

 handsome flowers compared with their humbler relative. 



The ovary of the willow-herb is situated below the 

 flower, and develops into a long narrow pod, very little 

 thicker than the flower-stalk. On a transverse section it 

 is seen to consist of four cells with the seeds attached in 

 the inner angles at the centre. Each cell of the ovary 

 is a long narrow chamber, containing a single vertical 



Fia. II. — 1. Longitudinal Section tlirough Loculus of Willow-herb 

 Fruit, showing the parting of tlie Arillodes. 2. Two ValveB with Sejds 

 suspended by Hairs. 3. Seed witli Arillode. Magnified. 



column of seeds closely packed one on the top of the 

 other. The arillode of hairs on each seed is parted in 

 the middle, the two portions being crushed between the 

 next overlying seed and the walls of the loculus, as shown 

 in Fig. II. 1. The hairs are four or five times as long as the 

 seed itself, and, at their upper extremities, are caught and 

 firmly clasped by the edges of the folded carpels (Fig. I.). 

 The narrow, elongated capsule, when ripe, splits into four 

 very slender splint-like valves, two or three inches in 

 length. The fruit first opens at the top ; the separation 

 of the valves is very gradual ; as they dry they diverge 

 from the central axis and from each other, and curve out- 

 wards, but remain united by their bases until long after 

 all the seeds have been discharged (Fig. III.). But when 

 the capsule splits in this way, the seeds, though detached 

 from the placenta, do not fall out ; they are tethered by 

 their hairs to the four slender segments into which the 

 capsule has separated. Each pair of valves resembles the 

 letter V, the seeds being hung between its two arms 

 (Fig. II. 2). This cresset-like arrangement might roughly 

 be compared to four ladders set in a square. On the 

 middle of each rung of the ladders a seed is perched, the 

 rungs themselves being the hairs of the arillode, which 

 in each seed is parted in the middle, one half being attached 

 to one valve and the other half to the next. The manner 

 in which the seeds are superimposed on one another gives 

 to each of the four vertical series a herring-bone appearance. 

 One effect of this curious arrangement is that as the 

 valves curl in drying, and become more and more divergent, 

 each seed in turn is lifted up out of the loculus and is at 

 the same time pulled away from the central axis. The 

 uppermost seeds in each row have their hairs tightly 

 stretched, but with the ends still firmly grasped in the 

 grooved valves. This condition continues some little time, 

 but as the hairs get drier they are one by one released 

 from the clutch of the carpels. The latter gradually 

 slacken their hold, and the seed slowly spreads out its 

 little flossy fan-like arillode. The topmost seed is at length 

 held fast by a single hair ; finally it shakes itself free and 

 escapes. The others in each of the four series follow in 

 regular succession. 



