THE FRUIT AND SEED-DISPERSAL 287 



seen in the parasitic Rafflesiaceae or Balanoplioreae. In sucli cases the large 

 number of seeds co^■ers the risk of germination. 



Sucli differences in number and size of the seeds must profoundly allect the 

 meclianical problem of seed-distribution. The nrinute seeds of an Orcliid 

 can be scattered by a breatli. but this would have no appreciable effect upon 

 a Coco-Xut in its natural husk. 



DlSSEMINATIOX OF SeEDS. 



Seeds being themselves immobile require some agent outside 

 themselves for their dispersal. Various means are effective. Partly 

 the dispersal may be dependent upon the structure of the pistil, as 

 in explosive fruits. But usually it is carried out by means of transfer 

 outside the plant, such as currents of air, or water, or the movements 

 of animals. There is an obvious analogv in this between seed-dispersal 

 and the transfer of pollen. In both cases the immobile plant depends 

 upon external agencies for transfer of essential parts. But the two 

 present quite distinct problems. In pollination the end is to deposit 

 the pollen-grains upon the stigma, and the more accurately this is 

 done the better. In seed-dispersal the end is simply the wide separa- 

 tion of the individual seeds. Accuracy is of no importance. 



In primitive types of fruit, as the carpellary wall matures, its tissues 

 commonlv become hardened. Where the seeds are numerous they 

 require to be set free for germination. The natural course is by 

 splitting or deliiscence oj the carpel ; and this is carried out in various 

 ways. A split most naturally follows along the line of the coalescent 

 margins of the folded leaf. This is the case in the follicle of Aconite 

 (Fig. 227). A similar split mav also sever the carpel at its midrib, 

 as in the pods of Peas and Laburnums. The case of svncarpous fruits 

 is not so simple. Their capsules open sometimes by longitudinal slits, 

 as in Cardamine (Fig. 93) or Hura (Fig. 94I, sometimes by transverse 

 slits, sometimes by valves or pores : and this may occur either in 

 superior or in inferior fruits. The seeds can thus escape, leaving 

 the carpellary structure behind as an empty husk. On the other 

 hand, in cases where the carpels are separate, and the number 

 of the seeds in each is reduced to one, dehiscence is unnecessary, while 

 the firm tissue of the carpel may form an additional protective wall. 

 The resulting achene, or nut, is then shed bodily, as if it were a single 

 seed, though it is strictly speaking a fruit. This is seen in the Butter- 

 cup, or Potentilla (Fig. 229). It may happen also in syncarpous 

 pistils where only one seed matures, as in the Hazel, Acorn, or Coco- 

 Nut. A peculiar case is that of the L'mbelliferae, where the inferior 



