450 FERTILIZATION AND FORMATION OF FRUIT IN PHANEROGAMS. 



and 431), the walls of their capsules being exceedinglj' thick and strong. In not a 

 few Steppe-plants the seeds within are protected by the nature of the inclosing 

 pericarp during the hot, rainless summer season. Remarkable amongst them are 

 the big Umbelliferous genera Prangos and Cachrys, the schizocarps of which are 

 protected by a thick spongy wall not unlike elder-pith in nature. Preserved within 

 these walls against desiccation, the embryo secures this further advantage, that from 

 the relative largeness and lightness of the fruits they are readily dispersed by the 

 wind over the Steppe. 



In dry, dehiscent fruits protection against unfavourable climatic conditions is 

 extended only so long as the seeds remain attached to the parent plant; in achenes, 

 nuts, and schizocarps it lasts longer, however. For in the latter classes of fruit the 

 pericarp accompanies the seed after severance, protecting and aiding it during its 

 passage, and often assisting it at germination. All those developments of the seed- 

 coat, met with in the cases in which the seeds themselves are liberated, are replaced, 

 in these non-dehiscent fruits, by the pericarp or other associated portion — calyx, 

 bracts, and the like. These structures are fashioned so as at once to preserve the 

 seed on its journey, be it by air or water, and to attach it to its germinating-bed 

 by various irregularities of surface — pits, furrows, warts, or even sticky excretions. 

 Further, it is important that arrangements be provided so that the young plant 

 should have access to water at certain spots on the fruit wall, and that on germi- 

 nation it should be able to push out its young rootlet without undue eflbrt, as, for 

 instance, in the Water-chestnut and Bulrush (c/. vol. i. p. 607, figs. 1443,4,11,12,13^ 



The stage of development at which the embryo is detached from the parent 

 plant is not the same in all cases. In the Maidenhair Tree (Gingko hiloha) the 

 embryo is but slightly differentiated at the time when the plum-like seed falls. The 

 egg -cell has been already fertilized, and the envelojjing tissues of the seed have 

 matured, but the differentiation of the embryo is postponed till after the seed has 

 fallen. So, too, in Orchids and in many parasitic and saprophytic plants, such as 

 Cuscuta, Orobanche, Monotropa, and Balauophoreis, the embryo, at the moment of 

 severance of the seed, is of the most rudimentary character. But in the majority 

 of Phanerogams the embryo shows a differentiation into plumule, and radicle, and 

 cotyledons. In CeratopJiyllumi the plumule has already slightly elongated and 

 exhibits a number of little leaves, and in Nelumhiwm the leaves show a differentia- 

 tion into blade and petiole. In the Mangrove Tree (EJiizopho7-a Mangle, see fig^, 

 341 ^) the embryo attains to a very considerable degree of development whilst still 

 attached to the parent plant. Its root penetrates the wall of the ovary (fig. S41 ") 

 and ultimately attains a length of 30-50 cm. and a thickness of 1-5 cm. and a 

 weight of some 80 grams. Finally, the young plant breaks away from its sucker- 

 like cotyledon and falls into the mud below, where it speedily unfolds a pair of 

 gi-een foliage-leaves (c/. vol. i. p. 604). Thus, in the Mangrove Tree, it is not the 

 seed but the embryo which is detached from the parent plant. Comparing the 

 Gingko to an oviparous animal, the Mangrove might be regarded as viviparous. 



The envelopes which surround the embryo at the moment of detachment vary 



