PROTECTION OF SEEDS AGAINST WEATHER. 440 



of slifcs arise near the base, while in the Butterworfc {Pinguicula vulgaris, fig. 340^) 

 the fruit splits into two valves. How all these capsules behave is shown in the por- 

 tion of the illustration with the shaded background (figs. 340 1' to 340**'). Closure 

 is so complete that the entrance of moisture into the capsules is impossible, and 

 the seeds have absolute immunity from premature wetting. Even in the case of 

 capsules with lateral slits, whei-e there is a possibility of moisture entering, the 

 same arrangement prevails, as is well illustrated in the Fragrant Orchis {Gynina- 

 denia Conopsea, cf. figs. 340^ and 340'). 



Of capsules with openings directed downwards there are but few, and in these 

 the dispersal of the seeds is not accomplished quite in the same manner as in those 

 just described. In the Funkias {FtmJcia ovata, Sieboldi, subcordata, &c.) the pendent 

 capsules open by three valves at their depressed tips, and concurrently some of the 

 seeds are shot out into the air — often to some little distance. Those which remain in 

 the capsule after the sudden opening are not, as one might expect, deposited in a 

 heap on the ground close by, but, being attached to the valves by delicate strings, 

 and having a flattened form, ofier considerable surface to the wind, which ultimately 

 loosens them and bears them far away. The fact should be emphasized that in 

 many cases fruits only open under the influence of dry winds, and that the same 

 wind which promotes their dehiscence at the same moment scatters the seeds. This 

 is very well shown by the Scotch Pine {Pinus sylvestris). Its cones only separate 

 their scales in the afternoon when the air is driest and a wind is blowing. When 

 the conditions for this are favourable one hears quite a series of noises in the tree- 

 top, caused by the separating of the scales, and at the same time the winged seeds 

 go spinning through the air. The scales protect the seeds in this case against 

 wetting, and indeed against other dangers as well. It should be remembered that 

 any protective arrangement is but rarely directed solely against one particular 

 source of danger. An envelope may protect the embryo at one time against the 

 wet, at another against excessive desiccation; sometimes the same envelope may 

 alsd ward off attacks from undesired guests of the animal world. 



Developments on fruits whose chief function is the protection of the seeds 

 against desiccation are, on the whole, scai'ce. In certain portions of Australia the 

 whole vegetation enters upon a sort of summer sleep ; during this period no rain 

 falls and the surface of the earth is hardly ever wetted with dew; such a dryness 

 of the air and of the soil prevails that plants are compelled to suspend ojserations 

 for a while. It has been already explained how the foliage is protected against 

 drying up (cf. vol. i. p. 296); but the embryo also, which has arisen within the ovule 

 during the period of vegetative activity, has to be preserved during this period. 

 This is achieved by the massive development of the ovary wall, which in this 

 respect exceeds anything met with in the Floras of other regions of the globe. The 

 seeds of the Wooden Pear {Xylomelwm pyriforme, see fig. 325 ^ p. 431) are inclosed 

 in a strong pericarp, the wall of which attains a diameter of 2 centimetres, and can 

 resist the greatest dryness for very long periods. So also is it with the seeds of the 

 Australian species of Banksia and Eucalyptios (see figs. 324 and 325 i^-^*' pp. 42& 



Vol. II. 79 



