HEEBERT SPENCER's BIOLOGY. 383 



upright growing ones, and this upright direction is brought about 

 in one of two different ways, the explanation of which, as given by- 

 Mr. Spencer, is very ingenious. The one way is simply by the 

 rolling up of the frond into a cylinder or imperfect tube, which of 

 course gives an increased stiffness and power of resistance to the 

 growing fronds. This involution is shown to exist in many Algce^ 

 in Riella lielicophylla, in Jungermannice^ etc. The successive fronds 

 thus rolled round, will very often cohere by their edges, so as to 

 form sheaths.* If during the successive formation of these sheaths 

 one above another, the united midribs continue to grow, and serve 

 as the channels of circulation between the uppermost fronds and the 

 roots, they will increase in diameter, and ultimately form a solid 

 axis wrapped round by the sheaths, as in Dendrobiwn^ but if, on the 

 other hand, the woody bundles of each succeeding midrib, instead of 

 remaining concentrated, are distributed all round the sheath, then 

 the structure will eventually be that of a hollow cylinder, Ps in 

 grasses or sedges. The other way in which a mutually dependent 

 series of fronds may acquire sufficient rigidity to maintain an erect 

 position without involution, is by the thickening and hardening 

 of the fused midribs, these latter, moreover, elongate, so as to allow 

 of the separation of one leaf to a sufficient distance from its fellows. 

 In this way such stems are produced, as those of Jungermannia 

 decipiens and J". Hoolceri. If the separation of the leaves in this 

 way be incomplete, then we get " decurrent leaves," in which the 

 continuity of the axis with the midrib of the leaf is manifest. If 

 the internodes grow more rapidly, then the leaf will become am- 

 plexicaul simply, or if the midrib develop rapidly, an ordinary 

 stalked leaf is the result. Such result being the consequence of the 

 survival of the fittest, inasmuch as the formation of a leaf- stalk T^dll 

 throw the leaves further away from the axis, and so diminish the 

 shading of leaves one by another. Here may be mentioned other 

 arrangements that possibly have reference to the non-interference 

 of one leaf with another, e.g. the spiral arrangement of leaves, one 

 leaf being in course of formation, the one next to it is in process of 

 growth also, but the axis is at the same time lengthening, therefore 

 the second leaf is placed above the first ; but a third leaf has also to 

 be provided for, and .another after this. 



* It is more consistent with truth to say that the sheaths are developed as such; 

 the so-called cohesion of the edges being congenital. 



