THE OUTER TISSUES OF PLANTS. 239 



be even thought surprising that this tendency can 

 be counteracted to so great an extent by changed con- 

 ditions. Such a degree of modifiability becomes compre- 

 hensible, only when we remember how little a plant's 

 functions are integrated ; and how much, therefore, the 

 functions going on in each part may be altered without 

 having to overcome the momentum of the functions through 

 out the whole plant. But this modifiability being as great 

 as it is, we can have no difficulty in understanding how, by 

 the cumulative aid of natural selection, this primary differen- 

 tiation of the surface in plants has become what we see it. 



§ 273. "We will leave now these contrasts between the 

 free surfaces of plants and their attached or imbedded 

 surfaces, and turn our attention to the secondary contrasts 

 existing between different parts of their free surfaces. Were 

 a full statement of the evidence practicable, it would be 

 proper here to dwell on that which is furnished by the 

 inferior classes. It might be pointed out in detail that 

 where, as among the Algce, the free surfaces are not dis- 

 similarly conditioned, there is no systematic differentiation of 

 them — that the frond of an Ulva, the ribbon-shaped divisions 

 of a Laminaria, and the dichotomous expansions of the Fuci 

 that clothe the rocks between tide-marks, are alike on both 

 sides ; because, swayed about in all directions as they are by 

 the waves and tides, their sides are equally affected. Con- 

 versely, from the Fungi might be drawn abundant proof that 

 even among Thallogens, unlikenesses arise between different 

 parts of the free surfaces when their circumstances are unlike 

 — that in such laterally-growing kinds as are shown in Fig. 

 196 b, the honeycombed under surface and the smooth 

 leathery upper surface, have their contrasts related to con- 

 trasted conditions ; and that in the adjacently- figured 

 Agarics, and other stalked genera, the pileus exhibits a 

 parallel difference, explicable in a parallel way. But passing 

 over Cryptogams, it must suffice if we examine more at 

 Vol. IT. 11 



