THE GENERAL SHAPES OF PLANTS. 127 



standing alone, and in positions where the winds do not injure 

 them or adjacent objects shade them, shrubs and trees develop 

 with tolerable evenness on all sides, is an obvious truth. Equal- 

 ly obvious is the truth that, when growing together in a wood, 

 and mutually interfered with on all sides, trees still show 

 obscurely radial distributions of parts ; though, under such 

 conditions, they have tall taper stems with branches directed 

 upwards a difference of shape clearly due to the different 

 incidence of forces. And almost equally obvious is the truth, 

 that a tree of this same kind growing at the edge of the wood, 

 has its outer branches well-developed and its inner branches 

 comparatively ill-developed. Fig. 197, which very inaccur- 



ately represents this difference, will serve to make it manifest 

 that while one of the peripheral trees can be cut into some- 

 thing like two similar halves by a vertical plane directed to- 

 wards the centre of the wood a plane on each side of which 

 the conditions are alike it cannot be cut into similar halves 

 by any other plane. A like divergence from an indefinitely- 

 radial symmetry towards an indefinitely-bilateral symmetry, 

 occurs in trees that have their conditions made bilateral by 

 growing on inclined surfaces. Two of the common forms 

 observable in such cases are given in Fig. 198. Here there 

 is divisibility into parts that are tolerably similar, by a vertical 

 plane running directly down the hill ; but not by any other 

 plane. Then, further, there is the bilateralness, similar in 

 general meaning though differently caused, which we see 

 in trees exposed to strong prevailing winds. Almost every 



