314 THE LIFE OF THE PLANT 



ful to the organism ; it is only the struggle for existence 

 and selection which control variation in a definite way 

 by destroying harmful deviations and preserving useful 

 ones ; so that scarcely perceptible variations, when 

 accumulated during a long series of generations, in the 

 end are considerably accentuated. Let us try to explain 

 by means of an illustration what part of the phenomenon 

 may belong to variation as such, and what part to the 

 subsequent action of selection. A while back we tried 

 to explain the course by which a symmetrical flower is 

 derived from a regular one. We pointed out a whole 

 series of transitional forms, but of course this does not 

 explain the primary origin of bilateral symmetry, the 

 primary deviation from regularity. We may conjecture 

 with a certain degree of probability that this transfor- 

 mation has taken place under the influence of the force 

 of gravity acting upon the branches as they developed. 

 We have seen ^ that growing organs change the direction 

 of their growth under the influence of the force of 

 gravity, and this change of direction depends upon 

 irregularity of growth in the upper and lower parts of 

 an organ. The same effect is also manifested in 

 another way : horizontal branches become unlike on 

 their upper and lower sides ; the leaves, uniformly 

 distributed on all sides of the erect main stem, 

 in horizontal branches distribute themselves in one 

 horizontal plane, and so on. 



A considerable number of similar facts concerning 

 flowers apparently justify this opinion. Flowers of one 

 and the same plant have been observed to be regular or 

 slightly symmetrical, according to the place they occupy 

 on the flower axis. Thus, for instance, plants with 

 regular flowers, having lateral flowers either in an 

 almost horizontal or in a drooping position, often 

 acquire a slightly bilateral form ; while the apical 

 flowers of the same cluster, or erect flowers, preserve an 



1 See chapter vii. 



