36 PHYLOGENY 



gress at large, as well in practice as in nature, whilst intraspecific 

 selection is the base of highly valuable but local and transient im- 

 provement. 



This principal difference between mutative and fluctuative selec- 

 tion can be supported by a critical examination of the processes of 

 fluctuation themselves. Fluctuations are the responses of the indi- 

 vidual plants to the outward influences to which they are subjected. 

 Of old, all these factors were thrown together under the name of 

 nourishment, and already a century ago, Knight has pointed out how 

 largely the variability depends on this factor. Since then, the term 

 nourishment has been replaced by that of life-conditions, which is, 

 of course, more appropriate on closer analysis. 



Light and space, soil and water, temperature, and numerous minor 

 factors, determine the growth of the plant and of all its parts. Quite 

 obviously the development must depend upon them, and at least a 

 large part of the observed fluctuating variability finds its cause and 

 its explanation in these influences. It is readily granted that in 

 observations of plants and animals, taken from their native localities, 

 these influences are liable to escape the observer. In the experiment 

 garden, however, exactly the same fluctuations occur, and statistical 

 studies find a material, which is in no way inferior to that which is 

 afforded by nature. Here the climatic conditions are daily seen at 

 work. The differences in soil and manure, in space and exposure, are 

 in large part dependent on the conscious will of the experimenter. 

 Partly, of course, they escape his direction, but even then they are 

 followed and controlled with utmost care, not to say with great 

 anxiety. In the same bed one individual is affected by them in this 

 way, and another in a diverging direction, but these relations, though 

 often unavoidable, are commonly obvious, at least in their main 

 features. Thence it comes that the experimenter is strongly impressed 

 by the dependency of fluctuations on outer conditions, whilst the 

 observer takes the variability as a fact of dubious and hypothetical 

 explanation. 



In our gardens we observe our plants during the whole period of 

 their development. At each moment they undergo the influences of 

 the prevailing conditions. But it is evident that each part of the plant 

 must respond to them in its own way. One branch may be exposed 

 to the sun, whilst others are more or less shaded. The first will 

 enjoy all the effects of full light and vigorous assimilation, whilst the 

 latter have only a scant supply of organic food. On richer branches 

 the flowers will be more numerous and larger, and their variable 

 parts produced in greater abundance. On the poor sides reduction 

 must be the rule. 



It seems quite superfluous to work out this discussion any further. 

 It leads directly to the conclusion that fluctuating variability is, at 



