VARIATIONS AND MODIFICATIONS 79 



which arises from a change in the germ-cell is a variation. But it 

 is conceivable, indeed, as we shall see, it is certain, that the germ- 

 cell may be affected (e.g. by nutrient fluids or by poisons), and the 

 child which springs from it altered, while the hereditary poten- 

 tialities of the gzrm-plasm, which passes on to the descendants, 

 remain unchanged. 1 In such cases the alteration in the child is 

 just as much a 'modification,' and just as little transmissible, as if 

 the agency which produced it had acted, not on the germ-cell, but on 

 its cell-descendants that is on the tissues of the child. Evidently, 

 then, if we accept the modern meaning, and the word variation is 

 to have a signification sufficiently clear and precise to make it 

 valuable in science, it should be limited to those changes in the 

 germ-plasm which mark real alterations of the hereditary ten- 

 dencies ; thus it should not be applied to those temporary alterations 

 in the well-being of the germ-plasm which result from changes in 

 nutriment or from the presence of toxins and the like alterations 

 which may result in modifications in the child that springs from 

 the germ-cell in which they have occurred, but are not reproduced 

 by the descendants. It is in this precise sense we shall use the word. 

 127. Either variations are 'spontaneous' or else they result 

 from changes in the hereditary tendencies due to the direct action 

 of the environment on the germ-plasm. By the term spontaneous, 

 it is not intended to imply that variations ever arise without 

 antecedent cause, but merely that the cause or causes of them are 

 inherent in the germ-plasm itself. Thus, if the germ-plasm gained 

 or lost qualities as a consequence of the normal nuclear activity, 

 growth, and change which precede cell-multiplication, or if, owing 

 to inherent causes, the division of it between the daughter-cells 

 were qualitatively inexact, the variations thus resulting would be 

 spontaneous. Again, when offspring have two parents, a father 

 and a mother, 2 the variations from one, or other, or both parents 

 that may result from that mingling of germ-plasms which occurs 

 in the conjunction of sperm and ovum, are to be reckoned as 

 spontaneous, for they do not arise from the action of the environ- 

 ment external to the germ-plasm. By variations which result from 

 the action of the environment are meant those variations which 

 arise through the direct and immediate action of its surroundings 

 on the germ-plasm. In multicellular organisms, especially warm- 

 blooded organisms, the cells of the soma and the substances (food, 

 toxins, etc.) contained in the nutrient fluids constitute the main 

 effective part of the environment of the germ-cell and its contained 



1 See 132 and 136. 2 Some offspring have only one parent, see 233. 



