THE CAUSE OF SPONTANEOUS VARIATIONS 101 



165. Darwin, following Pallas, maintained that " domesti- 

 cation eliminates the tendency to sterility which is general in 

 species when crossed." l Doubtless he was right. When 

 species are partially infertile, when their respective develop- 

 mental tendencies run on nearly but not quite parallel lines, 

 it is evidently probable that some sperms and ova may vary 

 so that the parallelism is more complete in their cases than 

 in those of others. They are therefore able to unite and 

 develop into individuals, whereas the others are not. Cross- 

 ing partially infertile varieties consists in nothing more than 

 a process of selection of such sperms and ova. When varieties 

 are crossed the offspring are said to resemble one or other 

 parent type more often than when species are crossed. 2 In 

 the latter case they are said more often to be intermediate 

 in character. This is what might be expected. Exclusive 

 inheritance, like blended or particulate inheritance, is pos- 

 sible when the two parents are not very widely separated 

 by descent. When they are widely separated there is more 

 often reversion to the common ancestral type, which in many 

 respects appears intermediate between the two. 



166. We have now finished the first section of our work. 

 Before proceeding to the study of disease, which will form 

 the next section, it will be useful to summarize in a convenient 

 form the conclusions at which we have already arrived. 



I. All the characters of living beings, mental or physical, 

 are either inborn or acquired. 



II. An inborn character is one which takes origin in the 

 germ-plasm. 



III. A variation is an inborn deviation from the parental 

 type. It is therefore a new inborn character, and as such 

 takes origin in the germ-plasm. It may be progressive or 

 regressive in type. 



IV. A progressive variation constitutes a deviation from 

 the parental and ancestral type, which speaking generally is 

 in the direction of increased magnitude and complexity. It 

 results from the complete recapitulation of the parental 

 development plus an addition. 



V. A regressive variation constitutes a deviation from the 

 parental towards the ancestral type. As a rule it is in the 

 direction of diminished magnitude and complexity. It 

 results from an incomplete recapitulation of the parental 

 development. The only occasions when it is in the direction 

 of increased magnitude and complexity are when it has been 



1 Animals and Plants, vol. ii., pp. 88-9. 

 2 Origin of Species, p. 407. 



