THE PURE LINE AND SELECTION 155 



any differences found between individuals of a pure line 

 are entirely somatic and not hereditary. 



The importance of the problem of pure line selection 

 for any general consideration of the mechanism of evo- 

 lution is at once apparent. There have been many re- 

 cent investigations besides those of Johannsen to test 

 the result of selection within the four kinds of "pure 

 lines." Some of these investigations are enumerated in 

 the table on pages 136 and 137. 



It is apparent in the first section of the following 

 table that the pure line sensu stricto, that is, the pure 

 line of Johannsen, must be studied with plants alone, 

 since among animals only certain highly specialized 

 parasitic worms, which do not lend themselves readily to 

 selection experiments, produce offspring by means of 

 self-fertilization. The work of the other authors 

 upon plants, mentioned in the first group of the table, is 

 in entire agreement with the work of Johannsen. 



The noteworthy contribution of L. de Vilmorin con- 

 sists in a detailed comparison of preserved specimens 

 of certain pure lines of wheat which were isolated in 

 France about 1840, with their lineal descendants of 

 to-day. In spite of continuous selection for better- 

 ment within these self-fertilized strains during more than 

 60 years, their constancy has been maintained. 



B. CLONES 



With respect to selection within a clone there is an 

 apparent conflict of results. 



