540 THE PURE LINE HYPOTHESIS. 



a prioii it is quite as likely that a visible variation in a character 

 of the adult may really correspond to some modification of the 

 factor, representmg the character in the germ-cell from v^diich the 

 individual arose, as that the variation is simply due to some 

 change in the environment during embryonic life. 



6. Pure Lines. 



An almost necessary corollary to the supposed non-inheritance 

 of fluctuating variations would be the existence of the so-called 

 " pure lines " of Johannsen. This observer raised several self- 

 fertilised generations of beans, and he interpreted his results as 

 proving that the hereditary transmission of a character remained 

 the same from one self-fertilised generation to another, although 

 the visible expression of the character might vary considerably 

 among the individuals. In other words, he sought to show that 

 selection within a self-fertilised family had no result. Karl Pear- 

 son has pointed out, however, that the expermients were not at 

 all conclusive, although Morgan* and his collaborators state thcit 

 Johannsen worked '* by methods that have not been questioned." 



7. Modification of Factors. 



The question whether a factor can be gradually modified by 

 the selection of individuals exhibiting small variations in the 

 expression of the character is a most imi^ortant one, and depends 

 upon the circumstance as to whether or not the variations in the 

 character are reflections of variations in the factors in the germ- 

 cells from which the individuals arose. The testing of this 

 problem is not as easy as might at first be suiJi>osed. In the case 

 of animals which are not self-fertilising, a new factor for the 

 character studied is introduced at every generation unless resort 

 is made to in-breeding. 



Utilising self-fertilised generations of plants simplifies the 

 enquiry, and Johannsen worked with such material. His potion 

 of pure lines involves the denial of the power of selection to 

 modify the factors. 



Let us suppose that the red colouration of the flowers of a 

 gametically pure plant is represented by some simple factor. In 

 the fertilised egg-cell from which the plant arose there were two 

 of such factors, one in the chromosome derived from the female 

 nucleus of the ovule, and one in the homologous chromosome 

 derived from the male nucleus of the pollen tube. If the pollen 

 grain came from an unrelated plant, the two homologous factors 

 are doubtless not identical. 



In the formation of mature germ-cells the homologous 

 chromosomes separate from each other, so that finally in the 

 matured germ-cell there will be only one of the pair of homo- 

 logous chromosomes remaining in the nucleus and containing the 

 factor in question. As far as is known, which of the two honiu- 

 logous chromosomes remains in the ger m-cell is a matter of 



* Morgan, T. H., and others : " The Mechanism of Mendelian Heredity." 

 New York, 1915. 



