EXPERIMENTAL STUDY OF HEREDITY 559 



either appearing in their fullness, or not appearing at all. 

 From more careful observations we know that such is 

 not the case. A blossom may, for example, be more or 

 less pink, an odor more or less strong, dwarfs are not 

 all the same height, but fluctuate around a mean. We 

 conclude therefore that characters do not behave as 

 units, and that the conception of ''unit-characters" is 

 erroneous. The evidence does, however, seem to justify 

 the conclusion that the factor or factors, whatever they 

 may be,^ that are causally related to the given character 

 do behave as units. We may therefore designate them 

 as character -units. They are commonly known as genes. 

 Quite probably, in many if not all cases, more than one 

 gene or character-unit is involved in the production of 

 any given character. 



481. Applications of MendePs Law. — Over loo pairs 

 of structural and color characters have been found, in 

 plant breeding, to behave more or less closely in accord- 

 ance with the Mendelian conception. In peas alone over 

 20 pairs of characters are expressed in successive genera- 

 tions, in accordance with this law. Among the more 

 striking results which are explainable upon Mendelian 

 theory are the following: 



1. Mottled beans have been produced in the Fi genera- 

 tion by crossing two varieties, neither of which had mottled 

 seeds. Vaiious types appeared in the F2 generation. 



2. Jet black beans have appeared in the Fi generation 

 from a cross between two varieties, one of which had pure 

 white seeds, the other light yellow. Various shades and 

 colors appeared in the F2 generation. 



3. In one case three distinct varieties of beans, breeding 



* Substance or condition, we know not what, within the germ-cells. 



