FUTUEE OF THE SCIENCE OF BREEDING 175 



Quantitative Cases.— Many of the characters with which 

 the breeder deals are of the sharp, clean-cut order. The 

 alternative characters are easily distinguished. A flower is 

 purple or it is red, a bantam is white or it is black, an animal 

 is male or it is female. The difference is a qualitative one and 

 readily appreciated. These are cases of the simplest kind, 

 but, as every breeder knows well, a cross between two pure 

 strains may give a great number and confusion of forms when 

 carried to the Fj generation. Take for example the case of 

 the sweet pea. If we make a cross between the salmon- 

 orange " Barbara " and a lavender such as " R. F. Felton," 

 the resulting plants are a reddish mauve and quite unlike 

 either parent. In the next generation a considerable range 

 of colours and shades appear. Some are like the original 

 parents, some hke the first cross, and many new forms turn 

 up of which some can be recognised as varieties already 

 standardised and named. 



With a little care and a few years' work we can sift out and 

 fix these many forms. What appeared at first sight a com- 

 pound mass of blending colours can be analysed into a number 

 of separate and distinct components. In spite of the appear- 

 ance of grading and blending the differences are all qualitative. 

 Such a case as this of the sweet pea offers no special problem 

 or difficulty for the ordinary methods of Mendelian analysis. 



There are, however, large groups of results which cannot as 

 yet be expressed clearly in terms of quahtative factors. They 

 are to be met with more especially where such characters as 

 size, weight, extent of pigmentation and so forth are concerned. 

 To put it crudely, the problem with which they challenge us 

 is a quantitative rather than a qualitative one. Size inherit- 

 ance may be straightforward. Many cases are known where 

 a, dwarf plant behaves as a simple recessive to a tall one. In 

 all such cases it will probably be found that the dwarf is not 

 merely a tall with all its parts reduced to scale. It will differ 

 in habit as well as in bulk. The factor on which the size 

 difference here depends is fundamentally a qualitative one. 



But there are other cases in which the study of the inherit- 

 ance of size brings to light a more complex result. In illus- 



