LUTHER BURBANK 



Until we can produce a white blackberry that 

 is "homozygous" for size-factors as well as for 

 color-factors, we shall not obtain a fruit that will 

 breed true to size as well as color. 



A similar analysis might be applied to the 

 various other pairs of unit characters that are 

 represented in any given fruit or flower. And the 

 essential principle, stated in Mendelian terms, to 

 be aimed at by the experimenter who would fix 

 a newly developed type of plant so that it will 

 breed true from seed, must be to render the plant 

 "homozygous" for the factors of each pair of unit 

 characters involved. If that can be done, the 

 plant will breed true; if that cannot be done, the 

 plant will not breed true. 



In the olden phrasing, this would be spoken of 

 as "line" breeding — a method long familiar to 

 every breeder of plants or animals. 



Fixing a Type in the Second Generation 



In actual practice, where only two or three 

 unit characters are involved, it may be possible 

 to produce a new type that breeds true, or is fixed, 

 in the second generation. In such a case the time 

 element may be ignored. 



Take, by way of illustration, Professor Castle's 

 guinea pigs, to which reference has more than 

 once been made. Suppose we have as parent 

 stock a black guinea pig with a smooth coat, and 



[242] 



