32 APPLICATION OF PRINCIPLES OF HEEEDITY TO BREEDING. 



seed. Taking the yield of the unselected seed as a basis, the results 

 were as follows: 



Source of seed. 



1904. 



1905. 



1906. 



1907.a 



1908.& 



From high-yielding hills 



122 



127 



147 



125 



171 



From unselected hills 



100 



100 



100 



100 



100 



From low-vielding hills 



70 



55 



77 



66 



91 



a See Report, American Breeders' Association, vol. 3, p. 191 et seq. 

 b See Circular No. 90, Ohio Agricultural Experiment Station. 



SELF-FERTILIZED SPECIES. 



The effect of selection on self-fertilized species, such as wheat, 

 barley, and oats, is essentially the same as it is on species propa- 

 gated vegetatively. Doctor Nilsson in his remarkable work at 

 Svalof , Sweden, has many times taken an unselected lot of seed from 

 some cultivated variety of wheat or oats and planted each seed indi- 

 vidually to study the character of the plants produced. He rarely 

 finds two plants exactly alike. But when he saves the seeds from 

 these plants separately the next year the progeny of each plant is, as 

 a rule, found to be so much like the parent plant as to be indistin- 

 guishable from it except for such fluctuations as may be due to en- 

 vironmental influences only. 



Occasionally in work of this kind a plant is found which is not 

 homozygote in all its characters. In other words, it is not abso- 

 lutely pure bred. These plants split up in the next generation 

 according to the law of recombination. Furthermore, their presence 

 indicates that there is occasionally cross-fertilization in wheat and 

 oats, so that ultimately in a wheat field there may be found prac- 

 tically every possible combination of all the characters present in 

 the field, and in tim.e every one of these combinations will come to 

 exist in some individuals in homozygote form; for, as will be seen 

 later, a self-fertilized plant tends to split up into all the fixed forms 

 which can be made from the various combinations of the characters 

 present in it. 



Neglecting for the present the occasional cross-fertilizations in a 

 field of wheat or oats and the resulting heterozygote plants that are 

 produced in this manner, which will be considered under the next 

 heading, selection without cross-fertilization in self-fertilized species 

 can have no effect except to enable the breeder to find those indi- 

 viduals which are best among the population with which he is deal- 

 ing. After he has found these individuals he can not improve them 

 by selection. On the other hand, he may be able to hold them up 

 to a high standard by means of selection, for presumably, as in the 

 case of potatoes, the hereditary characters present in wheat may 

 change in their tendency to develop. Especially may characters 

 that are present get into the habit of failing to develop and thus 



165 



