44 APPLICATION OF PRINCIPLES OF HEREDITY TO BREEDING. 



vided they all yield well, so as to insure as much cross-fertilization 

 as possible in the seed plot. 



The fact is, we know very little indeed about the relation between 

 yield of corn and type of ear. Prof. A. E. Grantham, of the Delaware 

 Agricultural Experiment Station, recently called the writer's atten- 

 tion to the fact that in communities that have been unaffected by 

 modern ideas about corn breeding, the best ears of corn are usually 

 of the so-called slick" type. He suggests that this may be a case 

 of the survival of the fittest. Farmers have from year to year 

 selected sound ears for seed, paying little or no attention to type. 

 The prevalence of slick ears may therefore represent a case of the 

 survival of the fittest. In testing local varieties of corn unaffected 

 by modern ideas of selection alongside of the improved varieties, 

 Professor Grantham states that the local varieties yield about as 

 well as the others. The writer can partially verify this statement 

 for southwestern Missouri. On his own farm there is a variety of 

 corn that has been grown there for at least thirty years. The best 

 ears of this variety are decidedly slick; on good land it has yielded 

 80 bushels per acre. A single year's test of one of the noted improved 

 varieties in comparison with this local one indicates that the improved 

 variety will outyield the other considerably, but it is important to 

 note that the improved variety, although selected for excellence 

 for fifty years, has never been selected for uniformity of type and 

 has not a few slick ears in it. 



The amount of careful investigation that this question of relation 

 of type to yield has received is wholly inadequate. It ought to 

 receive careful attention at the hands of our best investigators. 



There is another method of breeding open-pollinated crops, like 

 corn, that, while it has not been extensively tried, seems to deserve 

 consideration. This method consists of maintaining two pure strains 

 w^hich are not close kin, and raising each year seed which is a cross 

 between these two varieties and using this seed for the field crop 

 the next year. This method was first proposed by Dr. G. H. Shull^ 

 in an article read before the American Breeders' Association at its 

 Washington meeting in January, 1908. Shull's results have been 

 confirmed by Dr. E. M. East, of the Connecticut Agricultural Experi- 

 ment Station, whose work is referred to later in these pages. A simi- 

 lar plan was recently proposed by Mr. G. N. Collins, of this Bureau, 

 in a buUetin entitled ''The Importance of Broad Breeding in Corn."^ 



a See Report, American Breeders' Association, voL 4. 



& Since the above was written. Mr. Charles P. Hartley, of this Bureau, has called 

 the attention of the writer to some recommendations made in 1893 and 1894 by Prof. 

 G. E. Morrow and his assistant, Mr. F. B. Gardner, of the Illinois Agricultural Experi- 

 ment Station. In Bulletin 25 of the Illinois station these investigators say: "The 

 fact that increased yields can be obtained by crossing two varieties is pretty certainly 

 165 



