24 GEORGE HARRISON SHULL 



row was collected and applied at the appropriate time to the silks of all the 

 plants in the second half of the row. Then the mixed pollen from the plants 

 in the second half of the row was applied in turn to the silks of all the plants 

 in the first half of the row. It was realized that this still involved a con- 

 siderable degree of inbreeding, but it seemed about the only way of carrymg 

 on a continuing program of crossing while still keeping the breedmg com- 

 pletely under the operator's control. 



Two major observations made on the 1906 crop were: (1) that every one 

 of the seven families from selfed parents could be readily detected by their 

 less height, more slender stalks, and greater susceptibility to theattack ot 

 Ustila2o maydis. When the ears were harvested each lot was weighed and 

 it was found that cross-fertilized rows produced on an average about three 

 times as much grain as the self-fertilized. (2) The family A3, from a self- 

 fertilized ear having 12 grain-rows, was practically all flint corn, showing that 

 to be probably recessive. This occurrence of a rather obvious segregation in 

 the 1906 crop remained at the end of the season only an isolated observation 

 which led to no generalization. From the fall of 1905 until his retirement 

 in 1943 Charles Leo Macy assisted me in many of the technical details of my 

 experimental cultures. While I handled the planning and breeding operations 

 as well as the actual poUinations, Macy prepared the plants for selfing and 

 crossing, and counted the grain-rows and weighed the ear corn. The results 

 of these counts for the 1906 crop are given in Table 2.2. _ 



The following quotation from my notebook seems justified here, since it 

 includes the first formulation of the considerations and conclusions which 

 appeared in my report to the American Breeders' Association in 1908, on 

 "The composition of a field of maize": 



^- A c;,, 1007 p'; in 1906) namely each self -fertilized row was 



The same plan was continued, ^i" l^^^;^^ ^^^^^^^ row was divided in half, 



the offspring of a smgle self-fertilized far, and each cross termzea ^^^^ 



each half coming from a single cross-fertilized e^^^' f^^^.^f" ^"J^'e ^Vhlr ear coming from the 

 the first half of the corresponding row of the precedmg > ear, the other ear c n g 



second half. ... . mnA tv,o coif fprtili/pd rows being invariably 



The obvious results were the same as in ^^O^, the self -fert^ed o^^ s g ..^ 



smaller and weaker than the corresponding """f "^^ ^iliz'-^j^^f l^^^ upon me 



evidence on the self-fertilized A v^ry df erent explana ion of^ h^ P^.^.^^ 



by the fact that the several self- ertlllzedro^^s differ rom each ^^^^^^^^ elementary 



morphological characteristics, thusindicatng that the, be^^^^ ^^^^^ ^^.^.^^ ^^^, 



strains. The same point appeared last >ear in the c^se oi tne ^ ^^^^ 



almost a uniform flint corn, but the ^^^'^^^""^^f.^^^J^Z^^SSr^^^ that my corn- 

 It now appears that self-fertilization simply ^^'^f. ^ P^"^ > j^^' but between pure strains 

 parisons are not properly between "^^J^^f .^/^"^Jj^'^j^^^^^ 



and their hybrids; and that a well '•^g^^^^^f ,^^^ ^ °4e^^ ^ of 'corn must have as 



its ?b.-t?^rn!?inraTc^ ^^s^ h^d ^^^^^^^^^^^^^ to\e most vigorous and 



productive and give all desirable qualities of ear and grain. 



The ideas in this quotation represent a discovery in complete disagree- 

 ment with my preconception that my white dent foundation stock, which 

 had been the progeny of a single ear, was essentially a ^^^^'^^^^y f'^l?^'^^' 

 I had before me seven distinct biotypes, clearly distinguishable m their sev- 



