36 VALUE OF FIKST-GENEEATION HYBRIDS IN CORN. 



From the standpoint of the investigation the faihn-es of such 

 experiments are often of even greater interest than the successes, 

 since they may lead to better understandings of the factors involved- 



In reporting results it would seem desirable to state the facts 

 bearing upon as many of the following points as ])ossible: 



(1) Names and descriptions of varieties crossed. — While the names 

 of commercial varieties are almost hopelessly confused, some desig- 

 nations are necessary for purposes of reference, and if these are accom- 

 panied by careful descriptions many errors may be avoided, as well 

 as a needless duplication of work. 



(2) History of the varieties. — This should be traced as far back as 

 possible to throw light on the degree of relationship that exists between 

 the varieties crossed. 



(3) Sources of seed and previous methods of breeding. — Important 

 differences may be expected even where the same varieties are used, 

 depending on whether the seed has been self-pollinated or cross- 

 pollinated; also whether it was the result of mass selection in the 

 field or crib or was derived from a single ear. 



(4) Size of the hybridizing plat and the plats or rows in which the 

 yields are tested. — The ratio between the area devoted to each variety 

 in the breeding plat and that in which the yield test of the same 

 variety is made should be recorded, since it is a measure of the oppor- 

 tunity for selection. If the breeding or hybridizing plat is small in 

 proportion to the area to be planted, it will be necessary to save a 

 large part of the seed for planting and the opportunity for selection 

 will be correspondingly small. The failure to take this fact into 

 consideration is one of the reasons why large field plantings of pure- 

 bred varieties so frequently fail to meet expectations of high yields 

 indicated in the breeding ])lats, where a more rigid selection was 

 practiced. 



(5) Extent of self-pollination in the parent varieties. — Many varieties 

 produce pollen so little in advance of the silks that a considerable 

 proportion of the seed is self-pollinated, and this operates to diminish 

 the yield of the resulting plants. In such cases a part of the increase 

 that might be ascribed to the crossing of two varieties would in 

 reality be due to the depressed yields of the })arent varieties with 

 which the cross is compared. To determine the increase actually 

 due to the crossing, seed from detasseled plants of the parent varie- 

 ties should be included in the yield test, together with ordinary wind- 

 pollinated seed of the same varieties. 



(6) The method by whicli the yields are compared and the precautions 

 against experimental error. — In this connection it should be borne in 

 mind that large })lats do not insure greater accuracy. The larger 

 the ])lat the greater the difliculty of obtaining equal conditions. 



191 



