32 CKOSSBEEEDING COEN. 



WORK OF 1910. 

 TEST AT SHERMAN, TEX. 



The test at Sherman was located upon fertile black upland of 

 uniform appearance. All the land used had been cropped the same 

 the previous year, had been broken deep, and was in good condition 

 when the corn was planted. 



The planting was made in hills 3^ feet apart each way; the crop 

 was cultivated frequently, and the ground kept free of grass and 

 weeds. The season at Sherman was extremely dry and the yields 

 were very light. 



The order in which the varieties and crosses were planted and their 

 field-row numbers are shown in Table IX. Because of lack of uni- 

 formity in number of stalks per row of the different varieties, the 

 comparisons have been made on the basis of production per stalk 

 rather than on row yields. This method is followed in all cases. 



The comparison is made between the yield of the cross and the 

 higher yielding of the two parents. For practical purposes the cross 

 can not be regarded as an improvement upon existing conditions or 

 as worthy of propagation unless it is superior to the better parent. 

 Where the comparison is between the cross and male parent the 

 average of the two rows of the male parent nearest the cross has been 

 used. This was also done in the test at Waco and Corsicana. The 

 seed of the female parents used in the experiment is in every case 

 taken from the 1908 crop, that is, from the same lot of seed that was 

 used for the beginning of the experiment in 1909. The seed of the 

 crosses is from the 1909 crop. The seed of Chisholm (the male 

 parent) is in part from the 1908 crop, but mostly from the 1909 crop, 

 entirely so at Sherman and Corsicana, and also in the greater part of 

 the Waco test. 



When the crop was harvested, determinations showed the per- 

 centage of moisture to be approximately the same for the different 

 varieties. Taking into consideration the unavoidable percentage of 

 error, it was believed that notliing would be gained by calculating 

 the yields to a water-free basis. In computing the production per 

 stalk only the main stalks were considered, as the suckers were not 

 productive at any of the three places. 



218 



