96 COBN CROPS 



Selecting the ears. — If yield is to be the principal object 

 of selection, it will not be necessary to hold closely to some 

 one type of ear. In fact, since we do not know definitely 

 what particular type of ear in a variety may do best in a 

 new locality, it would seem wise to select several types, 

 the main consideration being that the ears are sound and 

 well matured. 



Number of ears to select as foundation stock. — Excep- 

 tional ears are not common, there being probably not more 

 than one in every fifty to one hundred ears. Therefore, 

 if one starts with only a small number of ears, twenty- 

 five to fifty, he may not find a single exceptional yielder 

 in the lot. Not less than one hundred ears, and pref- 

 erably two hundred should be tried out in the prelimi- 

 nary trial. 



The test plat. — Great care should be exercised in pro- 

 curing a uniform piece of land for the test plat, as everj^- 

 thing depends on being able to compare in an accurate 

 way the yields of the different ears. The land should not 

 be exceptionally rich, but only of the average fertility of 

 the region. If the land can be plowed twice — say fall- 

 plowed, and then backset in the spring — and disked sev- 

 eral times, this will do much toward equalizing conditions. 



Size of plat. — Half an ear will plant a row 16 to 20 

 rods in length. However, there will be less error if the 

 rows are duplicated, and it is best to plant two rows 8 rods 

 long from each ear. One hundred ears ^vill make two 

 hundred plats 8 rods long. This will take a piece of land 

 32 by 11 rods or 16 by 22 rods ; or two test plats one-half 

 this size on different parts of the farm may be used, dupli- 

 cating the experiment in each. 



Check plats. — No matter how carefully the land is 

 selected, it may lack uniformity; for this reason, check 



