THE TESTING OF SEED 59 



were not. And since but twelve good ears are needed to 

 seed an acre, it was not difficult to choose enough ears to 

 plant a good many acres. The method is as follows. 



The ears are all numbered and set carefully in order on 

 shelves or the attic floor. From each ear are taken six 

 kernels, beginning at the butt and going spirally around 

 the ear to the tip. This is because one end or one side of 



^mm^mmmw;;m& 



FIG. 36. A SAND-BOX CORN-TESTER, SEVEN DAYS AFTER PLANTING. 



Five kernels were planted in each of 42 squares. Which show perfect ears ? 

 Was the seed well grown, or well kept ? 



the ear may be bad. Then, keeping each set of six kernels 

 separate, they are all tested. 



But if one is testing a hundred ears or more, it is very 

 clumsy to do the work with dinner plates. Other methods 

 have been devised. I will speak of the three best. 



The first is the sand-box tester. Make, out of a soap box, 

 a tray an inch and a half deep. Fill it with moist sand 

 level with the top. On the sides, an inch and a half apart, 

 drive tacks. From these tacks lace twine back and forth 

 across the tray, going both ways, so that the sand is marked 

 off into squares of an inch and a half. It is most convenient 

 if there are ten squares to one side of the box, so that the 



