363 



Rainfall 

 in inches. 



28 days immediately preceding April 29, 0.00 



9 " " " May 14, 0.00 



7 " " " May 22, ..... 0.00 



18 " " " June 21, only 0.22 



14 " " " July 6, only 0.13 



24 " " " Aug. 16, only 0.26 



27 " " " Sept.2L, only 0.05 



Some of these periods of drought appear short, but many 

 of them were in reality longer than they seem, for the 

 showers separating them were light and altogether insuffi- 

 cient. 



The effect of the dry season is shown by the low yields 

 obtained in nearly all experiments conducted on upland. 



II. Vakiety Test of Corn, 1896. 



For this test sixteen plots were used. The land was quite 

 xmiform in fertility as was indicated by the close agreement 

 between the duplicate plots. Fertilization, culture, etc., 

 were identical for all plots. The distance, 4% by 8 feet, 

 or 14 square feet per plant, is probably less than is advisa- 

 ble for most of the upland of this vicinity. 



The following table gives the number of pounds of thor- 

 oughly dry unshucked corn required to afford 56 lbs. of 

 shelled corn, the percentage of grain in the unshucked corn, 

 and the yield per acre of each variety, arranged in order of 

 productiveness: 



