STMM A!;\ OF THE FOUR YEARS* WORK. 57 



Of all the meteorological data, the distribution of the rainfall 

 appears to he the most potent factor. At the Maryland and Florida 

 stations there fell during the growing season (June and July for the 

 the former, April and May for the latter) apparently just enough rain 

 to satisfy the requirements of the growing crop. The South Carolina 

 station had a fair distribution of rain, and the Connecticut station an 

 excessive rainfall at the period of growth when the sugar was forming. 

 The total rainfall for Maryland was far below the average for this 

 Mation. The relation between the sugar content and these data is 

 discussed in connection with figures 7 to 11, inclusive (pages 62 to 67). 



SUMMARY OF AVERAGE DATA FOR 1908. 

 Analytical data for Crosby and Stowell Evergreen corn at Jive stations, 1908. 



Meteorological data, 1908 Jor fire stations. 



Data not given for March. 



Sunshine given for Jacksonville, Fla. 



< Sunshine given for Atlanta, Ga. 



d Sunshine given for Washington, D. C. 



SUMMARY OF THE FOUR YEARS' WORK. 

 GENERAL DISCUSSION. 



In reviewing the investigation, the fact that the corn with the 

 highest percentage of sugar was grown in South Carolina seems of 

 prime importance, since it has always been held that corn grown in 

 the South soon lost its sweetness. South Carolina has held the dis- 

 tinction of producing the highest percentage of sugar in the Crosby 

 corn for three out of the four years, and in 1906 it was practically 



