22 BULLETIN 698, U. S. DEPAETMENT OF AGRiCULTUEE. 



tative period and covers the time elapsing between heading and 

 ripening. 



The height of the plants is the average of measurements made at 

 several, usually 10, points in the plats. 



The varieties are all harvested with a corn binder, leaving a stubble 

 varying in height from 5 to about 8 inches (1 to 2 dm.), depending on 

 the height of the standing plants. The bundles are shocked in the 

 field and allowed to cure for four to six weeks before being tlu-ashed. 

 The shocks are then hauled to the scales and weighed. 



The heads are cut from the bundles by means of a large knife 

 fastened to the edge of a box. Very dwarf strains are not headed, 

 but are thrashed in the bundle unless the weight of the heads is 

 desired. Since the stalks often are not of uniform height, the piece 

 of the peduncle or stem left attached to the heads also varies in 

 length. In general it averages about 10 to 12 inches (2.5 to 3 dm.) 

 long. Where it is desired to determine the proportion of heads in 

 the total crop, the heads are weighed before thrashing and the per- 

 centage computed. 



Thrashing has been done with a small separator. The thrashed 

 seed is weighed as it comes from the separator and the acre yield 

 computed therefrom. 



The yields are based on 60 pounds to the bushel of kafir and 58 

 pounds of all other varieties. After weighing, the thrashed seed is 

 run through a fanning mill and the bushel weight then determined 

 with the standard tester. 



Environing Conditions. 



To aid in an understanding of the results ootained during the 

 9-year period from 1908 to 1916, inclusive, a brief summary of 

 seasonal conditions is given and their effects on crop growth and 

 production are noted. Figure 2 shows the annual and seasonal 

 (April to September, inclusive) precipitation during the 10-year 

 period from 1907 to 1916, inclusive. 



The season of 1908 was favorable to good yields. The precipita- 

 tion in the last three months of 1907 was slightly above normal, but 

 in the first three months of 1908 it was considerably below normal. 

 This left little or no stored moisture at the end of March. However, 

 the rainfall in April, May, and June was about normal and fully 

 sufficient for plant growth. With a rainfall of nearly 2 inches above 

 normal in July and August, most of which fell in July, conditions 

 were favorable for a heavy grain yield. Dwarf milo averaged over 

 41 bushels and Blackhull kafir a little less than 34 bushels per acre. 



The season of 1909 was variable and unfavorable. It consisted of 

 a dry spring and a dry summer with a wet June intervening. Pre- 

 cipitation in the winter of 1908-9 was decidedly below normal. The 



