32 HATCH EXPERIMENT STATION. [Jan. 



The dift'erences in yield are large, but the scale upon 

 which the varieties were grown is small, — too small, in 

 my judgment, to justify sweeping conclusions as to the 

 relative merits of the several sorts. 



The " Dakota" closely resembles the " Early Harvest; " 

 the "loose-headed" variety of the "barn-yard" millet is 

 much less leafy and less valuable than the common form. 

 The so-called "Golden Wonder" cultivated appeared to 

 be like the "Golden." The "Japanese Glutenous," from 

 ^^ Mukodamashi" is very late, and does not mature with 

 us. The variety of glutenous millet from Hokkaido ap- 

 pears to l)c a valuable sort. Moisture tests which are 

 being made will very likely change the relative position 

 of some varieties. 



5. Japanese Millets for Seed. 

 A small area of each of our three leading varieties of 

 Japanese millets was grown for seed. The soil was fer- 

 tilized for each variety at the following rate per acre, the 

 fertilizer being sown broadcast and harrowed in : — 



Manure, 4 cords. 



Nitrate of soda, 125 pounds. 



Dried blood, 100 " 



Tankage, 200 " 



Superphosphate, 250 " 



Muriate of potash 200 " 



The season was not very favora])le for these crops, and they 

 were somewhat injured on several occasions by the Avash- 

 ing of the soil, due to excessive rain-fall. 



Barn-yard Variety {Panicuni cms galli). — The area 

 sown to this variety was .633 acres. The seed was sown 

 May 27, in drills, and was cultivated and hand-weeded. It 

 yielded 1,370 pounds of seed and 4,360 pounds of straw, 

 which is at the rate of 40 bushels of seed and 3 tons 888 

 pounds of straw per acre. 



Japanese Broom-corn Millet {Paniciim miliaceurii) . — 

 The area of this variety was .248 acres. It was planted 

 and cultivated like the preceding variety. The yield was 

 535 pounds of seed and 1,620 pounds of straw, which is at 



