BREEDING DEOUGHT-EESISTANT SORGOS. 



21 



can not be secured unless it reaches the point of flowerinp^ at the time 

 of harvesting. The purpose in the breedino; work described here has 

 been to obtain a drought-resistant and productive strain which will 

 mature early. Such a strain would extend the sorgo-growing area 

 north of its present limits. 



The breeding work with sorgo at Highmore and Bellefourche has 

 been done with a saccharine sorglnmi of the Minnesota Amber type, 

 South Dakota Xo. 341. This strain has slender stalks and rather 

 long, narrow leaves. The plants stool quite freely, having from two 

 to six suckers per plant. The seed panicles become open and spread- 

 ing as the seed ripens. The seeds are reddish yellow in color when 

 separated from the glumes. The glumes, however, are black and 

 either smooth or slightly hairy. In thrashing, manj' of the seeds 

 separate from the glumes. The stock of this variety was found at the 

 Highmore substation in 1903 under the name of ' 'Montana. " This is 

 all that is known about its history. It was grown at Ilighmore in 1906 

 in comparison with two other amber types and proved to be two 

 weeks earlier than the varieties with which it was compared. The 

 earhness of the type has made it valuable as a stock from which to 

 "work. Two valuable selections (PI. II, fig, 1) were made in the course 

 of the breechng w^ork at Highmore, and seed of these has been increased 

 and is now on the market. 



The two selections referred to were very marked in point of earliness 

 and in uniformity of the progeny. It is probable that the early 

 flowering of the mother plant in each of these selections prevented 

 cross-pollination from any of the surrounding plants, which were ten 

 days or more later in flowering. This insured self-fertilization and 

 the resulting uniformity of progeny. 



Yields of sorgo, South Dakota No. 341, at the Highmore substation 

 for three seasons, 1906 to 1908, inclusive, and at the Bellefourche sta- 

 tion for 1908 and 1909, were furnished by the Office of Dry-Land 

 Agriculture Investigations. These pelds are from each of two j^g-acre 

 plats used in the rotation experiments of that office and are as follow^s: 



Table IV. — Yield per acre of air-dry /odder at Highmore and Bellefourche, S. Dak. 



Place and year. 



Yield of 

 rotation 

 No. 33. 



Highmore: 



190G. .. 

 1907 a.. 



1908 



Bellefourche: b 

 1908 



1909 



Average per acre of all plats . 



Pounds. 



11,140 



4,940 



8,150 



2.. 330 

 4,280 



Yield of 

 rotation 

 No. 34. 



Pounds. 



10, 810 



5,700 



7,250 



4.200 

 7,500 



Average 

 vield. 



Pounds. 

 10,975 

 5,350 

 7,700 



3,265 

 5,920 



6,642 



o The low yields in 1907 at Highmore were due to a poor stand of plants in both plats. 



b The lower yields in rotation No. .33 than in No. 34 at Bellefourche for both years, 1908 and 1909, are due 

 to the poorer type of soil where the plats of rotation No. 33 were located. The soil there is very poor in 

 spots, being liable to puddling and to extreme baking when dry. 



196 



