42 BULLETIN 698, U. S. DEPARTMENT OF AGEICULTUEE. 



is thus evident that feterita can scarcely compete with milo as a pro- 

 ducer of grain; and inilo, it should be remembered, is consistently 

 outyielded by Dwarf milo. 



WHITE DURRA. 



The durra subgroup differs from the milos chiefly in the broader 

 and more hairy glumes, not transversely wrinkled, and the flatter 

 seeds. Two durra varieties were introduced into Cahfornia in 1874 

 under the names "White Egyptian corn" and "Brown Egyptian 

 corn," it being reported that they had come from Egypt. They 

 have since become known as White durra and Brown durra, respec- 

 tively. Like the milos, they differ from each other only in the color 

 of the seeds. 



The durras are low-growing plants with dry stems and few leaves. 

 The heads are ovate or oval, rather compact, and commonly recurved 

 or pendent. The broad seeds are lenticular in shape and much more 

 flattened than in the case of milo or feterita. Very similar forms, no 

 doubt the progenitors of these, are still fomid in cultivation in North 

 Africa and in western Asia under the Arabic general name Dari. 

 The forms of Turkestan are taller, with much more compact heads, 

 and are known there as Dzhugara, a name probably from the same 

 root as jowar, the name used in India for sorghum. 



The white form of durra was rather extensively grown in the early 

 days of dry farming in Kansas under the name Jerusalem corn. The 

 brown form never has been commercially important. There are 

 real objections to these varieties as farm crops. Among them are 

 the recurved heads, the easily shattering seed, and the covering of 

 hairs on the glumes. The first makes harvesting difficult, the second 

 causes heavy loss of seed in wind or storm, as well as in harvest, and 

 the third is very irritating to workmen in thrashing. Careful selec- 

 tion has eliminated the pendent heads and Nos. 27 and 81 have 

 heads completely erect The shattering habit and hairy glumes 

 have not yielded to simple selection, and none of the many hybrids 

 has been entirely satisfactory in other respects. 



Nine different lots or races have been included in the experiments, 

 seven being included at the start in 1908 and two having been added 

 since. In the last three years all but one have been discarded, and 

 that one has been retained chiefly for illustration and comparison. 

 Of the nine lots tested, Nos. 27 and 81 are selections from the ordinary 

 White durra found in the United States, while No. 283 is an unse- 

 lected lot of the same. No. 129 represents the Dzhugara from Tur- 

 kestan. No. 244 is a form with less compact heads, called Edra, 

 from Arabia. Nos. 247 and 248 are from India and not closely 

 related to any of the others. The results of these experiments are 

 shown in Table XIII. 



