140 ANNUAL REPORTS OF DEPARTMENT OF AGRICULTURE. 



(3) Dwarf ness is the next most important factor in the adaptation of these 

 crops. 



(4) The combination of earliness and dwarf ness is extremely efficient in 

 insuring adaptation to environmental conditions which include frequent periods 

 of drought. 



(5) Dwarf milo, Dawn (dwarf) kafir, and Sunrise (early) kafir are shown 

 to be well-adapted varieties. 



(6) Dwarf milo and Dawn kSfir are meeting with wide approval on the farms 

 of the high, dry plains. 



(7) Germination and stand are governed largely by local conditions at 

 sowing time. 



(8) Tillering, or the production of suckers, is a varietal or group character 

 to some extent. In part, it is correlated with stand and seasonal conditions. 



(9) The production of erect heads is largely a group or varietal character, 

 but is influenced by the same factors as tillering. 



BROOM CORN. 



Experiments in tlie improvement and culture of broom corn have 

 been conducted for several years at the Woodward Dry Land Field 

 Station in western Oklahoma. The results indicate that in that dis- 

 trict the crops should be sown in the first half of May or the last half 

 of June. These seeding periods bring the crop into head either be- 

 fore or after the hot dry period which usually occurs about the mid- 

 dle of August. Spacing the rows 7 feet apart, with the plants twice 

 as thick in the roAvs, gave no advantage over rows spaced 3.5 feet 

 apart with the usual stand in the row. When harvesting is done 

 while the seeds are in the dough stage, a higher yield and apparently 

 a better quality of brush are obtained than w^hen harvesting occurs 

 at an earlier or later stage in the development of the plant. 



PASTURE IMPROVEMENT. 



The problem of improving pastures is becoming increasingly im- 

 portant, as the high prices of foods are more and more forcing atten- 

 tion to the only really cheap feed pasturage. Confirmation of the 

 bluegrass-pasture experiments, in which it was shown that heavy 

 pasturing is much the best method, both agronomically and economi- 

 cally, has been secured by two investigators working on other pasture 

 grasses. Better handling of northern pastures is a matter of great 

 importance, but, unfortunately, available facilities for experiments 

 and demonstrations are wholly inadequate. 



On the coastal plains of the South, unless the soils are much af- 

 fected by drought, carpet grass, supplemented with lespedeza, makes 

 excellent pasture. As a rule, P as f alum dilatatum (golden crown 

 grass) is also excellent. Efforts are being made, with the aid of a 

 new stripping machine, to place carpet-grass seed on the market in 

 adequate quantity. Several newly introduced grasses are promising 

 as pasture possibilities, but the absence of any facilities for experi- 

 mental trials makes progress very slow. The pasture problem on the 

 Coastal Plain must be solved in order to bring about agricultural 

 utilization of these lands at the present time. 



RANGE IMPROVEMENT. 



There is a constant demand for grasses that will increase the 

 amount of feed on the semiarid range lands, over and above that 

 which may be secured by rational grazing of the land under fence. 

 Sweet clover has proved very valuable, but much of the area is too 



