864 EXPERIMENT STATION RECORD. 



lan«l i>l()wi-(l (il. in. deep yieltled 76 lbs. of seed cotton and 10 lbs. of lint more per 

 acre than land plowed :U in. deep. Subsoiling did not materially affect the yield of 

 lint. CottdU planteil on ridges yielded 344 lbs. of lint per acre as compared with 

 272 lbs. when planted on beds in the usual way. Ridge i)lanting, however, is consid- 

 ered more likely to suffer from drought. The use of a cultivator gave practically as 

 large a yield as the use of the sweep. An additional cultivation given September 19 

 apparently increased the yield of lint c(jtton per acre by 40 lbs. over a plat laid by 

 August 2(). 



Slag phosphate and acid phos])iiate each applied at the rate of 200 lbs. })er acre on 

 black slough land faileil to increase the yield J)ut hastened maturity to some extent. 

 Different forms of potash were also used without material effect. In a test with 

 nitrogenous fertilizers, plats receiving 160, 80, and 40 lbs. of nitrate of soda per acre 

 yielded 112, 64, and 48 lbs. of lint cotton more per acre than the check plat. The 

 increase in yield from 100 lbs. of cotton-seed meal per acre was the same as that 

 obtaine<l from 40 lbs. of nitrate of soda. An application of 3,200 lbs. per acre of 

 barnyard manure apparently increased the yield of lint by 80 lbs., and the same 

 application of corn stover by 16 lbs. per acre. 



The best-yielding varieties of corn under test were Shaw Imi)r()ved, IJlount Pro- 

 lific, and Cocke Prolific, producing 37.7, 35.4, and 34.9 bu. of corn and 2,176, 1,360, 

 and 1,840 lbs. of stover per acre, respectively. Two plats subsoiled on April 4 

 yielded on an average 2.5 bu. per acre less than 2 plats not subsoiled. No material 

 difference resulted from different methods of preparing the soil and planting. In a 

 cultivation test the best yield was obtained from the use of the 12-in. sweep at the 

 first cultivation and the subsequent use of the 20-in. sweep. Owing to a very wet 

 season the continuous use of the turnplow gave very good results. 



Among different fertilizer applications, 88 lbs. of nitrate of soda per acre applied 

 at the second cultivation gave the largest increase in yield. The same increase was 

 obtained from 44 lbs. nitrate of soda and 100 lbs. of cotton-seed meal applied together 

 at the first cultivation. The increase obtained from 88 lbs. of nitrate of soda per 

 acre applied at planting was smaller by 2 bu. than when the application was given 

 at the second cultivation. Manure applied before planting at the rate of 2,702 lbs. 

 per acre yielded a return of $2.20 per ton. In a second test 5,408 lbs. of manure per 

 acre gave an increase of only 4 bu., while 200 lbs. of cotton-seed meal and 80 lbs. of 

 nitrate of soda increased the yield by 7.1 and 7 bu. per acre, respectively. Nitrate 

 of soda was apjilied at the third cultivation, June 12, 2j months after planting. 



Direct api)lication of nitrate of soda and cotton-seed meal failed to increase the 

 yield of Mexican June corn, but the residual effect of 1,600 lbs. of cotton-seed meal 

 applied to the immediately preceding crop of oats showed a gain of 7.2 bu. i>ev acre. 

 Sul)soiling apparently decreased the yield. 



Selecting and improving corn, A. M. Soule ( Univ. Tennessee Record, 7 {1904), 

 No. 1, jyp. 13-27, figs. 10). — The value and methods of selecting and improving com 

 are discussed at some length, and results of planter and variety tests are reported. 

 The planter tests indicate that uniformity in the size of kernels insures unifonnity 

 in the number of kernels dropped by the machine, and consequently tends to pro- 

 duce an even stand. Attention is called to the importance of selecting varieties 

 adapted to the locality, by the fact that on the University farm the varieties leading 

 in yield in 1901, a dry year, were not among the leaders the following season, which 

 was favorable, and others not represented in either year had a better average for 

 the 2 seasons. Tlie vitality test showed a low percentage of germination for butt, 

 tip, and deformed kernels and soft corn. A score card for judging corn in the ear, 

 in use at the College of Agriculture, is presented. 



Cowpea hay, C. L. Newman {Arkamas Sta. Bui. 80, pp. 69-82).— T\\q average 

 yield at the station of cowpea hay for 5 years, from 1899 to 1903, as computed from 

 all plats and all varieties recorded, was 3,169.4 lbs. per acre. This in general did 

 not include the weight of peas and hulls. In 1902, 16 plats yielded over 6,000 lbs. 



