FTET.D CEOPS. 829 



nictliod ;iii(l tlK> tiX'.-itiiuMit was identical cxi-cpl in llio ])()iiits wliidi disliiiiniish 

 tilt' Williaiusdii method. 



Tile yield of shelled corn was .".O.n Im. iu>r acre on tlie Williamson plat and 

 20.rj hn. on the check i)lat. On the Williamson i)lat each plant averaged 0.r54('. 

 lb. of shelled corn and on the other O.cr.T lb. The average weijiht of ear on 

 the check plat was 0..")4 Ih. as compared with 0.4;") lb. on the Williamson plat. 

 One hundred plants jxrown by the Williamson method ])rodnced 00 ears and 

 nubbins as compared with lilt*, ears and nubbins in tiie dieck test. The average 

 height from the ground at which the ears grew on tlie check plat was 444 in. . 

 and on the Williamson plat 3(!.| in. The Williamson plat showed 2'.) per cent of 

 broken stalks while the check plat showed only 14 i)er cent. 



Applications of 200 lbs. ootton-seed meal, 200 lbs. acid iihosphate, 400 lbs. 

 kainit, and 200 lbs. nitrate of soda per acre were used, but the yield was not 

 sufficient to make the use of such large quantities of fertilizers profitable. It is 

 believed that the proportion of kainit in the Williamson fertilizer is too high. 



It is also suggested that the uimsually large yields obtained by this method 

 on upland in South Carolina may be due more to the freipient plowing under of 

 a croi) of cowpea vines, to the liberal use of nitrogenous fertilizers, and to close 

 planting than to the dwarfing of the plants through omission of cultivation and 

 withholding until late the application of fertilizers. 



Variety tests with cotton and corn, J. F. Duggar and L. N. Duncan (Ala- 

 haiiiu Cdl. Shi. Jiiil. 1SS, jij). <J7-102). — Of ?>2 varieties of cotton under test Toole, 

 Cook Improved, Cleveland, Bancroft. Herlong, and Christopher ranked fii'st in 

 1905, and of 20 varieties compared Cook Improved, Cleveland, Layton, Toole, 

 and Pullnot were the leaders in 1900. Sunflower and Floradora, w'bich were 

 grown in this experiment, are long staple varieties and conmianded in the home 

 market a premium of about 4 cts. per pound. 



In 190.^>, estimating the lint at 114 cts. per pound and the seed at 7 cts. per 

 100 lbs., the total value per acre ranged from ,$20.04 to .$07.04. and in 1900, the 

 lint being regarded as worth only 10 cts. per pound, the total acre value varied 

 from .$4.">.0l to $09.30. 



The leading varieties of corn in a list of 18 and tly'ir yields were as follows : 

 Sanders 28.9 bu. per acre, Marlboro 28.3 bu., Mosby 2f'>.0 bu., and Henry Grady 

 2.5.7 bu. Sanders, Marlboro, and ]Mosby are several-eared sorts. Early northern 

 varieties were again shown to be worthless for Alabama. 



Cotton culture, R. J. Redding and J. M. Kimbrough ((Georgia Sfa. Bid. 

 7.7, pp. 211-240). — Twenty-six varieties of cotton wvvr imder test. In value of 

 total products per acre the leading varieties, given in decreasing order, were as 

 follows : Layton Improved, Cleveland Big Boll, Broadwell Double- Jointed. Cook 

 Improved. These varieties all produced values of over .$0.") per acre. The 

 average total value of lint and seed ])roduced by the 20 varieties was $00.01 per 

 acre, the range being from .$47.1. "► to $7.').i')0 per acre. The average yield of seed 

 cotton per acre was 1,517 lbs., and the range 1,315 to 1.7.54 lbs. 



The average results of this work for 13. years show that the best half of the 

 number of varieties grown each year produced 34.9 per cent of lint and the 

 others 32.5 per cent. In 190(5 the best half of the varieties grown produced 

 30.8 per cent of lint as compared with 37.8 for 1905. The best yielding varieties 

 also showed the larger size of IioUs, although in some years there were exceptions 

 to this rule. Earliness was in favor of the lietter yielding sorts during 4 

 seasons and in favor of the i)oorer yielders the remaining 9 years. 



A high jiercentage yield of lint apiiears very closely related to a high value of 

 total products. The size of the iioUs is to be considertnl in relation to the cost of 

 liicking. 



It was found that a fertilizer aiiplication, of which the nitrogenous ingredient 



