FIELD CROPS. 735 



kernels of each ear into yellow starchy, yellow sweet, white starchy, and white 

 sweet groups. The 4 ears used were selected at random from about 2 bu. of 

 ears bearing Fs kernels, and secured by crossing a white sweet with a yellow 

 dent variety. Both varieties used were pure in the sense that each bred true 

 to the general type to which it belonged. 



Tables state the names of the observers, the classifications of kernels made 

 by each, the range of variation shown by the classification, the mean deviation 

 from the mean (total and net), and the total deviations of the counts of all 

 observers. The original and second counts of certain observers are also stated. 



In view of the data presented the author concludes that no two observers 

 agreed in the observed F2 Mendelian ratios because their classification of 

 kernels disagreed by reason of (1) somatically intermediate kernels, and (2) 

 the personal bias or idiosyucracy of the observer. These differences were so 

 great as to demonstrate " that the personal equation is a factor which can not 

 safely be neglected in work of this character." The personal equation was 

 smallest in the case of the observers most experienced in the observation of 

 variation in connection with the particular class of material here observed. 



No evidence appeared of the difference between the progeny of somatically 

 intermediate kernels and that of distinctly nonintermediate kernels of the 

 same gametic constitution. 



Corn culture, J. M. Kimbkough (Georgia Sta. Bui. 97, pp. 93-101). — Earlier 

 results of these or similar experiments have already been noted (E. S. R., 22, 

 p. 633; 25, p. 232). 



In a test of 17 corn varieties for which seed was donated by the growers, 

 the yields ranged from 15.07 to 26.15 bu. per acre. Vallenweider Blend, Harley 

 Improved, and Whatley Improved yielded over 25 bu. per acre. 



In the nitrogen test in which " 3 formulas were made to carry precisely the 

 same amounts of phosphoric acid, potash, and nitrogen " the 3 years' aver- 

 age yield secured from the cotton-seed meal formula was 28.94 bu. of shelled 

 corn per acre as compared with 31.57 bu. with nitrate of soda, 31.92 bu. with 

 sulphate of ammonia, and 25.38 bu. on the check plat. " The cotton-seed meal 

 formula made less corn than nitrate of soda and sulphate of ammonia formulas 

 every year during the test." The experiment has been conducted on a different 

 part of the farm each year. 



In another experiment the 3 years' average yield secured after the use of the 

 cotton-seed meal formula was 33.96 bu. of shelled corn per acre as compared 

 with 35.25 bu. after a tankage formula and 35.52 bu. after a half tankage 

 and half cotton-seed meal formula. From these 2 experiments the author 

 feels warranted in concluding that " there is a great loss by the southern 

 farmers using cotton-seed meal as a source of nitrogen." 



Detasseled corn produced an average yield of 68.04 bu. of shelled corn as 

 compared with 67.25 bu. when detasseling was not practiced. The 2 years' 

 average yields of seed from the butts, middles, and tips were 25.7, 22.85, and 

 25.97 bu. of shelled corn per acre, respectively. The average yield for 2 years' 

 planting of 2 stalks per hill was 55.04 bu. per acre as compared with 42.45 bu. 

 from only 1 stalk. The hills were 3A ft. apart in each direction and all plats 

 were fertilized with 300 lbs. of 16 per cent acid phosphate, 60 lbs. nitrate of 

 soda, and 60 lbs. of muriate of potash per acre. The 3 years' average yield 

 secured by the Williajnson method was 32.62 bu. per acre as compared with 

 34.18 bu. secured by the ordinary method. The methods followed are described 

 in detail. 



Cotton culture, J. M. Kimbrough (Gcor0a l^ta. Bui. 98, pp. 101-122, fig. 1).— 

 The earlier results of these or similar experiments have already been noted 

 (E. S. R., 22, p. 635; 25, p. 233). 



