436 EXPERIMENT STATION RECORD. [Vol. 40 



and 11 low-yielding strains were selected for examination. The high-yield- 

 ing strains produced 20.77 bu. more per acre than the low-yielding strains in 

 1916 and 14.72 bu. more in 1917. 



The data relative to the various ear characters were in close accord for the 

 two years and in general favored the high-yielding strains. A significant corre- 

 lation was observed between yield and length, average circumference, average 

 circumference of cob, uniformity of exhibit, shape of ears and trueness to type, 

 character of tips, uniformity of kernels, and shape of kernels and size of eerm, 

 while the correlation between yield and ratio of butt to tip circumference, per- 

 centage of grain, number of rows, average length of kernels, character of butts* 

 space between kernels, and space between rows was small. The results are said 

 to indicate that points emphasized in the corn score card may be of value in 

 selecting high-yielding strains, and that high-yielding strains are high-scoring 

 strains. 



A fifth pair of factors, Aa, for aleurone color in maize, and its relation to 

 the Cc and Rr pairs, R. A. Emkrson (New York Cornell Sta. Mem. 16 (1918), 

 pp. 231-289, fig. 1). — Observations of Fi. F,, and F. individuals are said to indi- 

 cate the existence of a fifth pair of aleurone color factors in addition to the 

 four factor pairs concerned in the development of aleurone color in maize, pre- 

 viously described by East and Hayes (E. S. R., 25, p. 736) and East (E. S. R., 

 28, p. 634). Considerable data are presented in tabular form, and certain other 

 factors, genetic and otherwise, concerned in aleurone color development are 

 discussed. 



The fifth factor pair designated as Aa was found to bear such a relation to 

 the other factors C< . Rr, It, and Prpr, that dominant A, C, and R, and duplex 

 recessive i, must all be present for the development of any aleurone color. 

 Duplex pr, together with the other factors, resulted in red aleurone, while Pr 

 gave purple. F 2 ratios of colored to colorless approaching 27:37 are noted and 

 are compared with the well known 9 : 7 F a ratios. Variation in the percentage 

 of colorless individuals was sufficient to cause an overlapping of the two 

 classes, the range for the 9:7 class being from 36 to 53.7 per cent and for the 

 27:37 class from 44.9 to 66.5 per cent. The mean 1% percentages were found 

 to be 42.91 ±0.28 and 57.79 ±0.21 and the theoretical percentages 43.75 and 

 57.81, respectively. 



The hypothesis that color developed only in the presence of dominant A. C, 

 and R, and that all three were heterozygous in Fi was subjected to every 

 genetic test known to the author with results said to be quite in accord with 

 expectations, as follows: 



(1) Colorless F 2 individuals bred true colorless in F,. while colored F, indi- 

 viduals were found to be of four kinds with respect to their behavior in F«, 

 giving ratios of colored to colorless of 1:0, 3:1. 9:7, and 27:37 in approxi- 

 mately the expected numerical relation of 1 : 6 : 12 : 8, respectively. Colored Fi 

 individuals of the 9:7 class were tested and were found to give F« ratios of 

 1:0. 3:1, and 9 : 7 in about the expected numerical relation of 1:4:4, respec- 

 tively. The results in F 4 , so far as determined, were in agreement with the 

 hypothesis. (2) The seven classes of colorless individuals expected on the 

 basis of the hypothesis, namely. aCR. AcR, ACr, Act, aCr, aeR, and acr, have 

 been found, demonstration of their existence having been made possible by 

 the use in crosses of the three classes aCR, AcR, and .ICr, known as aleurone 

 testers, after these had first been isolated by random intercrosses of colorless 

 Individuals. The results expected from crossing the three aleurone testers 

 with each of the 27 possible genotypes involving A, C, and R are noted, and ex- 

 amples Illustrating some of these results are given. The effect of the degree 



