530 EXPERIMENT STATION BECORD. [Vol. 43 



group. All characters may be arranged in such a way that they form a single 

 group iu which there is no coellicient of correlation lower than ±0.31. 



The arrangement of the .spil<elets in the pistillate inllorescence showed the 

 nearest approach to Mendelian inheritance. In maize the female spikelets are 

 borne in pairs (double female alicoles) ; in teosinte they occur singly (single 

 female alicoles). While dominance of the maize character was complete in 

 the first generation, this did not hold true in the second generation, many 

 plants possessing both single and double female alicoles; but tlie uuml)er of 

 individuals in which double female alicoles predominated was approxinuitely 

 tliree times the number possessing more single female alicoles. 



The characters of the pistillate inflorescence were found to be subdivided in 

 transmission to a remarkable degree. The maize ear, instead of behaving as 

 a unit was subdivided into a large number of separately inherited units, such 

 as number of rows, closely crowded seeds, and shortened peduncles, all of 

 which were inherited more or less independently. Number of rows was still 

 further resolved into paired or single spikelets and the number of rows of 

 alicoles in which they were borne. A surprisingly large number of plants 

 combined the abundant production of suckers characteristic of the teosinte parent 

 with the sturdy, upright character of maize, resulting in very leafy, compact 

 plants of a desirable forage type. 



Lint frequency in cotton with a method for determination, E. A. Hodson 

 (Arkansas Sta. Bui. 168 (1920), pp. 3-11). — This bulletin defines the term lint 

 frequency, briefly reviews literature bearing on the subject, describes iu detail 

 the method followed in determining this factor for 25 varieties of cotton, and 

 reports in tabular form the averages of the data so secured for 10 selected 

 plants for each of the varieties. Tlie rank of the 25 varieties in lint fre- 

 quency, index, percentage, and length is also given in a table. 



Lint frequency is defined as " the weight in grams of the fiber of uniform 

 length produced per square centimeter of seed surface," and the lint index 

 as " the average amount of lint produced on one seed without I'egard to the 

 size of the seed." It is pointed out that the quantity of lint produced on a 

 cotton seed depends upon the frequency and the length of the lint hairs and 

 upon the area of the surface. High lint frequency was found closely corre- 

 lated with short lint, and attention is called to the fact that it is necessary, 

 therefore, in making selections for higli lint frequency to consider length and 

 percentage of lint. Boykin, which ranked first in lint frequency, second in 

 lint Index, and second in lint percentage, stood twenty-fifth in length of 

 lint, while Dix-Affi, ranking twenty-fifth in lint frequency, lint index, and 

 lint percentage stood first in length of lint. 



Correlations of certain characters in cotton, E. A. Hodson (Arkansas Sta. 

 Bui. 169 (1920), pp. S-15, figs. 3). — Correlations were determined for number 

 of base limbs with number of fruiting branches, number of bolls per plant, 

 and height of plant ; for number of fruiting branches with height of plant 

 and number of bolls per plant ; and of length of fruiting period and other 

 characters studied. Based on the data from material studied, the correlation 

 coefficients for these and other characters, including length of lint and per- 

 centage of lint, weight of boll, and weight of .seed ; weight of seed and weight 

 of boll ; and percentage of lint and weight per boll are reported. 



The results indicated that the number of base limbs per plant depends 

 upan the variety and the environment, without being significantly correlated 

 with any physical character. The number of fruiting branches per plant was 

 found clo.sely correlated with the height of the plant, which is in turn dependent 

 upon the environment and the variety. A very low correlation was found to 



