148 kempton: endosperm color and albinism in maize 



obvious that in a population of yellow and white seeds having 

 an association between white endosperm and albinism the per- 

 centage of homozygous yellow plants would be greatly increased 

 owing to the differential death rate between the yellow and 

 white. 



The five classes of seeds planted with the resulting seedling 

 classes are shown in Table 1. 



Arranged in the form of a fourfold table by combining the 

 four grades of yellow the classes are as shown in Table 2. 



TABLE 2 



The coefficient of association by Yule's method 3 is 0.795 ±- 

 0.072 which is certainly too large to be ascribed to chance. 



Using Pearson's formula 4 for the mean square contingency, 

 C 2 = 0.395 ± 0.057, which is a deviation above a zero correla- 

 tion of 6.9 times the error. Although the number of individuals 

 is small, the fact that the proportion of albino seedlings increased 

 as the intensity of the endosperm color diminished may be of 

 some significance. If it is assumed that the differences in shade 

 between the four classes of yellow endosperm are due to sepa- 

 rate factors, it would seem that these several factors are corre- 

 lated with albinism, which on the linkage hypothesis would 

 necessitate that these associated factors be located on the same 

 chromosome. 



As with many simple characters there are several degrees of 

 albinism which grade from white to yellow by imperceptible 

 stages. The present albino seedlings, while they were easily 

 separated from the normal plants, varied among themselves 

 from almost pure white to an appreciable amount of yellow. 



3 \ule, G. Udny. On the methods of measuring association between two attri- 

 butes. Journal Royal Statistical Society, 75, Part VI, May, 1912. 



4 Pearson and Heron. Biometrika, 9: 167. 1913. 



