166 



Pearl and Surface. 



tion due to environieut. lu the Fj genei'ation twenty-five percent of 

 the plants would be AA or 28 inches tall, fiftj' percent would be Aa 

 or 20 inches tall and twenty-five percent aa or 12 inches tall. If more 

 than one pair of factors are present the number of F* classes will be 

 greater and the percentage frequency will be different. 



Let us see if a somewhat similar theory will interpret the growth 

 facts brought out in this paper. Let us assume first that there are 

 two pairs of independent factors: that dominance is absent and that 

 the effect of each factor is cumulative. The corn plant is open ferti- 

 lized so that a random sample of corn plants at any time would have 

 the constitution of a stable MendeUan population in which all possible 

 matings take place at random. Pearson (:Ü4) has shown tliat this con- 

 stitution will be the same as that of the individuals in the segregating 

 generation. 



A random sample of corn .plants would then have the constitution 

 obtained bv matiug at random, 



cfcf gametes 



By the usual recombination tliis gives the combinations and pro- 

 portions given in columns one and two of table 23. 



Table 23. 



Showing the zygotic constitution of a random sample of corn 



plants together with the theoretical mean quintile position of 



each class of plants under each of two assumptions for the 



values or the factors concerned. 



