CONTRIBUTIONS TO THE GENETICS OF PISUM 



357 



the thickness of stem may hardly have so great an influence upon 

 the length of internodes as to he the main cause of the difference 

 between two so different types as semi-dwarfs and tails. In crosses 

 I — II the total length of 433 individuals was determined. The num- 

 bers obtained are given in table III. As it will be seen, the total 

 length of the Le indivduals in F, widely transgrcedes the length of the 

 Le parent. The transgression of the /,. part of Fn over the /,. parent 

 is not equally strong, but evident. The coefficient of correlation 

 between stem thickness and total length has been determined. In the 

 Le part of Fo is r = 0,i2.'î + O,or)(i and in the /,- part r := 0,258 + 0,o85. 

 These results, too, make it probable that thickness of stem to a certain 



TABLE III. Total length of plants in parents and F, of crosses I + //. 



degree influences the height of stem, but not enough to explain the 

 great transgi;essions in this respect. (The mode of segregation in total 

 length within each part of F2 cannot be determined by the numbers 

 obtained). It seems very probable that the difference which causes 

 the great transgression in total length is the different number of 

 internodes. Unfortunately, the number of internodes was not deter- 

 mined, but there was a great variation and undoul)tedly segregation 

 in this character in Fo. (An attempt at getting hold of the number 

 of internodes by dividing total length by average internode length 

 failed to give results, only indicating the great variation in internode 

 number). White discusses in his collective work on genetics in peas 

 (White 1917) the problem and seems to be inclined to doubt the 

 interpretation of K. & P. He says (1. c. pag. 546): »Hence it seems 



