CONTRIBUTIONS TO THE GENETICS OF PISUM 359 



fluence, however, is not great enough to be the single cause of tlie 

 transgressions in internode lengtli, observed in the crosses here re- 

 corded. 2) Stem tliickness may, as stated by Kkkhij-: and Pellkw, 

 be mainly determined by one factor, T, wliose dominant state pro- 

 duces thick stem. The results obtained in the crosses here recorded 

 neither confirm nor contradict this view. The stem thickness is 

 further influenced by Lc or by a factor coupled with Le.' Le or the 

 other factor has an increasing effect uj)on stem thickness. 3) As to 

 the height of stem it is, as already known, a compound character. 

 Great influence upon plant highness is naturally caused by L,- and 

 also to a smaller degree by the factors that influence internode length. 

 It may be true that the factor (factors?) for stem thickness to some 

 degree inlluences the length of internode and thus the height of stem, 

 but this influence is to small to explain the difference in height 

 between e. g. a semi-dwarf pea and a tall one, both being of the 

 constitution Le- It seems to us, in agreement with White, that the 

 number of internodes to a high degree influences the height of stem. 

 As to the genotypical basis of the number of internodes our results 

 allow no conclusions. 4) The factor for internode length. Le, the 

 main factors for stem thickness and for internode number seem to 

 segregate independently, as far as they are involved in the crosses 

 here recorded. 



THE PLACE OF THE FIRST FLOWER. 



In respect to the number of sterile nodes there is a great varia- 

 tion between different lines of pease. Pure lines are cultivated here 

 at Svalöf with an average of 6 — 7 sterile nodes and further lines with 

 nearly every average number, up to such lines that have 24 — 25 

 sterile nodes. In different years the number of sterile nodes may 

 vary a little in the same line, but on the whole the character is well 

 fixed. In all four crosses here recorded this character has been 

 involved, and the results reached are given in table IV. In crosses III 

 and IV the monohybrid segregation is evident. In cross III there 

 may be some doubt as to the calculation of the two minimum classes, 

 viz. 12 and 13. If, however, the curves are drawn, it will be seen 

 that at least the main part of the 12 class belongs to the left curve, 

 the 13 class to the right curve. Thus segregation occurs in 386 with 

 high and 136 with low number of sterile nodes, corresponding, after 

 3:1 ratio, to an expected segregation in 391,5:130,5 and D/m = 



