CONTRIBUTIONS TO THE GKNETICS OF PISUM 355 



will in tlu' following, accordingly, be called the L,. \n\\[, tiie short 

 internode part tiie /,■ i)art. 



From the transgressions in /•':,, however, it is evident that mcjre 

 factors concerning the length ot internodes must be involved in the 

 crosses. It is true that Rmjunda-S in 1920 has remarkably short 

 internodes, the length of six internodes in this year varying between 

 16 — 45 cm., but in 1919 in the same line between 40 — 70 cm. (In 

 1920 tlie peas at Svalöf in general were rather low, owing to un- 

 favourable weather conditions). None the less, the Lv part of F 2 has 

 longer internodes and the h part shorter internodes than either parent. 

 How is this to be explained? It must be remembered that the geno- 

 typical constitution of Ragunda and 0254 may differ in a great many 

 factors. It is then possible that Le — which with certainty is intro- 

 duced by Rdijunda — together with the sum of factors in ()23A pro 

 duces — under the environmental conditions, given at Svalöf — 

 longer internodes than together with the factorial complex of Ragunda, 

 and that h in connection with the factors of Ragunda gives shorter 

 internodes than in connection with the 023i factors. It might, 

 however, be one single factor, dominant in 0234 and recessive in 

 Ragunda that exercises this influence upon the effect of L,, or h- 

 According to K. & P. this factor should be a factor for thick stem, 

 introduced in dominant state by 0234. Before this opinion may be 

 discussed the results concerning 



THICKNESS OF STEM 



must be taken into consideration. In all four crosses the parents 

 have been different in this respect, and segregation has occurred in 

 F2. We have, however, not been able to find the evidently mono- 

 hybrid segregation, recorded from their crosses by K. & P. The data 

 obtained are given in table II. (The number total in each cross is 

 not quite the same in respect to different characters, due to difficulties 

 by collection of data. When determining correlations between cha- 

 racters only those individuals have, of course, been calculated, all the 

 data of which are known). If the low numbers in the V2 mm. classes 

 are disregarded, the curve is continuous and gives no possibility of a 

 distinction between individuals with thick and thin stem. Naturally 

 the segregation may be monohybrid none the less, only that we cannot 

 with any certainty classify the plants in this respect. In the following 



