Section 13 — Plant Genetics and Breeding 



nexion with the Gatersleben mutation work with 

 barley. As cross parents distinct mutant lines 

 of high yielding capacity were used, belonging to 

 the following groups: (a) stiff-strawed erectoides, 

 (b) early maturing, (c) naked-grained, (d) 

 smooth-awned. If two of the characters were to 

 be combined, e.g. a b, isolation of the desired 

 double recessives followed in F2 generation. 

 These were used for a second crossing step, e.g. 

 (axb) ■ c or (axb)x(cxd), if three or all the 

 four characters were to be combined. The most 

 promising progeny lines of numerous isolated 

 F 2 plants were tested in yield trials for several 

 years. The yielding capacities of these combi- 

 nation lines on an average are at about the level 

 of the parent mutants and original varieties. 

 These forms are, consequently, distinguished by 

 two, three or even four desired specific characters 

 and sometimes by increased yields too. Each 

 of them seems to be more usable, in at least one 

 respect, than the original varieties or any of the 

 parent mutants. 



13.81. Inbreeding and Hybrid Studies in Marrow- 

 stem Kale Biassica oleracea L. var. Acephala 

 D.C. T. D. Johnston (Aberystwyth, Great 

 Britain). 



In this naturally outbreeding species it was 

 found that inbreeding causes a marked depres- 

 sion in yield of leaf to an average at I3 of 50 

 per cent of the original population. The depres- 

 sion was due to decrease in leaf size and not to 

 reduction in number per plant. 



On hybridization mean yield was restored and 

 diallel cross analysis of the data for six inbreds 

 and their hybrids showed additivity of gene 

 effects on leaf number with no heterosis effects, 

 whereas leaf size showed marked heterosis. 



All possible double crosses were also produced 

 from the Fi hybrids in an attempt to determine 

 a means of predicting the best combinations 

 for commercial exploitation. As would be ex- 

 pected from this genetic situation the number of 

 leaves per plant could be forecast with useful 

 accuracy from the values for constituent in- 

 breds or Fj hybrids. Leaf size could not how- 

 ever be so predicted. Although the mean leaf 

 size of the three possible double-cross combina- 

 tions obtainable from a given set of four inbreds 

 showed a significant correlation with the mean 

 of those inbreds, the range of values among the 

 combinations was very large and reference to Fi 

 data did not indicate which permutation would 

 be best of the three. 



Thus, as no correlation was found between 

 number of leaves per plant and their size, it 



appears that to obtain the highest yielding 

 double crosses severe selection can be carried 

 out among potential inbred parents for leaf 

 number, but for leaf size no useful preliminary 

 selection can be performed. 



13.82. The Genetic Control of Growth Rate in 

 Tomato. W. J. Whittington and W. E. Peat 

 (Sutton Bonington, Great Britain). 



A diallel cross experiment was carried out in- 

 volving Lycopersicon pimpinellifolium and sev- 

 eral varieties of L. esculentum in order to in- 

 vestigate the genetic control of relative growth 

 rates. One investigation followed the relative 

 growth rates for total dry weight and leaf area 

 in Fi and F2 generations; another followed 

 growth rate for leaf number in Fi, F 2 , Backcross 

 and F3 generations. Growth rates were calculated 

 as bi' and b 2 ' from polynomial regression equa- 

 tions fitted to the log measurements of the data 

 and as initial growth rate from the exponential 

 period of growth. The results for dry weight and 

 leaf area were similar in that bi' was found to 

 be inherited with complete dominance. Although 

 there were significant differences between cros- 

 ses in values of b 2 ', the large environmental com- 

 ponent of variation masked the genetic effects. 

 There was heterosis for initial relative growth rate 

 in the crosses involving L. pimpinellifolium and 

 negative heterosis in two of the crosses between 

 varieties of L. esculentum. This was probably 

 due to gene interaction. For leaf number bi' 

 and initial relative growth rate were found 

 to be inherited additively with varying degrees 

 of dominance. 



The results showed that the inheritance of 

 growth rate could be studied successfully by the 

 methods of biometrical genetics and also that the 

 successive measurements of growth with time 

 allowed a greater understanding of the genetic 

 and physiological data obtained. The experiment 

 continues that described previously/ 1 ' 



1. Kheiralla, A. I. and Whittington, W. J. 

 Genetic analysis of growth in tomato. I. The 

 Fi generation. Ann. Bot. 26, 489-504. 



13.83. Effect of Growing Conditions of Fi Soft Spring 

 Wheat on Variability in F 2 . T. Ya. Zaru- 



BAILO (U.S.S.R.). 



There were made crosses between soft spring 

 wheat varieties with different duration of their 

 growing periods. 



237 



