FIXING TRANSGRESSIVE VIGOR IN NICOTIANA RUSTICA 171 



ments could be related to the heritability of a character as determined from 

 F2 and first backcross data. 



Results on the three main characters studied were similar in that there 

 was no indication of complex genie interactions, and that estimates of the 

 number of effective factors were low and of the same order of magnitude in 

 each. If we wish to assume that the "reach" or selection differential (in terms 

 of standard deviations) was the same for each character, and this is approxi- 

 mately correct though exact records on this point are lacking, then the gain (in 

 terms of standard deviations) due to selection should be roughly pro])ortional 

 to the heritability. The gain was calculated as the difference between the 

 mean of selection Al and the mid-parent value, divided by the standard 

 deviation of the F2 (1949 data. Table 10.2). 



The relationships between heritabilities and gains can be observed by com- 

 paring the last two columns in Table 10.6. With regard to plant height and 

 leaf length, both heritability and gain are higher in the former character; 

 though the gain is less in plant height than would have been anticipated from 

 the relative heritabilities. Some possible explanations for this latter result 

 could be that the selection differential for plant height was lower than for leaf 

 length, that there was a relatively more rapid reduction in heritability, or 

 that an approach to a physiological limit for tallness was made. 



The gain in node number is disproportionately high in relation to its 

 heritability. Some possible explanations for this result could be that the 

 selection differential was higher, that there was a genetic correlation with 

 plant height, or that the selected character was determined by a preponder- 

 ance of recessive genes (see Fi result), and individual plants selected for high 

 node number were largely homozygous for recessive+genes. 



DISCUSSION 



The experimental results have shown that first generation crosses among 

 different varieties of Nicotiana riistica exhibit different degrees of character 

 expression ranging from the smaller parent value to above the larger parent. 

 By selection and inbreeding it was possible to develop an essentially true- 

 breeding improved line which exceeded the best Pi or Fi in most character- 

 istics measured. 



This same type of result has also been obtained in our experience with the 

 commercial species, N. tahacum, and it may be generally characteristic of 

 self-fertilized plants, as, e.g., Phaseolus ■vulgaris (Malinowski, 1928), soy- 

 beans (Veatch, 1930), and Galeopsis (Muntzing, 1930). 



Crossbreeding 

 There have been relatively few fundamental changes in the standard 

 domestic varieties of N. tabacum over a long period of years, except for recent 

 development of types resistant to destructive diseases (Garner, 1946). 



