60 STUDIES IN INDIAN TOBACCOS. 



homozygotic combination be produced equal or greater than 

 the sum of both parents, this may be explained by the 

 greater vigour produced by hybridization or by supposing a 

 different rudimentary height in different plants, or that the 

 factor giving height exists in all plants but has no definite 

 external expression. Both these hypotheses, however, would 

 introduce many difficulties. 



The results in the cross Type 9 x Type 51 (Table IV) need 

 not be considered in such detail. The range of variation in the 

 Fo Avas not greater than the range of the parents combined. In 

 the Fjj generation, the eight cultures showed a great difference 

 in tlie range of variation, but only one appeared to be uniform. 

 Culture 740 (Plate XV) resembled the parent, Type 51, in 

 every respect, both in the limits of variation and in the average 

 height. The average height in the case of 740 was 195.2 cm., 

 and the limits of variation 155 to 240 cm. ; in the case of 

 Type 51 the average was 195.4 cm. and the range 160 to 

 230 cm. Both cultures were grown side by side in the 

 field and there was no difference in their appearance. A 

 form resembling one of the parents was therefore isolated 

 in the F^. The difference in the range of variation in 

 the F2 and F3 generations distinctly indicates a segregation, 

 as well as the probable occurrence of homozj^gotic combinations 

 representing heights intermediate between the two parents. 

 One such was isolated in the F4 generation (738-58), as well as 

 a culture resembling Type 9 (738-96). From their appearance in 

 the field, their range of variation and the form of the curve 

 obtained by graphic representation, both these cultures appear 

 to be homozygotic. The average value of the height in 738-96 

 is 60.1 cm., that of 738-58 is 98.7 cm. The height of culture 

 738-96 is probably the same as that of Type 9 (average height 

 57.5 cm.) ; the other forms a novum. The remaining cultures 

 of 738 are not uniform and contain these two combinations with 

 possibly an intermediate. The F4 cultures of 694 are also not 

 homogeneous, but their limits of variation are very different to 



