98 STUDIES IN INDIAN TOBACCOS. 



the base is different (see Plate XXI, in which both parents 

 and typical leaves from the Fg generation are shoA\Ti). The 

 measurements were obtained by measuring the lamina on 

 either side of the mid-rib and obtaining the mean value. 

 This is preferable to any measurement across the leaf as it 

 eliminates the width of the mid-rib, which is probably an 

 independently varying character. The outline of Type 23 

 is almost straight, while that of Type 38 is suddenly contracted. 

 On h^'bridization, the F^ has a bigger indentation than either 

 parent. In the Fg, petiolate forms occur in which the amount of 

 lamina on either side of the petiole is less than .3 cm. All stages 

 between these and Type 23 were found. Nine cultures were 

 carried on to the Fg. The petiolate forms, Nos. 117 and 213, bred 

 true in the third generation. If the indentation in the parents be 

 due to two different factors or combinations of factors, then the 

 effect of their combined presence might be to reduce the lamina 

 to a negligible amount. These forms possessing both factors 

 would be homozygotic and breed true. Two cultures, Nos. 155 

 and 204, which in the F2 possessed a small amount of lamina gave 

 a progeny of petiolate and sessile forms which could be easily 

 separated both by eye and by measurement into three distinct 

 groups in the ratio 1:2:1. The actual numbers were as follows: — 



Culture 159 — petiolate 25, intermediate 52, sessile 27. 



Culture 204 — petiolate 24, intermediate 48, sessile 33. 

 The parent plants were probably homozygotic for one factor and 

 heterozygotic for the other. 



Three other cultures gave progeny in which a certain 

 number of petiolate forms occurred, but these formed a 

 series with the sessile forms, and the number of such 

 plants was much less than a quarter of the whole. In two 

 cultures, Nos. 104 and 111, no petiolate forms at all occurred. 

 If the presence of both indentations is possible, the absence of 

 both must also be possible. Some leaves with an outline 

 showing even less indentation than Type 23 were found, but 

 the range of variation of the latter was, in 1913, so great that 



