74 STUDIES IN INDIAN TOBACCOS. 



of the lobes (type XLVIII, Plate XLVIII), (3) slightly divided 

 (type XXXVII, Plate XL), (4) divided to about half its distance 

 (type XXV, Plate XXIX), (5) deeply divided (type XVII, Plate 

 XVIII). The undivided portion ot the liinb may be quite iiat 

 (type XXXVIII, Plate XL) or may form slight folds or ridges 

 at the line of junction of the lobes (type XLVIII, Plate XLVIII ). 

 The shape of the lobes varies somewhat. They may be more or 

 less triangular with somewhat straight margins (type XXXIX, 

 Plate XL) or much rounded at the base (type XVII, Plate XXII). 

 The apex may be acute or prolonged into a point (longer or shorter 

 according to the type), which may be straight or oblique. A 

 o-ood example of such apical points is seen in type XVII (Plate 

 XVIII), type XXXIV (Plate XL), and in type IX (Plate XII). 

 Capsule. The capsules in the Indian forms of N . tahacum 

 vary in size, shape, nature of the apex and in the degree to which 

 they are covered by the persistent calyx. The shape may be 

 cylindrical or conical, the apex blunt or pointed. In some types 

 the capsules have a tendency to fall off when ripe together with 

 the peduncle instead of remaining in situ and shedding the seed 

 by dehiscence. In such cases it is necessary to protect any crossed 

 or selfed capsules by a bag just before they are quite ripe. 



2. Modes of pollination. 



The frequency of cross-pollination in Nicotiana tahacum 

 has already been discussed on page 64. Knuth,^ in his Hancl- 

 huch der Bliitenhiologie, drew attention to the fact that the arrange- 

 ments for pollination differ somewhat in the various races. The 

 pollination of all the Indian types of Nicotiana tahacum has 

 been studied to obtain as much information on this point as 

 possible. Of the five stamens, one is always much shorter than 

 the others and is generally well below the stigma. In a few 

 cases, however, it may be long enough for its tip to touch the 

 stigma. The pollen of this stamen is, f herefore, negligible for 

 self-fertilization. The remaining four stamens may be approx- 



l Knuth, Handhwh (In- Bliiteniwlogie, Bd. II, 1899. 



