12 EXPERIMENTS IN THE HYBRIDISING OF INDIAN COTTONS. 
representative leaves. The uniformity of the plants of a bed 
was most striking ; this was particularly the case with lot No. 3, 
which were all small plants and sparingly branched like their 
Jari parent, with large bright yellow flowers like Jowari 
(fic, 15). 
Seed was collected from these beds during the months April 
to August, and from lot No. 3, which were the earliest, some were 
sown in April. It is from this lot that the plants of the 3rd and 
4th generations of this cross were raised. The others were sown 
in September in smal] seed-beds and planted out in rows and beds 
as explained above. 
It was at once seen that the herbacewm (of Gammie) 
type of leaf, of which there was no sign in the first generation, 
(fig. 15) had reappeared in its original form in a large number of 
plants of the next. 
Of those of No. 3, out of 620 plants 370 had leaves like 
Jari, 128 like Jowarz, 121 more or less intermediate. That is 
to say 21% had the character which was not present in the first 
hybrid generation. 
From No. 4, of 102 plants 54 were like Jari and 22 like 
the herbacewm, a percentage of 21°6. Of lot No. 5 however 
a larger proportion were classed as intermediate—of 163 plants 
114 were like Jari, 33 intermediate, and 16 herbaceums. I refer 
only to the general form of the leaf. Jari and Jowarr leaves 
differ also in texture and surface, but it was difficult in the short 
time available to discriminate these characters. Some classed 
above as like Jar7 had for instance even narrower lobes, and had 
the soft feel and surface of Jowari. Ina few the margins of the 
leaves were waved, a strikingly exaggerated instance of this 
being shown in fig. 10. 
Table LI] shows the behaviour of the plants from neglectum 
(Jari-like), and of the herbaceuwm or Jowari-like plants in the 
next generation, 
