364 
which bore these seeds were exposed); those which survived develop¬ 
ed into the best plants which I had yet seen, the conditions of 
growth being good almost throughout; in spite of a dry November, 
during which however the rain that fell was well distributed. Dur¬ 
ing the last week of January, just when the seeds of this crop were 
ripening, there fell more than 8 inches of rain. The early part of 
February however was rainless, and I was able to get many of the 
seeds dry. 
By this time the process of selecting the strongest offspring of 
certain crosses appeared to have had a considerable effect in the 
direction of producing vigorous plants. For the third and fourth 
generations of the original uncrossed strains, grown at the same time 
as the above and from seeds taken in each generation from the 
strongest surviving plants showed practically no advance upon the 
original generation, as grown at first from English seed, and were 
very poor in comparison with the cross-breds, and this effect was 
still more clearly seen in a further generation. 
(6) This arose from seeds sown on April 6th, 1904, and underwent 
a drought lasting from the 14th to the 27th of the month, during 
which time the young plants were with difficulty kept alive by 
constant watering and shading. Even then very little rain fell until 
the south-west monsoon “ broke ” with considerable violence on 
May 20th. After this constant heavy rain continued to fall during 
the whole of June and July and the plants ripened' their seeds in 
very wet weather. Thus the conditions during the whole of the life 
of these plants appeared to be distinctly unfavourable. Practically 
every seed sown germinated, a fact which evidently depends upon 
the favourable conditions enjoyed by the previous generation ; and 
although a good many seedlings perished—especially the offspring 
of certain combinations—enough were left to make a very good 
show at the time of flowering. The wet weather which prevailed 
at the time of ripening caused much damage ; nevertheless the 
crop was by far the best obtained throughout the whole series of 
experiments, and would doubtless have been most valuable as a 
source of profitable varieties if it had been possible to dry the 
seeds. 
