C. A. BARBER 115 



two-thirds of the animal rainfall descends. These rains are often heavy and 

 accompanied by strong winds, so that many arrows are broken or the pollen 

 is destroyed by thorough wetting. Sometimes, for a week or fortnight on 

 end, no collecting can be done, and this, considering the short time during 

 which fertile arrows can be obtained, is a serious disadvantage. Many of the 

 arrows during the earlier years were obtained at places from five to ten miles 

 from the farm, and in order to ensure their daily inspection and bring the 

 matured arrows in safely, a great deal of travelling under trying conditions 

 was necessary. 



The protection of the flowers from cross-pollination was instituted durin^ 

 the 1912 season, but it was found impossible properly to supervise the distant 

 cages, and this supervision was only effective in cages over the few arrows 

 in the Botanic Garden. The cages are made of iron rods or, preferably, strips 

 of bamboo covered with fine, close muslin, as the pollen is capable of passing 

 through the finest meshes. They hang over the inflorescences from a tall 

 gallows-like support, and can be raised or lowered at will. The whole appa- 

 ratus is a close adaptation of that used in the Java cane-breeding work (PI. V). 

 Because of heavy winds, it has often been found necessarv to place an addi- 

 tional upright bamboo or even two and tie the swinging cage to them. The 

 operation recpiires considerable care, because it has been found that even the 

 slightest permanent bend of the long, fragile inflorescence stalk destroys the 

 fertility of the whole arrow, and this is by no means an infrequent occurrence 

 in the heavy weather at Coimbatore at this time of year. From past experience, 

 a general impression has been gathered that the very fact of caging is preju- 

 dicial to the full development of the inflorescence, although large numbers of 

 Ncedlings are sometimes obtained in this way. Caging must be done before 

 the arrow emerges from its enveloping sheath, for the stamens have been found 

 sometimes to open before this emergence occurs. Enclosing the inflorescence 

 thus takes place some time before the introduction of foreign pollen. The 

 cane varieties are closely watched about this time of year and the approach 

 of flowering is foretold some time before it occurs by the elongation of the 

 terminal internodes and the presence of small leaves at the top of a cane shoot. 

 The cane is then said to be in '" short-blade " and its record is carefully gone 

 over to see in what way the inflorescence may be used. If, on emergence, the 

 flowers are found to contain few or no open anthers, crossing may be attempted, 

 but we have at present no criterion by which to tell whether the female organs 

 are fully formed, as we have in the male. Our only guide is the successful 

 rearing of the seedlings in the pan after sowing. To illustrate the need for 

 daily inspection of the cages, one fact may be n^entioned here. On several 



