llO STUDIES IN INDIAN SUGARCANE SEEDLINGS 



pans, led to a detailed study of the caue inflorescences. These were carefully 

 examined for ripe seed without success. It was noted, however, that in 

 many of the flowers the stamens were poorly developed and the anthers un- 

 opened, while inside their loculcs there was a mass of undeveloped pollen 

 mother cells. In Java the method of determining whether sugarcane pollen 

 is fully formed (and presumably fertile) has been to test it with iodine solution 

 for the occurrence *of starch. The presence of this substance indicates that 

 the pollen grains are healthy, and a blue coloration by iodine therefore 

 shows that the pollen is useful for fertilization. An examination of the pollen 

 grains in the unopened locules at Coimbatore showed absence of starch, while 

 those emerging from split locules were found to be full of this substance. 

 The Java method was therefore replaced by the simpler observation as to 

 whether the anther locules were open for, in that case, fully matured 

 pollen grains were found to be invariably present, and by this means the 

 percentage of fertile stamens in any inflorescence could easily be determined. 

 Judging from the analogy of the inflorescence of the pepper vine 

 {Piper nigrum),^ it was thought possible that, although abundant pollen 

 was to be found in the arrows of other canes in the same field, it was of 

 first importance for the production of seed that good pollen and recep- 

 tive stigmas should be found side by side in the same flower. The arrival 

 of a set of arrows of a cane growing at Bangalore, in March 1912, with plenty 

 of good pollen, was made a test case and all the arrows available of this 

 variety were immediately sown, with the result that a further lot of 32 seed- 

 lings was readily obtained (cf. Plate II). 



In later work, tlie stamens of all cane inflorescences have been examined 

 in the following manner. The stamens are shaken out on to a clean piece of 

 paper and preserved in a small envelope. 200 of them ai-e taken and passed 

 lapidly under a dissecting lens and divided into three classes : — Opened, closed, 

 distorted. It has been found tliat, as in the pej)pe]', if the anthers do not open 

 at the time when tlie flowers mature, they remain permanently closed under 

 all conditions and, once opened, they remain so permanently. They can thus 

 be examined at any time. The percentage of open anthers in each inflores- 

 cence dealt with is entered in the following tables wherever it was obtainable. 

 The failure of previous sporadic attempts at laising cane seedlings in India 

 can be readily explained, in that the state of the anthers was not observed. 

 Almost all such attempts a])pear to have been made in Northern India and 



' Barber, C. A. "TJie Varieties of Cultivated Pei)per." Depl. Aqr., Madras, Vul. HI. BiiH- 

 50, 1906. 



