C. A. IUrhkr 217 



and foniale organs mature at practically the same time. For certain 

 crossing it will be necessary to emasculate all the flowers of one parent 

 arrow. Even if we succeed in cutting off the n^ajority of the flowers 

 without injuring the rest, each experiment would mean the erection of 

 a lofty staging with practically a glasshouse at the top in which an 

 assistant could use a dissecting microscope. The state of the weather 

 would usually render futile any less cumbrous arrangement. This is 

 at present out of the question. C^ertain, direct crossing does not at 

 present appear to be feasible at Coimbatore and, added to this, the 

 outlook for such work is discouraging because the results obtained at 

 Barbados appear to have been unsatisfactory, one after another of the 

 carefully nurtured crosses having been abandoned on being tested 

 in the field. 



There are two ways which suggest themselves of approaching the 

 problem of obtaining known crosses indirectly. It has been noted that 

 the different cultivated varieties of sugarcane vary a great deal in the 

 development of their essential organs. Some have never been observed 

 to flower: others have flowers which do not emerge from the en- 

 veloping sheaths : yet others have partial or total sterility in male 

 (or female?) organs, while a number produce healthy arrows with good 

 pollen and fertile ovules. The commonest case at Coimbatore is that 

 only a certain percentage of stamens open. If we could obtain a variety 

 in which no pollen is formed at all or no stamens open (and there seems 

 to be a close relation between the two, stamens which do not open not 

 containing fully formed pollen and the converse), crossing that variety 

 with one producing good pollen w^ould be easy. One or two such cases 

 have already been observed and crosses obtained, but these happen to 

 be of little value from the economic point of view. The best local cane 

 at Coimbatore, the 'Vellai,' arrows freely every year but produces 

 comparatively little good pollen. Advantage has been taken of this 

 during the past year to pollinate this variety with about a dozen others, 

 in the hope of obtaining crosses between them. The withered flowers 

 of each arrow are kept and carefully analysed as to the percentage of 

 open anthers, and the probability of obtaining crosses or selfed seedlings 

 is calculated therefrom. But, unfortunately, some open anthers are 

 always noted (often about 2 %) in Vellai, and we shall have to depend 

 on further study of the seedlings before we can definitely say whether 

 we have been successful. 



A second way of approaching the problem indirectly is opened for 

 us by the fact, now observed for two years, that certain varieties produce 



