292 



[October, 1912. 



wet-ploughing, The result is attributed to the alteration in "Bacterial 

 content" of the soil, the wet cultivation helping the beneficial bacteria. 



Breeding Improved Races. 



The last method I have referred to is the breeding of improved races 

 of seed possessing the best qualities as regards yield, disease-resistance, 

 suitability to special localities — high or low, dry or wet, &c. This is work 

 for the scientist and not for the general cultivator, and it is most grati- 

 fying to find from a note published in the Tropical Agriculturist that 

 this matter is now receiving the attention of the scientific staff of the 

 Agricultural Department. 



SUPPLEMENTARY NOTE BY DR. R. H. LOCK. 



The following supplementary note by Dr. R. H, Lock, Assistant 

 Director of Botanic Gardens, was read by Mr. M. Kelway Bamber : — 



I have been asked, as a student of plant improvement, to make a few 

 remarks on the subject of the paper which has just been read. Mr. de 

 Silva has properly pointed out that seed selection is an operation which 

 requires only a small expenditure of capital and energy if only the method 

 of doing it is understood. Mr. de Silva has described four important 

 methods of seed-selection. To anyone who has followed the develop- 

 ments of scientific agriculture during the last twelve years it is a truism 

 that these methods, important as they are, are of subsidiary value in 

 comparison with a fifth method which Mr. de Silva has not mentioned. 

 This method is tbe selection of seed from the best plants regarded as seed 

 parents. Mr. de Silva states that increase of crop by as much as 36 per 

 cent, can be got by simply sorting the grain by a mechanical method. I 

 have no personal experience of this method, but it is clear from general 

 principles that any improvement obtained in such a way can only be 

 temporary. The method must be repeated in every generation with 

 every bushel of paddy used for seed. As regards change of seed the 

 method requires to be tried experimentally with rice in Ceylon before 

 it can definitely be said to apply to rice in Ceylon. It is, of course, highly 

 important to choose a variety which is suitable for the particular soil and 

 district in which it is to be grown. To a considerable extent this is doubt- 

 less already done. The breeding of improved races of the kind referred 

 to by Mr. de Silva is an elaborate and uncertain process, and, considering 

 the large number of valuable races already existing in Ceylon, little 

 stress need be laid on this until other methods of improvement have 

 been exhausted. 



Seed Parent Selection. 



The fifth method, that of seed parent selectiou, although based 

 on the scientific knowledge, is not a scientific process in the sense that 

 any special training is required in order to carry it out. The process 

 is perfectly simple and not very laborious. I have myself, by an 

 expenditure of time amounting to not more than six working days of 

 nine hours each, obtained in two generations in a transplanted crop 

 an improvement in yield estimated at 50 per cent. This strain is 

 now being grown for seed, and in two more generations enough can 

 be raised to sow a whole province. The Essence of the Process Employed 

 is tc select from among a number of the best plants those whose seed 

 when sown gives rise to the largest crop. For this purpose all that is 

 necessary is to sow the seed from each plaat separately, and, after traas« 



