166 AGRICULTURE IN THE TROPICS [PT. Ill 



so many crops are perennials, and do not generally flower for 

 some years. Not only, therefore, is there a long time between 

 each two generations, but the cultivator will not root out 

 established plants to make room for better oues. 



If this line of work is to become, as all present indications 

 point to its becoming, one of the most important duties of a 

 tropical agricultural department, it will obviously be necessary 

 to have proper seed-producing gardens, fully equipped with 

 trained breeders and all necessary fittings, from which seed 

 can be supplied to the cultivators. And this the more especially 

 at first, because the improved kinds will cross with the inferior 

 kinds cultivated round about, and the resulting seed be con- 

 sequently inferior. 



In many cases, of course, it is possible that varieties better 

 suited to local conditions may be obtained from other countries, 

 and this is especially probable in a country where local agri- 

 culture is backward. This sort of work is of course one which 

 should be attended to in Experimental Gardens, but the trials 

 should be on a commercial scale, should be thoroughly thought 

 out in advance, and should deal only with one or two things at 

 a time. Local varieties should of course be tested against the 

 introduced ones, under as many conditions as possible. Once 

 the Experimental Garden has shown the probability that a 

 certain variety, whether introduced from abroad or bred in 

 the garden, is apparently an improvement upon the local ones, 

 further tests as to its suitability may be made through the 

 school gardens, and if it pass successfully through these, it may 

 then be safely recommended to the villagers. 



It is of very great importance in dealing with eastern 

 natives to be quite sure of the result before recommending 

 any action to them, especially in matters of agriculture. Even 

 when we have shown that a particular variety of some locally 

 cultivated crop is superior in quality and yield to the native 

 forms, it may not be easy to get the villager to adopt it, 

 prejudices or customs standing in the way. An amusing 

 illustration has lately been afforded in Ceylon. The variety 

 of Indian corn locally grown is what is technically known as 

 a "flint" corn, with rounded ends to the grains. This was 



