CH. IV] CROPS AND METHODS OF PEAS ANT. AGRICULTURE 165 



want of selection. If even educated Europeans with special 

 taste for gardening cannot be induced to select seed from the 

 best parents, we cannot expect that the villager will do so. If 

 improvement of the quality of local strains is to be made by 

 selection, then some such method as that which is, we believe, 

 in force in some places must be used. Seed-parent plants are 

 marked by Government inspectors, and their seed separately 

 collected by the Government, and paid for. The cultivators 

 have to save their own seed and exchange it against the 

 Government supply. In this way only good seed is used. 



Another way of attaining the same end would be the 

 establishment of definite seed gardens at suitable places, where 

 the breeding of improved varieties could be carried on, and 

 seed produced in large quantities. These seeds, whose value 

 would then be accurately known, could be exchanged against 

 the villagers' seed, and the latter sold in the ordinary market 

 in partial repayment of the expense incurred. Without some 

 such system as this, carried out over long periods upon a 

 definite plan, it is idle to expect any improvement in village crops 

 so far as varieties grown are concerned, except in cases where 

 one can introduce a " fixed " strain, such as many " gardeners' 

 varieties " represent, which will not seriously deteriorate under 

 village methods of cultivation, being independent of selection 

 for the retention of its peculiar excellences, though selection 

 will always improve these. It is well, however, to point out 

 that such varieties are rare in field crops. 



In the last few years enormous advances have been made 

 in the science of plant breeding, and by the careful application 

 of the principles first enunciated by Mendel we are able to 

 produce more certain results in much less time. 



Breeding has been but little practised in the tropics. While 

 the cultivated plants of the colder zones are often incapable 

 of surviving if left to themselves in competition with the wild 

 vegetation, this is but rarely the case in the tropics. 



There is every reason to hope that by skilled scientific 

 work great progress may be made in the improvement of 

 tropical crops, and that within a reasonable time. But there 

 are many difficulties in the way. One, for instance, is that 



