INHERITANCE— I M PRO VEMENT B Y BREEDING 2 1 9 



taken from the biggest ears from the bulk crop. Big ears are CHAP. 

 more likely to reproduce their kind than big seed, and this is V ' 

 better than no selection ; but the farmer soon finds that it does 

 not bring him beyond a certain point. This is because the 

 ears in the ordinary bulk field have been cross-bred with inferior 

 strains. In an ordinary commercial field of maize the pro- 

 portion of good, typical ears is very small. Not long ago the 

 writer went through a South African field of what looked like 

 a good maize crop, running probably ten bags to the acre. In 

 the course of about an hour he could find only two plants, in 

 the best parts of the field of 1 5 acres, which could be con- 

 sidered ideal for seed purposes. And though this was much 

 superior to the ordinary crop of the country, he could not find 

 100 plants in an acre that were worth picking for seed. An 

 acre carried over 8,000 plants, and at this rate there would be 

 80 poor plants to every good one. As the maize plant is 

 usually not self-pollinated, but depends on cross pollination, it 

 is probable that every plant in that field was crossed with 

 pollen from one or more of the many poor plants with which 

 it was surrounded. Deterioration in the quality of the seed 

 produced must inevitably have followed. Such deterioration 

 might not show in the ear produced that year, but it would 

 show in the following crop. Deterioration is constantly taking 

 place where breeding is not practised. 



It is not only big ears that produce heavy yields. A big 

 ear is better than a little one, but big ears often produce 

 small, light grains ; medium ears with deep heavy grains 

 usually produce the heaviest yield per acre. 



In practical plant breeding, three processes are usually 

 followed in the selection of maize for the breeding plot : — 



(a) Field-selection of original mother plants ; 



(b) Selection of ears in the seed-store ; 



(c) Selection by continuous performance-record in the 

 breeding plot. 



177. Importance of Care in Selection. — Selection should not 

 be done carelessly, nor be left to the ignorant. Hand picking 

 by Kaffirs may be better than no selection, but it will not carry 

 forward the work of improvement. One season's careless 

 handling of the seed crop may undo all the good which has 

 been accomplished in three or four years. The man who is 



