INHERITANCE— IMPRO 1 'EMENT B ) ' BREEDING 1 3 7 



fact that 8 o/.s. of grain per ear is a common return, and CHAP, 

 that ears yielding 15 ozs. of grain are not unknown, an v - 

 average of 3^ ozs. is absurdly low, and suggests great possi- 

 bilities for improvement. 



"Nubbins" such as are shown in Fig. 62, must tend to 

 bring down the yield enormously. 



103. Average Weight of Grain per Ear. — Such variations 

 in weight of ear indicate that the composition of the average 

 lot of seed-maize is very unsatisfactory. A difference of 

 12 ozs. between the extremes is much too great; no stock- 

 man would breed 800 lbs. bullocks when he could produce 

 2,000 lbs. animals with the same amount of care and feed. 

 Although it may not be possible to bring the average up to 

 the maximum, still by breeding, the average can be greatly 

 improved, and the production of the 2\ to 4 ozs. nubbins (Fig. 

 62) which are so plentiful in South African crops at present, 

 can be minimized. The following tables (XIX to XXIX) show 

 the weights and yields of grain of 1 ,684 ears of maize grown 

 at Vereeniging, Transvaal, in 1908, and selected for seed. 

 They were carefully studied and weighed, at the end of the 

 dry season when the grain was as dry as possible. 



104. Need for Increase in the Weight of Grain per Ear, — 

 Allowing for losses to the extent of 25 per cent from external 

 causes, misses, etc., we should still have 6,534 bearing plants 

 to an acre, an average of one to every 7-i- square feet of soil 

 surface. As is shown in the preceding tables, the average 

 weight of grain from a medium-sized ear of Golden King in 

 1908 was 7'397 ozs. ; this was determined by weighing 250 

 ears which had been picked for seed, but from which all the 

 best ears had been removed for use on the breeding plot, 

 leaving 150 medium-sized ears. On well-farmed land, 10 

 muids of shelled grain per acre is not an uncommon average 

 yield without manure. To produce 10 muids would require 

 only 4,326 plants yielding 7 '379 ozs. each, or less than two- 

 thirds of the plants left on the acre. How are we to account 

 for the remaining 2,208 plants? At 7'379 ozs. each they 

 should produce an additional 1,020 lbs., or 5 muids of grain, 

 which would increase the yield per acre by 50 per cent. 

 There are three possible explanations: (1) that the average 

 yield of grain per ear is much lower than 7'379 ozs - ; ( 2 ) that 



