Mi HOW CEOPS GEOW. 



per cent, both of the. large and medium sizes, and 30 per 

 cent of the small ones. 



The total produce from the small seeds was less abun- 

 dant in all respects than that of the medium, and this 

 less than that of the large seeds. 



Calculated upon the same number of plants, the differ- 

 ences, though less in degree, are still very decided : 



Lehmann, in another experiment, found that from the 

 same weight of seed a larger crop is given by large seed 

 than by small, although the number of plants may be 

 considerably less. 



From the same weight (188 gm.) of seed-peas were 

 produced : 



Number of Weight of Kernels 



Seed-peas. Plants. per 92 sq.ft. Per lOOplants. 

 By small seed, 780 680 1690 234 



By medium seed, 630 605 2224 440 



By large seed, 384 360 2307 640 



DriesdorfE sowed separately, on the same land, winter 

 wheat, as winnowed, and the same divided by sifting into 

 three sizes. In April and May the vegetation from the 

 largest seed was evidently in advance, and at harvest 

 the relative yield for 100 of unsifted seed was 121 from 

 the largest, 105' for, the medium, and 95 for the smallest 

 seed. 



Improved varieties are often the result of continued 

 breeding from the heaviest or largest seeds, accompanied 

 by high culture on rich soil, and thin planting, so that 

 the roots have abundant earth for unhindered develop- 

 ment. 



Hallet, in 1857, selected two ears of Nursery Wheat, 

 " the finest quality of red wheat grown in England," con- 

 taining, togethei-, 87 grains, and planted the kernels 12 

 inches apart every way. At harvest one prime grain 



