54 EEPORT OF THE COMMISSIONER OF AGRICULTURE. 



IMPROTEMENT TJY SELECTION. 



Tlie selGCtio7i of the best for seed, or tlio rejoctioti oftlie i)oorest, has 

 alwayK been, 8(ii<;fly s))oakin<j^, tl)(3 only nietlind uf iiiiproA'iii}! varielics. 

 Thecarcfnl Hclectioii olseed was recoiniu ended in early times. Coliiniclla 

 and Celsius :ind V^irjiil .sj^'ui; of it. Injprovenient by weleetion is well 

 known in the abstract, bnt it cannot be loo often brought forward an<l 

 urged anew. 



Some seed-wheat of a clioice kind, wcigliing 58 ])onnds per bnishel, 

 was improved by careful selection and treatment nnti! an entire cn)i>, in 

 a few years, averaged G3 pounds per bushel. Every year tlie best tieid 

 was chosen for seed, the seed-whent extra cleaned by funning-mill to 

 blowout the lighter grains, then hand-sifted to get out the simdler 

 grains, then the imperfect grains skimmed from the surface of the strong 

 brine in which it was put as a preventive of smut ; thus only the largest, 

 plumpest, heaviest grains were sown. That the soil alone did not pro- 

 duce this change was shown by the tact that it took several years to 

 bring the improvement about; and that then the improvement was 

 shown when used on other farms. 



Dr. E. L. Sturtcvnnt, director of the New York exj^erimental station, 

 in speaking of the i'mportance of the carefid selection and change of 

 seed, says : 



Improve flus cliaraclor of the seed supply, eitlicr tlirongli selection or by cliaiige of 

 seed. At lb.0 New York statiou, in IbS'), one variety of oats averaged over 'JO bushels 

 per acre more than another, and frequent illustrations could be given of Iho doubling 

 of crops as between varieties, one better adapted to our conditions than another, or 

 else from diflerences inherent in the variety. The prol)lem for experinienlal action 

 can then l>e the question as to what constitutes thi.^ dinereiice of potency in the seed, 

 and in what way and to what extent can it be brought under our control, and wliat 

 are its relations to clinjate, fertility, and culture. 



DIFFERENCES IN THE SEEDS OF THE SA5IE PLANT. 



]\rr. Frederic F. Ilallett, of Brighton, some thirty years ago began 

 systematic experiments on improving wheat. A single head, chosen for 

 fine quality, irrespective of size or vigor, was 4f inch-es long and had 

 47 grains. ' These grains were carefully planted in rows, one seed in a 

 idace, 12 inches ajjart each way. Ar, harvest the plants were com- 

 pared, the best head of this best plant planted the next year in the same 

 way, and so on year after year, each year choosing the best head from 

 the most prolific plant. The tirst year the best plant bore 10 heads, 

 the second year 22, the third year '6d, and the fourth year 52, the best 

 head of which was 8^ inches long, and bore 123 grains. This was the 

 origin of the famous " Pedigree " wheat. He gave the name "Pedi- 

 gree" because the })rocess of improvement by selection was precisely 

 analogous to that of breeding animals to points and strengthening the 

 hcredTtv of the good points by " pedigree." His riper conclusions, given 

 later (iii the Trans. Brit. Assoc. Adv. Sci., ISCO, p. 113), are that every 

 fully-developed plant, whether of wheat, oats, or barley, has one ear 

 superior in reproductive power to any of the others on the plant; that 

 every such plant has one grain more productive than any other, and 

 that this best grain grows on the best ear; that the superior vigor of 

 this grain is transmissible to its progeny; that by selection thissupernjr- 

 ity accumulates; that the improvement is at first very rapid, but that 

 in successive years it gradually grou's less: that an improved type is 

 the result, and that by careful selection the improvement can be kept 

 np. Experiments conducted by iJr, Gustavo Marck at the experiment 



