98 MI* HIGAN ACADEMY OF SCIENCE. 



from .these plants was sown in rows one foot apart and the grains six 

 inches apart in the row. Each row was the progeny of a single plant. 

 At maturity the plants within a row were quite uniform in character but 

 a comparison of the rows showed very interesting results. There was a 

 variation between the rows of oats of about one week in time of ripening 

 and a marked variation in the character of the straw. There was an 

 interesting variation in the quality of the grain of the wheat both as 

 to plumpness and as to glutinous character. The yields of the several 

 rows of oats ranged from 40.3 bushels to 72.3 bushels per acre. 



When we take into consideration that the oats were grown on very 

 poor land and that they were sown very thin the results lead us to expect 

 some good results in the future, and when we consider further that in- 

 dividual plants were produced that yielded as high as 200 bushels per 

 acre, computing rows one foot apart and a plant every six inches in the 

 row it is hard to conjecture to just what extent the development may be 

 carried. The best plants from each of these rows were selected to furnish 

 seed for the 1902 plots. By this carefully selecting each year it is hoped 

 to develop some greatly improved strains. 



The certainty of improvement by careful selection has been well estab- 

 lished by seedsmen and experiment station men. The Illinois Station, 

 working with corn has shown that the size and height of stalk, width of 

 leaves, length of ears, size and amount of husks and even the chemical 

 composition of the grain may be changed almost at will by careful selec- 

 tion. Various seedsmen have developed varieties innumerable, particu- 

 larly of garden plants, by careful selection for certain characteristics. 



Provided the general grower is careful to fix his type of plant in mfnd 

 and selects uniformly for that type, the question, then, is not as to the 

 certainty of improvement but as to the economy of the practice. Doubt- 

 less the small grower would not be repaid from his few acres for the time 

 thus spent in careful selection. Just how small a business would war- 

 rant such method of selection is a question. Possibly the large grower 

 might find profit in some such system as is practiced by the Sea Island 

 cotton growers. A person may find growing the same season on one of 

 their plantations; 1. A small plot, which we may call plot number one, 

 of plants from seed of individuals selected from a similar plot the 

 preceding year; 2. Another and larger plot, number two, the seed for 

 which was the whole product of plot number one the preceding year and 

 the product of which will furnish seed for the succeeding general crop; 

 and 3. The general crop, the seed for which was the whole product of plot 

 number two the previous year. Thus the third season from the growth 

 of the selected individual the general crop is produced. The process 

 seems like a tedious and laborious one, bTit the general grower does not 

 need to keep records for publication which require much extra time 

 making exact measurements, getting exact weights, and saving every 

 grain. The extra labor would be required principally on the plot number 

 one in sowing carefully the seed from each ])Iant in separate rows and 

 then selecting and threshing the best individual plants at harvest. If a 

 man growing fifty acres of wheat yearly, could increase his average yield 

 two bushels per acre by such selection it would mean an added one hun- 

 dred bushels and that would j)ay for quite a little time spent in selecting. 



While the selection of individual ])lants is the most rapid and certain 

 method of developing an improved strain, the small farmer must adopt 

 some simple means of selecting his seed in bulk. Careful study and 



