^2 STATE BOARD OF AGRICULTURE. 



in the Lower Peninsula. Nearly every neighborhood has its favored 

 varieties, and many people have suffered by discartling an old variety' before 

 they Avere sure that something better was in sight. It is best to test the 

 new variety along side the old and make a change only when sure that the 

 new variety is superior. 



The Michigan Agricultural College has been breeding oats during the 

 past twenty years, and in recent years some of the improved strains have 

 been distributed through the Michigan Crop Improvement Association. 

 It is desired that the farmers of Michigan will plant these varieties along 

 side of their own and be the judge as to relative jdelds. In the end, there 

 may be no variety superior for all conditions, but if each farmer can have 

 the ])cst variety for his own conditions, the greatest good is attained. 



In order to draw reliable conclusions in regard to the productiveness 

 of a variety, it is necessary to try it side by side with a standard variety 

 under identical conditions. Because of the variability of soil, cultural, and 

 climatical conditions, the bushels per acre ai'e deceiving, unless the yield 

 of the standard variety is also given. To express this in one figure the 

 percentage method is valuable. 



THE EARLY WORK. 



The Alexander and College Wonder are two of the strains that origi- 

 nated in the work of Prof. J. A. Jeffrey who made sixteen individual plant 

 selections from an old variety known as the American Banner in 1900, 

 and so planted and selected them that the lines passed through an indivi- 

 dual plant annually until 1906. B}' that time nine of the original sixteen 

 races had been discarded as inferior strains. However, most of these lines 

 had been allowed to branch, each branch being continued through an 

 individual plant annuallj^ One of the lines contained six such branches. 

 In all nineteen strains of American Banner were handed to the writer in 

 1906. Similar selections were made from Big 4 in 1900, Init all but two of 

 these lines were dropped before 1906. These proved inferior to the Banner 

 selections during 1906-1908. 



In 1907, the writer planted Prof. Jeffrej^'s nineteen individual oat plants 

 in separate plats for comparison, and was convinced of the similarity if 

 not the identity of the lines, originating from a single plant in 1900. 

 This series was again run in 1908 with similar results. B}^ 1909 represen- 

 lalives of the various strains had Ijeen increased enough to plant in a 1-20 

 acre series, where the oat then grown on the College farm was used as 

 a standard or check. The plats were six feet wide and about 360 feet 

 long, with checks each fifth plat in the series. One of these strains was in- 

 troduced to farmers under the name of Alexander in 1911, and a second 

 was introduced in 1914. The latter is known as the College Wonder. 

 During the time that this work was being done, it was believed that oats 

 may be continually improved by selections within a pure line.^ In fact, 

 the stability of a pure line was not generally known by most agroano- 

 mists until 1912. The theory of the day was also practiced by the writer 

 until 1910 when the results of testing Prof. Jeffery's oats had shown the 

 fact that selections within pure lines are without i-esults. . Also, during 

 these years, it was found that when the progeny of ditlercnt individual 



- It was (luring these years that Johannsen of DcDmark was doing his epoch making work 

 on beans, which has now become renowned as "Joliannsen's Nineteen 15eans." 



