EXPERIMENT STATION BULLETINS. 



309 



TABLE II.— PERCENTAGE RELATION BETWEEN THE OLD (81302) AND THE NEW 



ROBUST (40511). 



The Nelson strain is not the original Xo. SI 302 bnt came from a later 

 selection which appears to be similar to 40511. The resnlts given in 

 Tal)le II are onlv the data from one year. The resnlts obtained are so 

 similar tliat one wonld liave great confidence in tlie average. Statistical 

 methods however shoTV the probable error is only 1.5 percent. This 

 informs us that true average yield lies between 86.2 percent and 89.2 

 l)ei- cent. If these are resnlts from the same strain of beans, the indi- 

 vidual determination should be within the range (8G.2 and 89.2). How- 

 ever four of the five results lie outside of these extremes. This leads us to 

 believe that these strains are not alike but possess different yielding 

 power, although origiualh" from Eobust No. 81302. 



Again, in 1918 individual plant selections were made from Robust 

 (No. 81302) fields in Genesee, Oakland, Oceana, and Jackson counties. 

 These selections were tested in the 1919 plant row series. After the 

 results were obtained over one-half of the Robust plant rows were dis- 

 carded as too poor to continue in the breeding work at the Michigan Ex- 

 periment Station. Twenty-seven of the selection tested in 1919 were 

 continued in the 1920 plats. Seventeen of the selections tested in 1920 

 were from Genesee county. The average yield of these 17 selections 

 was 71.34i.2.S percent of the yield of the 10511 strain. One of these 

 selectioms yielded 49.7 percent while more than one-half of them yielded 

 less than 68 percent and only one selection yielded as much as our check 

 (No. 40511). The other selections from fanners' fields gave similar 

 results. This shows the Robust bean in tlie hands of farmers lacks ani- 

 formity, indicating that a mixture has occurred somewhere along tlie 

 line, and that this mixture varies greatly in producing power. 



To sum up, investigation shows that many of the varieties termed 

 "Robust beans" in the hands of farmers are no better than the common 

 commercial varieties grown throughout the country. This is due to two 

 causes : first, the original strain of 81302 was not as good as our present 

 strain. Second, the mixture of other varieties has lowered the yield of 

 the 81302 strain. Growers with these strains of inferior value should 

 change their seed for highly productive Robust of unmistaken origin. 

 The Farm Crops Department will furnish information in regard to safe 

 sources of seed of pure Robust. The results of the 1917-20 varietal 

 series show that a new strain of Robust No. 40520 is yielding 13.4 percent 

 above the check (No. 40511) and one and a lialf times the yield that 

 has been obtained from Early Wonder. The No. 40520 strain will be 



