31 



abuudaiit, wlille up to tliat time none liad been found 

 on the plots not inoculated. 



Relative to a similar experiment in 1903, lie writes 

 under date of pTune 13, 1903, as follows: "I inoculated 

 all plots except No. 8, which seems to be the poorest of 

 any of thein.'" 



In view of Dr. Hopkins' conclusions, we can now rec- 

 ommend that earth from an old melilotu's field be used 

 for inoculating- alfalfa, where this is decidedly more 

 convenient tlian to use earth from alfalfa or bur clover 

 fields or than the pure culture of the laboratory. 



TIME TO sow. 



The following is a summary of results of sowing al- 

 falfa on the station farm during the eight years tliat 

 work has been under the writer's charge : 



We liave records of fall sowings on ten different 

 dates. In everv case when alfalfa was sown broadcast 

 after November 1, the stand was mined by cold. In one 

 case alfalfa sown as early as October 7, (1901), was al- 

 most completely winter killed. Plants from seed sown 

 as early as September 13 (1900), and as late as October 

 29 (1899) survived the winter, although in other years 

 a considerable proportion of the sowings made in late 

 October resulted disastrously. 



We have records of eight dates of spring sowing of 

 alfalfa on the station farm. These point to the first half 

 of March as better than a later date. 



In our co-operative experiments with alfalfa in 1903- 

 1904, arranged for in nearly every county in the State, 

 fall sowing was made undulv late by drought, and cold 

 weather came on unusually early, and has been unusual- 

 ly continuous. ]Moreover, in most cases there wasi in- 

 sufficient moisture to cause the young plants to grow 



