New YorK AGRICULTURAL EXPERIMENT STATION. 45 
From the above table it will be seen that among the thirty-two 
experiments cniy eight of the check plats produced enough hay to 
inake them a success while twenty-two of the adjoining plats which 
had been inoculated with soil from our alfalfa field produced 
enough hay to entitle them to be so classed. Since the two kinds 
of plats were in all cases in adjoining pairs and these pairs were 
treated alike except that inoculated soil was scattered over one of 
each pair of plats this difference in production can be correctly as- 
cribed to the influence of the inoculating soil. 
It should be remembered that the land for these experiments was 
selected by the farmers with little or no guidance in the matter 
and was fitted and sown with only very general directions. Some 
of the fields were not adapted to alfalfa on account of the poor 
drainage, while others were put in so hastily that the weeds over- 
ran and choked out the alfalfa. This result was often produced 
by a liberal application of stable manure which, while it added 
fertility, also added weed seed in abundance. 
Profiting by the mistakes which were made in 1905 the experi- 
menters in 1906 were urged to select their land for alfalfa with 
care as to its drainage and to fit it with a view to reducing its con- 
teut of weed seed as much as possible. Of the 67 farmers who were 
furnished with soil 32 conducted their experiments in such a man- 
ner as to give dependable results which are summarized in Table 
Ith 
