J oiirnal of Agriculture. 



[lo Feb., 1910 



infecting the roots of the pea, l>ut the\ were either not \ery vigorous' or 

 not ver\ numerous. The tubercles do not form so well under water as 

 in soil, but this is mainly a (juestion of aeration, and during the experi- 

 ment, proper care was taken that thi.s condition was fulfilled. It is 

 curious that the only tubercles should have been formed in the solutions 

 containing nitrates, for it is generally held that, in the soil, the abundant 

 presence of nitrates tends to decrease the formation of root tubercles. 



Although a single experiment is hardly conclusive, it is possible that 

 nitrates do not influence the development of root-tubercle bacteria on roots 

 grown in a culture solution in the same way that thev do in the soil. 



^V<: 



PEAS GROWN IN WATER CULTURE. 



The actual size of the roots as grown in the different solutions is shown 

 in the second illustration. 



In addition to the culture experiments, six small plots were prepared, 

 each 6 feet square, and laid out as in the i)lan on page 102. Care 

 was taken to see that the soil and other conditions were quite uniform, 

 and during the course of the experiments the plots were kept entireh- free 

 from weeds, and the soil open and pervious. The seeds were inoculated 

 before ])lanting, bv mixing the cultures with sterilized skimmed milk and 

 moistening the seeds with this mixture before planting. Care was taken 

 to insure even planting, and in the case of the peas each plot contained 



