CONDITIONS AFFECTING LEGUME INOCULATION. 79 



extract of uiitrinitt'cl soil i;avt' a growth of the alfalfa bacteria w liich 

 might be de.sionated ""fair,'' while '•.vith lime, with lime and humus, 

 and with hunuis alone the extracts o-ave uniformlv a orowth which 

 miji^ht be designated "excellent." In the pots, however, no nodules 

 developed in the humus series. The sterilization b\- heat of the 

 extract from a soil containing such a relativel}' large quantity' of 

 humus, one-fourth by volume, pro))ably <'hanges the material so that 

 the solution becomes favorable to the growth of the bacteria, but this 

 material in the soil pots, not heated, remains imchanged and may in 

 this form be unfavorable to the activity of the bacteria introduced 

 into the soil. 



Aside from the effect of the soil on nodule formation there seems 

 to exist a marked difference among different \ aiieties of legumes in 

 their susceptibility to infection — that is, their readiness to form 

 nodules. With a similar soil and one which is favorable to the 

 growth of the respective species, nodules will occur in abundance upon 

 one species or variety and another species or variety will exhil)it none 

 or only a few. This is particularly noticed with varieties of so}^ beans 

 and with certain species of Phaseolus, the Lima bean {PJiaseoI us 

 limatufi) and some varieties of sov beans being A'erv diiiicult to inocu- 

 late. It would seem that such species were actually resistant to 

 infection, especially in soils rich in nitrogenous matter. 



EFFECT OF HEAVY INOCULATION. 



The decided advantage of vciy heavy applications of pure cultures 

 reported from some field and pot experiments and reports of other 

 experiments showing that cultures diluted to almost infinitesimal limits 

 gave as good results as undiluted cultures seem at first diametrically 

 opposed. Greenhouse tests seem to confirm the beneficial effect to be 

 expected from heavy inoculation provided excessive cpiantities of cul- 

 ture are applied. The following experiment is fairly representative: 



Effect of light and heavy inoculation of garden peas on soil from Bine Hill, Me. 



It is here shown that heav}" inoculation not onl}^ doubled the average 

 number of nodules upon inoculated plants, but insured all plants 

 becoming infected; while with light inoculation in this unfavorable 

 soil one-third of the plants failed to produce nodules, and with no 

 inoculation nodules were almost entirelv lacking. 



100— VIII 



