339 



The infusions of lupine and pea soils and the pure cultures from pea 

 tubercles showed no bcueticial action up to the dose of the exi)eri- 

 nient. It Avas found exceedingly ditlicTilt, althou«ih not altojictlier 

 iiii]»ossib]e, to cultivate Ivoln'nia in sterilized soil, under the ])rccautions 

 iiieiitioiied above, without the formation of tubercles. IMants inocu- 

 lated after th(\v had jiassed a certain stage, altlnrngh they i>rodueed 

 tubercles, faih'd to recover froju the starvation ])eriod, and the author 

 lielieves they were no htnger capable of assimilating nitrogen even 

 when the tubercles were present. 



It was noticed Avhere the inoculation of Kobinia was successful that 

 in general the amount id" dry matter jyroduccd and the p«'reentage of 

 nitrogen in the same were larger than when the plants receivetl a dress- 

 ing of nitrogenous fertilizers instead. Thus, Avhile the ino<ulate<l 

 plants averaged .3.088 per cent of nitrogen, those manured Mith nitio 

 gen averaged only 1.312 i)er cent. 



The results of this series of exi)eriments show, therefore, that in the 

 case of peas and Kobinia each kind of plants responde<l most readily to 

 inoculation with extracts of soils in which a similar kind of plants Inul 

 l»een i)revi<msly grown, /. e. peas to pea soil extract, and Kobinia to 

 Kobinia soil extract. On the other hand the Kobinia soil extract w hen 

 ai>i»lied to i)ea ]»lants was felt latest of all and the pea soil extract failed 

 t(» ]»rodnce any perceptibh' etfect when ajiidied to Ifobinia. The autluns 

 ai-e forced to believe from these results that the infusions of ditferent soils 

 must contain bacteria which in some manner difler from each other. They 

 believe, liowever, that tlu' solution of the (piestion must come through 

 the study of pure cultures, as only indelinite results can be obtained 

 with crude infusions. 



AVhile the ])ure cultures from Ifoliinia tubercles Avere etl'ective on 

 Kobinia plants, the pure cultures from pea tubercles, as Avell as the 

 infusion of pea soil, were entirely without etfect on Kobinia. But the fact 

 that the' pure cultures from i)ca tubercles had almost no etfect on ]»ea 

 ]dants rendered further trials necessary, although the authois suggest 

 that the absence of acticm in the latter case may have been because the 

 inoculation came too late for the rapidly growing pea ]dants. 



tSccond serh'n <{f experiments. — Peas, lu]>iiu\ and beans were used, ]iure 

 cultures being employed almost exclusively. 



Pea plants were inoculated with pure cultures from the tubercles of 

 ]>eas, lui)ine, and Kol)inia, and from extracts of soils in which these 

 jtlants had been grown. The ])laiits inoculated with purecultiucs from 

 either pea soil or tubercles and trom hii»iim tubercles ])roduced i-oot 

 tubercles: the remainder faile<l to pioduce any tubercles. Those inocu- 

 lated with bacteiia from jjcas bore the tubeicles on the roots of the 

 second order; and those Avith cultures from lu])ine tubercles bore tulier- 

 cles on the roots of the third oider, and the tubercles were soim-Avhat 

 diflerent in ai)])earanc<' from those whe"e i»ca liacteria were used. These 

 observation!* on the locality of the tubercles were veriticd in a second 



