338 



infusions of tlie otlier soils. In preparing the pure mltnres of bacteria 

 from tuljerdes the tuberchis were tirst washed with a weak scdutiou 

 of corrosive sublimate to destroy any germs on the surface, cut oi)eu 

 with a sterilized knife, and a small amount of the tissue taken out with 

 a i)latinum Avire. At first only young tubercles were used, but later full- 

 gro\yn ones were employed with equally good results. In the inocula- 

 tion the soil immediately surrounding each ]»lant received 7 e. e. (»f the 

 decoction. 



First scries of e.rperimenis. — In this series, ])eas, IJoljinia, lal)urnum, 

 ami (lleditschia were used. The jihints of each genus were grown in 

 sterilized soil, being inoculated in separate cases with soil infusions of 

 lu])ine, peas, Gleditschia, Eobinia, and labui'num soils, and with juire 

 cidtures of bacteria from ])ea and from Kobinia tubercles, and in steril- 

 ized soil with a dressing of <-alcium nitrate or ammonium suli»liate. (tr 

 without the addition of nitrogen. 



Tlie results of this series <)f experiinents sh<»w. as suggested liy Ib'll- 

 riegel, that the extracts of ditlei-ent soils are (piiti" different in their 

 a(;tion on ditt'erent legumintius ])lants, and the authors lM'lie\ «■ that tliese 

 differences can not be account«*d for, as Frank* suggested, by the dif- 

 ference in the numbei- of l>a<'t<'iia ]»r<'sent. In the case of (ileditschia, 

 no symptoms of nitrogen starvation were noticed for oxer 1' months. 

 The |»hiiits grew, altiiough slowly, and luoilnced on an average altout 

 four timesasmuch dry matter asthe (tiiginal seeds contaim-d. and in some 

 cases a noticeable increase of nitrogen. No tubercle's were formed on 

 any of the roots, and no beneficial effects from the inoculation w«'re jiei- 

 cei>tible. 



The Ial)urnum jilaiits did n(»t glow well in the soil mixture used and 

 were harvested early. The indications \\cr<' that plants of tliis genus 

 resjiond slowly to inoculation. 



The first pea ]»lants to n-spond to the inotulation were those iimcu- 

 late<l with |»ea extracts (either fr<»m soil (»r pure cultures), the change 

 in the leaves being per<M'ptible L'O days after the inoculation. The 

 effects of Kobinia soil extract weie fell latest of all; tlic pure cnllnre 

 from Ifobinia tubercles met with an accident. In the i-ase of both 

 IvoViinia and peas ncarl\ all the jdants inoculated protluced root tubercles 

 and the tubei-clcs were confined almost e\clnsivel.\ to tln»se roots near the 

 surface. The jtea ]dants ]»r(Mluce<| a much laigei- numlx-r than the 

 liohinia. The roots of a single jtea itlant (inoculated with (ileditschia 

 soil infusion) juoduced 4,572 mirmal tubenles. 



The efiects of the inoculation of Kobinia were <)l)served fir.st where 

 ]mre cultures from Kobinia tid)ercles hatl been use«l (LM» days after 

 the inoculation) and !(► days later where Kobiina soil infusion had 

 been used. The effects where laburnum soil infusion was used were 

 noticed about the same time, and where (Ileditschia soil infusion was 

 used about 1(» days later, or 20 days after the Kobinia soil infusion. 



* Liiiiilw. .Taljil.., 19 (1800), p. 618. 



