288 FORMATION OF SLIME OR GUM BY RHIZOBIUM LEGUMINOSARUM, 



from dextrose and asparagin. Race "d," which produces bac- 

 teroidal forms freely, differs from all other races that I have 

 examined in being injuriously affected by the presence of alkalies. 

 Without the addition of alkali, in the experiment with different 

 forms of nitrogen, it produced 18 % of slime in a week, while with 

 the alkali only 3 % was formed in a fortnight. The favourable 

 influence of acidity explains the high yield with ammonium 

 phosphate, which soon produces an acid condition of the medium. 

 It also accounts for the low yields with the different sugars. 

 The two experiments were repeated with " d," but sodium car- 

 bonate was not added before pouring the plates; the results are 

 included in the tables under the heading (d, neutral). The stem 

 race produced no slime during the first week at 22°, but during 

 the second week while growing at 26 c , a decided although small 

 quantity of slime appeared on the plates. 



The discovery of a race capable of producing a maximum 

 amount of slime on an acid medium is of considerable importance, 

 for such a race would form slime much more readily in the 

 faintly acid nodular tissue than other races which thrive best in 

 slightly alkaline media. It was previously deduced that the 

 acidity of the nodular sap occasioned the formation of the 

 " infection threads " which stretch from cell to cell, and it 

 occurred to me that this was an excellent race with which to 

 test the point. Accordingly flasks containing 50 c.c. portions of 

 a medium containing dextrose 2 %, asparagin 006 %, and sodium 

 phosphate - 2 % were treated with increasing quantities of phos- 

 phoric and citric acids, and seeded with this acidophile race. 

 Growth occurred in the flasks containing - 033 % phosphoric and 

 0-14 % citric acids and in those containing smaller quantities. 

 No growth took place in the flasks with larger percentages. At 

 the end of a week, infection threads were found in the cultures 

 and especially in that containing 00165 % of phosphoric acid. 

 All the cultures contained bacteroids in considerable quantity, 

 but they were most numerous with the limiting amounts of acid. 

 The infection threads were noted upon examination at the end 

 of seven days, and upon a second examination, at the end of a 



