174 ROYAL SOCIETY OF CANADA 



minosse (see media). After three days the seeds were kept in the diffuse 

 daylight of the room, and aiter five days or more the seeds which had 

 germinated and which were found free from bacteria were planted in 

 the flasks of ash-maltose-agar, prepared as elsewhere described, the plug 

 was removed from the test-tube, the mouth of the test-tube was flamed, 

 and the liquid, if any, was poured away. The plug was then removed 

 an instant from the mouth of the flask containing the ash-maltose-agar 

 and the germinating seed was shaken from the test-tube into the flask, 

 and the plug was re-inserted into the mouth of the flask. The parch- 

 ment paper was tied in place over the cotton plug and the flask weighed 

 and set aside in diffuse daylight in the laboratory. A number of fla.sk 

 cultures were prepared in this way on the same day and were kept some 

 days before any were inoculated in order to observe possible contamina- 

 tions. Some of the flasks were then inoculated with a pure culture of 

 Ps. radicicola and some were kept uninoculated as controls. The flasks 

 were usually kept in a greenhouse. At intervals the flasks were weighed 

 to determine the loss of water by evaporation. The growth of the 

 plant, both root and stem, was observed and measured through the glass 

 and through the transparent medium. The presence or absence of con- 

 taminating fimgi and bacteria was noted. The growth of Ps. radicicola 

 in the inoculated flasks, the first indication o£ its invasion of the living 

 roots, the form and growth of nodules ajid the results on the root system 

 were studied and recorded daily or at convenient intervals. 



One series of flasks was kept under observation for eight months 

 and three other series for shorter periods. A final examination of the 

 flasks was then made, as follows : — 



The medium was examined for micro-organisms by staining and 

 by plate-cultures in ordinary agar and gelatin and in ash-maltose-agar. 

 A nodule, if any were present, was taken from the roots with sterile 

 forceps and plate cultures in ash-maltose-agar were made from it in 

 the manner already described (see isolation of Ps. radicicola from no- 

 dules). Stains were also made from this nodule and from others. 

 The roots were examined by staining, and when a general invasion was 

 observed, cultures were also made from the interiors of the roots so 

 invaded. Control cultures in the flasks not inoculated were treated in 

 a corresponding manner, stains and plate cultures being made from the 

 medium and from the roots. The plant was then drawn out entire 

 from the flask and the roots and stems were measured and sometimes 

 weighed. In some cases nitrogen determinations were made from the 

 medium and from the plants. The nodules were counted and weighed. 



The loss of water by evaporation was found not to var}^ much in 

 flasks of the same series. The average loss of seven flasks groAvn dur- 



