598 BACTERIA IN RELATION TO PLANT DISEASES 



when transferred from agar to agar, but that upon culture on agar after growth in 

 peptone bouillon they have a thin, wrinkled, dull surface. Furthermore, action of 

 cold, sodium chloride, or acids induces ready conversion into club-shaped, Y-shaped, 

 or variously branched involution forms. Furthermore, these involution forms are re- 

 ported to occur in tumors. In the bacteria of legume nodules, involution or degenera- 

 tion colonies appear after a time in the galls of legumes in which a balance seems to 

 have been established between parasite and host ; also in culture after a lapse of time. 



Gardner and Kendrick,' in describing what appears to be a typical rough (R) 

 colony of B. viridifaciens, point out that drying out of the agar substrate accentuates 

 ridging and piling up of the colony. Whether this is due to agar desiccation has not 

 been determined. 



Motility of cells of smooth and rough colonies. — The observation of Sharps that 

 roughness of Bact. phaseoli sojense is correlated with low motility or no motility of the 

 organism, while smoothness is correlated directly with high motility, is in accord with 

 many observations on bacteria other than plant pathogens. 



Virulence. — In consonance with the findings in the animal field, Sharp' observed 

 a correlation of virulence with culture forms. He found that rough colonies of Bact. 

 phaseoli sojense when inoculated into soy-bean leaves by needle pricks are less virulent, 

 as measured by percentage of takes and by size of lesions in a definite time, than the 

 smooth ones. He found further that passage of the rough strain through the plant 

 caused a partial reversion to the smooth type, and that the new smooth forms were 

 as virulent as the original smooth ones. 



Smith has taken pathogenicity for the proper host and production of character- 

 istic symptoms as one of the chief criteria in determining the identity of bacterial 

 pathogens. For this reason, certain statements by him^ are significant in this place. 

 He cautions workers against using any but smooth translucent colonies of Bact. tume- 

 faciens in making isolations from galls, or in conducting infection tests. In this con- 

 nection the finding of Link and HuU^ for their strain of Bact. tumefaciens is significant. 

 When agar-shake dilutions were made from galls or from agar colonies, this strain 

 produced colonies of which about one half were smooth and the other half rough, or 

 partly rough and smooth. Later all colonies became entirely smooth. Apparently 

 both types are pathogenic. 



One of the criteria of the virulence of an organism used by animal bacteriologists 

 is the minimal number of organisms of a given species in a definite state of culture 

 which, by a definite route, will produce infection or death of a definite host, in a def-| 

 inite period of time. Practically no studies using a comparable technique have been 

 made by phytobacteriologists. Smith'' in 191 1 called attention to the observation that 

 often a large amount of bacterial inoculum will produce infection when a smaller 

 amount will fail. He suggested that this phenomenon of mass action should be in-J 



' Gardner, M. W., and Kendrick, J. B.: loc. cit. 



* Sharp, C. G.: loc. cit. 



3 Ihid. 



■• Smith, E. F. : Bacterial Diseases of Plants. 



sLink, George K. K., and Hull, K.: loc. cit. 



^ Smith, E. F.: Bacteria in Relation to Plant Diseases. 



