Dorothy M. Cayley 469 



gave colonies of different appearance, but later platings showed uniformly 

 pure cultures. Among others Bacillus mycoidesiQ), Bacillus mega- 

 tkeriumil), are instances of organisms non-pathogenic to man, which 

 show considerable colony variability. 



Ps. seminum also shows considerable variability in size and 

 general characteristics of its colonies on artificial media. The fact 

 that the organism produces a thin capsule round the rods when they 

 come to rest after the motile stage, and has its spores embedded in 

 a kind of matrix, may partly account for the variability shown. In 

 liquid media it hangs together in long chains, and as I have already 

 pointed out, this phenomenon of incomplete division can also occur 

 on solid media, and may be a possible explanation of the occurrence 

 of stragghng outgrowths in the submerged, and lobing or radiations 

 in the surface colonies. 



The two forms of colonies most frequently met with in platings of 

 Ps. semmum are (1) surface opaque white or whitey-buff, more or 

 less circular, and (2) surface less opaque, more or less widely radiating 

 colonies, with their corresponding lens-shaped and straggling submerged 

 colonies (Plate VII, fig. 13). Repeated tests as to their reaction to litmus 

 milk showed identical results with both colonies. They also give the 

 same reaction with sugars. There is no doubt that the amount of 

 water in the medium greatly influences the size and length of the lobes 

 or radiations in the spreading colonies. Treatment of the organism 

 previous to plating may also have some effect. If the bacillus is iso- 

 lated from the young green cotyledon or the growing stem, it tends to 

 throw spreading colonies ; whereas more or less circular surface colonies 

 result from young cultures from dry cotyledons incubated in peptone 

 beef broths. If pure cultures of first platings are kept for some time 

 in hquid media, so that a pellicle or flocculent sediment is formed, the 

 resulting colonies then vary considerably from the original colony. 



To give illustrations of this, a tube of alkaline (1 per cent, normal 

 NaOH) peptone beef broth was infected with a surface opaque whitish 

 circular colony from a dry cotyledon, and, replated after five months 

 and twelve days on acid (1 per cent, normal HCl) pea agar agar, gave a 

 straggling surface colony. On the other hand a tube of sterile soil 

 extract was infected with a straggling colony and replated on neutral 

 pea agar agar, and gave the following : 



(1) More or less circular opaque colonies, inclined to be lobed. 



(2) One large semi-transparent spreading colony which almost com- 

 pletely covered the whole surface of the petri dish. 



31—2 



