225 



A few words may be said about each of these conditions, 

 which have been reduced to their simplest proportions. 



In Case I a population consists of two species of which one is 

 numerically dominant owing to the selective and favourable 

 action of the medium. The second species may exist in almost 

 infinitesimal numbers and therefore the characters of the 

 colony are entirely those of the dominant form, which we may 

 suppose to lack fermentative power. If now the conditions 

 be gradually changed by the addition for example of lactose 

 to the medium in increasing amount, the relative numerical 

 proportions of the two species comprised in the colony may be 

 reversed owing to this medium favouring the growth of the 

 previously subordinate species. At the end of the " educative " 

 treatment the characters of the colony will be those of the now 

 dominant form which we may suppose to possess the power 

 of fermenting sugars. The non-fermenting organism will still 

 exist in the colonies but its influence will not be visible and thus 

 the "organism" will by such treatment have acquired the 

 power to ferment sugar. If the conditions be reversed an 

 opposite process will occur and the original equilibrium be 

 restored. This explanation is probably applicable to most of 

 those instances in bacteriology and mycology where the 

 virulence of a specific culture fluctuates, and it may explain 

 results such as those obtained by Bernhardt and Markoff(4) 

 in which blue colonies of B. coli-mutabile on Drigalski-Conradi 

 agar gave rise to red colonies which if bred through mice and 

 rabbits again became blue. 



In Case II the population is again dominated by the 

 non-fermenting species. By the cumulative addition of lactose 

 however, the conditions may be so altered that the subordinate 

 species becomes dominant and the previously dominant species 

 dies out. Again the final result will be a colony able to ferment 

 sugar but, whereas in Case I the action was reversible, in this 



