NATURE OF THE ACTIVE PRINCIPLES 307 



forms of B. diphtheriae (gravis, mitis and intermediate types) , 

 having characteristic cultural and biochemical reactions and ap- 

 parently bearing relation to the clinical course of the disease. 

 The type gravis is responsible for severe forms of the infection 

 resistant to treatment. Type mitis produces the least fatal infec- 

 tion and is well controlled by serum therapy, \vhilst the inter- 

 mediate type appears to possess greater virulence than type mitis. 

 Robinson and Marshall (1934) also foimd that the gravis and 

 intermediate types possess greater po^ver to invade and persist in 

 the tissues of the animal body than the type mitis. On the other 

 hand, the type mitis may give better toxins than the other two 

 types and there is little correlation between toxigenicity in vitro 

 and virulence. It is also of importance that Park-Williams strain 

 No. 8, well known for its high toxigenicity, gave in the hands of 

 Robinson, predominantly colonies of type mitis. 



Again, similar results are obtained Avith the active principles of 

 the phenomenon. Conversion of stock strains of B. typJwsus to 

 their rough variants is accompanied by an unquestionable lo^ver- 

 ing of virulence. This was previously observed for the colon 

 typhoid group by Topley and Ayrton, Jordan, Orcutt, White, 

 Goyle, Schiitze and others (quoted by Hadley, 1927) . In my ^vork 

 the "rough" variants of low viridence produce, however, highly 

 potent reacting factors which differ strikingly in antigenicity and 

 neutralizability from the reacting factors derived from the 

 "mother" stock strains. 



It is common knowledge that no relationship exists bet^veen 

 the production of diphtheria toxin and the amoiuit of growth. 

 It is frequently obser\ed that certain peptones having no in- 

 fluence upon the rate and amount of gro^vth, either inhibit or 

 promote the production of toxin. The lack of correlation ^vas 

 especially obvious in recent experiments of Pappenheimer and 

 Johnson (1936) in which, for example, if the quantity of iron 

 present was allowed to exceed the concentration necessary for 

 optimum toxin production, no toxin was obtained in spite of a 

 good amount of gro\\th. Similarly, there is no correlation between 

 the amotmt of growth and the production of active principles by 

 meningococcus. Cultures in media enriched with various sub- 

 stances and inctibated for prolonged periods of time may yield 

 inactive preparations, whilst incubation for twenty-four hours of 

 cultures on j^lain agar may, on many occasions, yield prepara- 

 tions of considerable potency. 



