R. D. HOTCHKISS 



were active at the surface of cells in those strains capable of fermenting 

 these sugars. 



We have waited until now to mention a large group of other 

 substances, which, so far as known, behave in essentially the same way 

 as tyrocidine. It is not surprising that fixatives such as hot water, 

 alcohol, acetone, and trichloroacetic acid also liberate cell constituents. 

 But, in the enormous class of synthetic detergents, wetting, dispersing, 

 and emulsifying agents, etc., there are many bactericidal surface-active 

 agents; and, so far as they have been investigated, these show the same 

 phenomena as tyrocidine when they kill bacteria. Cell solutes are 

 liberated, respiratory activity is depressed, autolytic changes are 

 initiated, whenever the substance and its concentration are adequate 

 to kill. Indeed, tyrocidine deserves no special consideration except 

 for the fact that its behavior has been investigated more fully than that 

 of any of the synthetic products. And the diversity of chemical struc- 

 tures which produce this cytolytic type of killing allows us to conclude 

 that the structural requirements for activity are chiefly physical and 

 are broad and ill-defined. Basic (cationic) agents are generally more 

 effective, especially toward the Gram-negative bacteria, but they may 

 be aliphatic straight-chain, aromatic or heterocyclic quarternary 

 ammonium bases, or simple fatty amines. Various alkyl and aryl 

 sulfates and sulfonates have similar activity, as do some fatty carboxylic 

 acids. When in concentration sufficient to kill, phenols and cresols 

 have the same eff"ect, but this is greatly increased in alkyl-substituted 

 phenols, which have more enhanced surface activity. The small 

 class of nonionizing surface-active agents has apparently not yet 

 produced a representative with notable bactericidal activity. 



Somewhat systematic investigations have already been made 

 on the relation of structure to bactericidal activity in some of these 

 compounds, especially the phenols and the aliphatic acids. One of 

 the few conclusions that could be reached was that the most effective 

 substance in each series is usually one having approximately sixteen 

 to eighteen carbon atoms. Branched-carbon-chain compounds appear 

 to be commonly more toxic than those with straight chains. There 

 are almost no evidences of selectivity except that Gram-negative bac- 

 teria are mainly vulnerable only to basic substances unless the pH is 

 low enough that the cell proteins can combine with anionic detergents. 

 There are other small diff"erences of sensitivity among various species. 



