JAMES WALTER McLEOD 217 



in the presence of oxygen and catalase produce much more methemoglobin than any 

 concentration of H2O2 which is likely to occur on oxygenation of such extracts would 

 do, and they suggest the formation of an organic peroxide insensitive to catalase. 



The heated blood agar or "chocolate agar" first introduced by Cohen and Fitz- 

 gerald' is an extremely delicate indicator for peroxide formation by bacteria, and by 

 its use the interesting fact is established that anaerobes produce an oxidative effect 

 similar to that produced by bacteria which can be shown to produce HoOo by the usual 

 chemical tests. This effect can be elicited either as a green ring at the upper limit of 

 growth of an anaerobe in a deep tube of "chocolate agar" grown at the ordinary 

 atmospheric pressure (McLeod and Gordon^) or it may appear as shown in Plate I, C 

 as a faint green coloration around the bacterial colonies when the anaerobe is inocu- 

 lated as a surface culture and incubated under the greatest tension of oxygen which it 

 can tolerate without complete inhibition of growth. 



THE MECHANISM OF BACTERIAL OXIDATION 



It seems likely that all bacteria tend to produce peroxide and that this tendency is 

 proportioned to their reducing activities. Those which can be shown to contain cata- 

 lase and peroxidase presumably use it in their respiratory processes. It is interesting 

 in this connection that Keilin-' found that a reducing mechanism which cannot be 

 separated from peroxidase and may well be identical with it, is essential to the func- 

 tioning of the respiratory pigment cytochrome. 



A reducing mechanism, functioning possibly in combination with a catalyst con- 

 taining the sulphydryl group such as glutathione, may be responsible for anaerobic 

 respiration independently of any Fe complex. Whether H2O2 takes any part in res- 

 piratory processes of such bacteria as pneumococci is not clear. The lack of catalase 

 and the fact that these bacteria resemble the anaerobes in being relatively insensitive 

 to KCN (Burnet)'' which paralyzes cytochrome by fixing it in the reduced state 

 (Keilin),^ suggest that H2O2 does not at all events function in combination with a 

 peroxidase in these bacteria, and possibly is merely a by-product. The very interest- 

 ing problem is presented by the Shiga type of dysentery bacilli. These lack catalase, 

 ha\^ slight reducing powers and limited fermentative activities, do not produce detect- 

 able amounts of peroxide, give no direct oxydase reaction (Kramer), ^ and although 

 sensitive to cyanides (Burnet),'' they diverge in their reaction to this substance from 

 other bacteria tested. I have not found a record of a determination of their peroxidase 

 content. It is clear that none of the mechanisms mentioned could be considered to 

 explain the respiratory processes of this type of bacterium. 



It is not unlikely that the dependence of some strictly parasitic bacteria on the 

 presence of blood in the media in which they are cultured indicates their poverty in 

 peroxidase, which they require and which is normally supplied to them from the tis- 

 sues of the host on which they are parasites (Fildes,'' Webster^). 



' Cohen, C, and Fitzgerald, J. G.: Cenlralbl. f. Baklcriol., Abt. I, Orig., 56, 464. 1910. 



' McLeod, J. W., and Gordon, J.: /. Path, and Bad., 26, 332. 1923; ibid., 28, 147. 1925. 



3 Keilin, D.: loc. cit. 



''Burnet, F. M.: /. Path, and Bad., 30, 21. 1927. 



s Kramer, G.: loc. cit. *■ Fildes, P.: Brit. J. Ex per. Path., 3, 210. 1922. 



'Webster, L. T.: Proc. Soc. Rxper. Biol. ^ Med., 22, 139. 1924-25. 



