SALM. PARATYPHI B 725 



almost, non-toxic. The unhydrolysed toxic material, as prepared and tested by Martin 

 (1934), had an average lethal dose for the mouse of 0-5 mgm. The substance obtained by 

 Boivin and his colleagues is reported as being slightly more toxic. 



A point of considerable interest in regard to these findings is that there seems little 

 doubt that the toxic substance is itself the somatic antigen of the bacterial cell. Rabbits 

 immunized with the purified fractions produce characteristic somatic agglutinins. The 

 polysaccharides isolated from salmonella organisms by previous workers, and identified 

 with the somatic antigens, or with their hapten constituents, were in most cases found 

 to be non-toxic. But this was probably because the method of preparation had resulted 

 in the splitting of the polysaccharide from the phosphatide component. 



The same series of experiments have given an indication of one of the ways in which 

 these toxic substances produce their effects. It had already been shown by several 

 workers (see Chapter 44) that natural infections in man, and experimental infections in 

 animals, might be associated with an increase in the concentration of sugar in the blood ; 

 and Menten and King (1930) had found that rabbits injected with the fraction isolated by 

 them from Salm. typhi-rmmum developed hyperglycsemia. 



Delafield (1934), who had already made a careful study of the chemical changes 

 occurring in the blood of the rabbit following the injection of killed suspensions of various 

 bacteria (see Chapter 44), was able to show that the purified fractions prepared by 

 Raistrick and Topley induced a marked hyperglycasmia, followed in many cases by a fall 

 in blood sugar far below the normal level. These findings were confirmed by Boivin and 

 Mesrobeanu (1934fe). 



These observations suggest that the toxic substances isolated from Salm. typhi-murium 

 and Salm. enteritidis are representatives of a group of antigenic components, possessing 

 very similar toxic properties but differing widely in the chemical structure that deter- 

 mines their antigenic specificity, and that bacterial components of this type are very 

 widely distributed among different bacterial groups. Delafield (1931, 1932) found that 

 a wide variety of Gram-negative bacteria, including Bad. coli. Bad. aerogenes, Sh. shigce, 

 Salm. typhi, Salm. typhi-murium, H. influenzce, H. hronchisepticus. Past, muriseptica, 

 Proteus vulgaris and the meningococcus, induced hyperglycsemia in rabbits, while the 

 Gram-positive species studied did not ; and Boivin and Mesrobeanu (19356) have isolated 

 fractions of the same type as those obtained from Salm. typhi-murium from a variety of 

 bacilli of the coli-typhoid group, though the toxicity of most of these fractions has not yet 

 been determined. 



There is, of course, no reason to suppose that these particular substances are the only 

 toxic constituents of the bacteria in which they occur. It is possible that there are many 

 others. They do, however, afford examples of " endotoxins " that are definable chemical 

 substances, and not merely crude extracts containing all the multitudinous components 

 of the bacterial cells from which they were derived. 



We append a detailed description of the general characters of a typical member 

 of the Salmonella group — Salm. paratyphi B — and a summarized description of 

 each of the different species. 



Salm. paratyphi B 



Synonyms. — Bad. paratyphosum B, Bad. schottmiilleri. 



Isolation. — From cases of enteric infection in man by Achard and Bensaude (1896) and 



Schottmuller (1900, 1901). 

 Habitat. — Almost entirely a parasite of the human intestinal tract. 

 Morphology. — Bacilli with parallel sides and rounded ends, usually 2-3 /t long and 0-6 /t 



broad. Gram-negative ; non-acid-fast. Non-sporing ; usually non-capsulated, 



but may form a mucoid envelope. Actively motile ; flagella are stated to be 



peritrichate, but of this there is some doubt, 



