VARIATION IN THE CHARACTERS OF STREPTOCOCCI 595 



Str. pyogenes varies considerably in colonial appearance. Adopting the terminology 

 of Dawson, Hobby and Olmstead (1938), we may recognize the following forms : («) mucoid 

 colonies, probably corresponding to the pseudo-glossy colonies of Todd and Lancefield 

 (1928). The organisms tend to be imiform m size, arranged in pairs or short chains, and 

 to show capsulation when young. In broth, growth is diffuse or finely granular ; (6) matt 

 colonies, probably representing an intermediate stage between the mucoid and the smooth 

 colony forms. The organisms are slightly pleomorphic, non-capsulated, and are arranged 

 in short chains or small clumps. The growth in broth is coarsely granular but not floccu- 

 lent ; (c) smooth colonies, corresponding to the glossy colonies of Todd (1928a). The 

 organisms tend to be of uniform size and to be arranged in short chains. Growth in 

 broth is diffuse or finely granular; {d) rough colonies, first described by Eagles (1928), 

 flat and very irregular in outhne. The organisms are large, pleomorphic, and arranged 

 in long chains. Growth in broth is flocculent. The type-specific acid-soluble protein M 

 is usually present in the mucoid, matt, and smooth forms, but not in the rough form. 

 Its loss is associated with an absence of virulence to mice (Todd 1928ff, Todd and Lance- 

 field 1928, Lancefield and Todd 1928), but its presence does not necessarily signify that 

 the stram is vu'ulent. The change from mucoid or matt to rough or smooth corresponds 

 to the S — >- R variation that occurs in many other bacterial species. According to Todd 

 (1930), who studied the influence of oxygen pressure on this change, repeated subculture 

 on solid media under aerobic conditions favours the appearance of smooth avirulent 

 variants, but cultivation under anaerobic conditions prevents it. The effect of aerobiosis 

 is apparently dependent on the formation of bacterial peroxide, the smooth variants 

 being more resistant to this agent than the original matt virulent forms. In broth cultures 

 the effect of increasing oxygen pressure by aeration is somewhat different ; matt avkulent 

 variants tend to appear under anaerobic conditions, whereas the vii'ulence of the original 

 matt strain is maintained when the culture is freely aerated, even though smooth variants 

 make their appearance. The change from matt to smooth is not associated with a loss 

 of the power to produce haemolysin. 



Variations in other Groups of Haemolytic Streptococci.^Variant colonies of more or 

 less similar type to those found in Str. pyogenes have been described in other groups of 

 streptococci. The reader will find an illustrated description of them in the article by 

 Dawson, Hobby and Olmstead (1938). In addition, " minute " or dwarf colonies have 

 been observed in Groups C and G by Long and Bliss (1934), Lancefield (1941) and Morton 

 and Sommer (1944). Though, in general, the fermentation reactions of streptococci are 

 relatively stable, differences have been noted in the same species associated with varia- 

 tions in colony forms. For instance, in Group C, the smooth and dwarf colony forms 

 are said to ferment lactose and trehalose, but not sorbitol or mannitol : on the other 

 hand, the mucoid colony forms ferment sorbitol and mannitol, but not lactose or trehalose 

 (Morton and Sommer 1944). 



Variations in Str. pneumonise. — The more important types of variation that are en- 

 countered in the pneumococcus have already been noted in Chapter 9. The S — > R 

 variation is here associated with the loss of the characteristic capsule, and with it the poly- 

 saccharide antigen that confers type-specificity. Here, as elsewhere, the rough variants 

 of pneuraococci usually retain the characteristic bile solubility. They also retain the power 

 of producing a hsemolysin and a leucodicin (Oram 1934). 



We have also referred in Chapter 9 to the important observations of Griffith (1928) 

 on the conversion of a smooth strain of pneumococcus, belonging to a particular antigenic 

 type, through the corresponding rough variant to a smooth strain belonging to a different 

 antigenic type ; and we have noted that these observations have been confirmed by several 

 subsequent workers. Up to the present time this remains the only instance in which a 

 transmutation of one normal bacterial type into another has been demonstrated under 

 experimental conditions ; and it affords no grounds for the assumption that such types 

 are unstable under natural conditions. 



