193 



numerous subcultures. The same seems to apply on the whole, 

 to absorption, but the material in this case is small. The 

 variability is greater in heterologous agglutination. This applies 

 to strains which have constantly been cultivated on haemoglo- 

 bin (or Fildes) agar, as well as to those grown in symbiosis 

 with different bacteria. The strongest group agglutinations 

 are usually more stable than the weakest. 



The sera do not seem to have become either more or less 

 specific on keeping. 



The most important result of the agglutination experiments 

 reported here is the demonstration of the marked differen- 

 ces between the strains. The finding of strains with 

 identical agglutination reactions is the exception which proves 

 the rule. If identity could not be shown to exist at all we 

 should be on very unsafe ground in talking of heterogeneity. 



My results agree in all essentials with those reported from 

 other quarters. One fact has been brought to light which 

 does not seem to have been observed by other investigators, 

 — namely the demonstration of the wide occurrence of a di- 

 stinct type in a well-defined group of persons. In fact all 

 the strains which were able to absorb agglutinin from serum 

 I 36, except Ms 13, were found in the inoculations, in July 

 1920, from soldiers in the Jsegersborg camp. 



An investigation was made as to how the Pfeiffer's bacillus 

 carriers of identical type were distributed in the camp. Two of 

 them slept in beds side by side, and two others above them, but 

 the remaining carriers of this type were irregularly distributed over 

 this and three other barracks. The type we have to deal with here does 

 not seem to be very sharply defined. As far as can be seen there 

 exist gradual transitions from the rapidly absorbing strains through 

 the slowly absorbing ones to those without this property. 



It will be seen that strains belonging to the same type 

 have, on the whole, the same agglutination „spectrum", by 

 which is meant the complete picture of the agglutination reac- 

 tions of the 14 sera. Perhaps, under certain conditions, it 

 would be just as rational to describe different serological types 

 by their „spectra" as by absorption reactions. 



13 



