172 



ted in the table by a „-p, were all of the titre 5—6 except H 327 

 which was a little weaker (about 4). 



With the exception of series VI the difference between the 

 positive and negative reactions was always very sharp; either a 

 marked red colour appeared or the liquid remained a pure yellow. 



As a control that the reacting substance really was indol, the 

 medium, in series I in the Jan. investigations and in series VI, 

 was shaken up with amyl alcohol. In all the cases where a red 

 colour was present the colouring matter was taken up by the 

 alcohol, while the latter was never tinted red when the reaction 

 had been read as negative. 



The tubes in series IX— XII where the reaction was negative 

 or weakly positive, were shaken with chloroform. Both the positive 

 and negative reactions were confirmed by this test except that in 

 some of the negative tubes a suspicion of red appeared. But as 

 this was only observed in one of the tubes for each strain, and 

 only in series IX, XI, and XII and not in X where it would be 

 particularly expected that a weak reaction would first of all be 

 obtained, this cannot be regarded as an indol reaction. 



In using amyl alcohol for shaking out, a weak red colour 

 could sometimes be seen in the negative tests, but only after several 

 hours' standing, and never when the readings were taken immedia- 

 tely after the liquids had separated. As this separation takes place 

 quicker with amyl alcohol which also extracts the colour more 

 completely than chloroform, amyl alcohol seems to be preferable 

 to chloroform for the purpose of verifying indol tests by the shaking 

 out method. From my experience with Pfeiffer's bacillus this need 

 only be resorted to in the case of weak positive reactions, as 

 marked positive or negative reactions can be read with the same 

 accuracy without it. 



The reactions given for II, I 57 and the other strains where 

 two colonies of different appearance were generated, were partly 

 carried out with the original growth and partly with the two forms 

 of colonies. In all cases however, the different cultures gave identical 

 results. 



The first impression we get from the table is that the 

 reaction of each strain is usually either constantly positive 

 or negative, irrespective of the time it is done and whether 

 the growth took place on media of rather different constitution 



Some of the reactions however, are divergent. These are 

 printed in thick type. We will now consider them more clo- 

 sely. No explanation is forthcoming of the causes of the absent 

 reaction for Ms 9 in series I and the positive reactions for H 151 

 and H 205 in series V. The change from negative to positive reac- 

 tion in I 45 occurred simultaneously with a marked alteration in 



