IX 



duces a red and subsequently a blue colour, whilst no 

 alteration is caused by B. lyphosus. 



Rothberger 15 ), after expcrimenlingwith a large number 

 of colours, recommends neutral red as being far superior 

 to all the others for distinguishing B. lyphosus and B. coli. 

 The method employed was to melt the agar and then 

 add two or three drops of a sterilised concentrated solu- 

 tion of neutral red to each test lube. Within 24 hours 

 from the time of inoculation B. coli produced a light 

 yellow colour in the medium, accompanied by fluores- 

 cence. Rothberger objected to the use of bouillon as it 

 decolourised spontaneously. In a later paper he recalled 

 this statement and rightly so, adding that this sponta- 

 neous decoloration only occurred when a freshly made so- 

 lution of neutral red was used. He also said lhal no fluores- 

 cence appeared in bouillon, which statement is incorrect. 



Rothberger used 4 different slrains of B. lyphosus 

 and 20 ot B. coli and maintains that the reaction is con- 

 slanl for B. coli, bul does nol occur with B. typhosus, 

 B. Friedlaenderi and others. 



A number of investigators, relying upon Rothberger' s 

 results, have recenlly employed Ihe neutral red method, 

 and have found it both satisfactory and reliable. 



Wolff 16 ), endeavouring to explain the cause of the 

 reaction, concludes lhal il is due lo a reduction, since 

 the colour returns by reoxidation, and from this and 

 Ihe facl of the tubes being examined at different periods 

 of time after inoculation, accounls for Ihe diversity of 

 results obtained by different investigators. To overcome 

 the possibility of reoxidation when experimenting with 

 neutral red and B. lyphosus and B. coli, he covered Ihe 

 cotton wool corks with liquid paraffin and by this means, 

 obtaining the same result as Rothberger and Scheffler, 

 he regards Ihe method as a good one. 



