NEW METHODS OF USING THE ANILINE DYES. 405 



the only objection to its use being that the dilute eosin em- 

 ployed (though stronger than the eosin solutions in general 

 use) is incapable of staining the background of the tissue 

 strongly. I have sections stained in this way showing dark blue 

 bacilli on a nearly colourless background. 



Moreover, the eosin may be replaced by certain othrr 

 reagents, which have the property of precipitating methyl 

 blue ; for instance, by dilute alcoholic solution of picric acid 

 and by a dilute alcoholic solution of tetrabromofluorescein. 

 This last reagent is made from a strong watery solution of 

 eosin by adding hydrochloric acid, filtering off, and washing 

 the precipitate, and then dissolving it in alcohol.^ 



In the former case I noticed that if the picric acid solution 

 is too strong all the methyl blue is precipitated in the solution, 

 and no staining effect is produced. By diluting the picric 

 acid the bacteria were stained, and by still further dilution of 

 the picric acid no more effect was produced than in the first 

 experiment. These facts seem to show that the precipitation 

 of the methyl blue plays some part in producing the stain. 



Most of the aniline dyes in common use are more soluble in 

 alcohol than in water. It occurred to me that if a saturated 

 alcoholic solution was added to a strong watery solution of the 

 same dye a precipitate would occur, and that possibly bacteria 

 present would be stained. I first tried with Weigert's aniline 

 water solution of Spiller's purple, and obtained very poor 



• See 'The Chemistry of the Coal Tar Colours,' by Benedict aud 

 Knecht, published by Bell and Son. Since writing the above I have been 

 looking over some of the first specimens that I made by the eosin aud methyl 

 blue method about six months ago. I find that some of them are faded, but 

 in others the bacteria are still dark blue on a pink background. In the latter 

 case, instead of using the eosin solution above described, I had used a mixture 

 of three parts eosin saturated in alcohol, and one part of tetrabromo- 

 fluorescein saturated in alcohol. When first made the bacteria were stained 

 nearly black. 



The bacteria in question were small bacilli, which Dr. Klein found associated 

 with some as yet undescribed disease of a sheep. They were not very easy to 

 stain by this method. Scarcely any anthrax specimens stained in this way 

 show any sign of fading. 



VOL. XXV II, PART 3. NEW SliR. F J? 



