THE BLOOD DYES 271 



product, freshly dissolved, is somewhat metachromatic towards 

 strong chromotropes such as the granules of Mastzellen, but not 

 at all towards DNA. Certain facts, then, stand out clearly from 

 what has already been said in this chapter. On standing in solu- 

 tion, but particularly in the presence of alkaline carbonates, 

 methylene blue gives rise to new, metachromatic substances: 

 azure B, azure A, and methylene violet. It is these that are re- 

 sponsible for the purple or red colouring of the chromatin of 

 leucocytes by the Romanowsky techniques, while methylene blue 

 gives its own, contrasting colour to basiphil cytoplasm; eosin dyes 

 the eosinophil granules and red blood-corpuscles. In order to have 

 a solution of known and invariable composition, it is best to use 

 weighed amounts of methylene blue, azure B or A, and eosin, with 

 or without the addition of methylene violet. The basic compo- 

 nents should slightly predominate so as to help the dyes to remain 

 temporarily in solution in aqueous media. A stock solution should 

 be made by dissolving them in methanol or in a mixture of this with 

 glycerol. This solution should be strongly diluted with water when 

 staining is to begin, so as to permit ionization of the dyes. No dye 

 that does not arise spontaneously in the polychroming of methylene 

 blue has any special virtue in Romanowsky dyeing. 



Certain problems remain. The reader may have noticed that less 

 emphasis has been placed on neutral dyes towards the end of this 

 chapter than at the beginning. The Romanowsky dyes originated 

 by the application of Ehrlich's idea, and neutral dyes do in fact 

 exist in the dry state. It is doubtful, however, whether they play 

 any important part in obtaining the Romanowsky effect. When a 

 neutral dye is dissolved in water, ionization must occur, and the 

 dyes presumably act mainly as ions. Some undissociated molecules 

 may, indeed, remain, but these would be very unlikely to react 

 with DNA. They might conceivably dissolve in lipids, and this 

 could perhaps account for the colouring of the neutrophil granules ; 

 but the strongly acidic chromatin would attract only the basic 

 dye-ions. 



In an ordinary solution of methylene blue or the azures, 

 chloride ions are present; in an ordinary solution of eosin, sodium 

 ions. If an exactly balanced compound of methylene blue or azure 

 with eosin be precipitated, the precipitate dissolved in methanol, 

 and the methanol then diluted with water, there will be no 

 chloride or sodium ions in the solution. This fact cannot account 



