256 DYEING 



way one would be able to distinguish grades of metachromasy, 

 according to the ability of various tissue-constituents to resist the 

 anti-metachromatic effect of these solutions. A start in this direc- 

 tion has been made by Sylven.*^* 



Certain substances give metachromatic colours some resistance 

 to dehydrating agents. Potassium dichromate, potassium ferri- 

 cyanide, ferrous sulphate, and uranyl nitrate have been especially 

 recommended. ^^^'^*^ No satisfactory explanation of the action of 

 these salts has been provided. 



The aluminium lake of coelestine blue is remarkable for the 

 resistance of its metachromatic colour to extraction by ethanol 

 during dehydration. ^^^ 



There are several indications that polymerization may be con- 

 cerned in the metachromatic colour-shift. Thus dilution would 

 antagonize polymer-formation and it is also antagonistic to meta- 

 chromasy. The same applies to increase of temperature. If poly- 

 mer formation were the cause of metachromasy, we should have an 

 explanation of the fact that the metachromatic constituent of the 

 dye-solution cannot be isolated from the orthochromatic as a dry 

 substance. Non-conformity w^ith Beer's law implies polymeriza- 

 tion or the formation of complexes of some kind. 



The American chemists, Sheppard and Geddes,*^'^ suggested 

 that the cause of non-conformity with Beer's law was the coupling 

 of dye-ions in pairs. They considered that the conformist (ortho- 

 chromatic) dyes were prevented from making dimeric associations 

 of this kind by the particular shapes of their cations, for the ions of 

 certain dyes would fit together less easily than others. Since 

 resonance would occur differently in a monomer and dimer, 

 colour would be affected. 



The idea that metachromasy (whether of simple aqueous dye 

 solutions or of dyed chromotropes) is due to polymerization has 

 been especially advocated by Michaelis.^*^' ^^^ His contention is 

 that the a band in the absorption curve represents the monomeric 

 form of the dye, the g band its dimeric form, and the wide y band 

 its various polymeric forms, with overlapping effects. Sheppard 

 and Geddes *^^ look at the problem in a slightly different light. 

 They claim that the p band can just be detected even when the 

 dye is dissolved in organic solvents, in which dimers (and polymers) 

 were thought not to exist. In the dimeric form of the dye the reson- 

 ance that exhibits itself as the p band is greatly augmented. These 



