ON ABSORPTION SPECTRA OF ORGANIC COMPOUNDS. 225 



Hantzsch's theory, and indeed these are so fundamental that it becomes impos- 

 sible to accept the theory as it stands. In the first place, as was pointed out 

 above, the cardinal assumption on which the whole theory rests is that the absorp- 

 tion band shown by the metallic derivatives of ethyl acetoacetate is due to the 

 secondary valencies of the metallic atom and the carbonyl oxygen of the 

 carboxyl group. There are many cases of compounds in which secondary 

 valencies must be postulated in order to explain their very existence, and these 

 compounds do not generally show absorption bands in the visible and ultra- 

 violet. Some peculiar merit must therefore be attributed to the six-membered 

 'ring' of Hantzsch's formula, and it is difficult to accept this since the selective 

 absorption of such compounds as' the alkaline nitrates and chloropicrin obviously 

 cannot have any relation to a six-membered ring. 



More important still are two facts which appear to have escaped the notice 

 of Hantzsch. First, ethyl dimethylacetoacetate in the presence of alkali shows 

 an absorption band very similar to that shown by ethyl acetoacetate in the 

 presence of alkali. Second, ethyl ;8-ethoxycrotonate shows an inciprient absorp- 

 tion band in the presence of acid. It is obvious that these two observations are 

 in direct opposition to the Hantzsch formula as the correct explanation of the 

 selective absorption shown by the metallic derivatives of ethyl acetoacetate. 



Still more cogent arguments against the theory of correlation between 

 structure and absorption in the visible and ultra-violet are to be found in such 

 cases as pyridine and piperidine. Pyridine in the homogeneous state and in 

 solution in various solvents exhibits an absorption band with centre at 

 1/X = 3910. but in the vapom- state it shows an entirely different band with 

 centre at 1/A = 3587." Piperidine vapour shows a well-marked absorption band, 

 but in solution and in the homogeneous state it is completely diactinic. Analogous 

 dissimilarities between the molecular absorptive powers of liquid and vapour have 

 been observed with other compounds, and clearly on the structure-absorption 

 theory the structure of the molecules in the liquid and vapour phases must be 

 different. This would seem to be impossible at any rate in the case of 

 symmetrical molecules such as pyridine and piperidine. 



The evidence against the direct structure-absorption correlation theory as 

 developed by Hantzsch is overwhelmingly great, and this is equally true of the 

 quinonoid explanation of visible colour. The evidence of numerous colourless 

 compounds which cannot be quinonoid in structure is sufficient to condemn this 

 theory, even were there no other evidence against it. One of the most often 

 quoted instances in which the quinonoid theory is invoked is the well-known 

 case of aminoazobenzene. This compound gives with hydrochloric acid (one 

 equivalent) a salt which is more highly coloured than it is itself. This is 

 universally accepted as being due to the salt having the structure 



C1H,N=<: >-N-NH- 



because the colour and absorption spectrum is entirely different from that of 

 benzeneazophenyltrimethylammonium iodide. 



\N- </ \-N=H- 

 (T ^ ^ 



which of course corresponds to the normal form of the hydrochloride. 



On the other hand, the trimethylamnionium compound also gives a salt which 

 is more highly coloured than it is iipelf, and obviously this cannot be due to a 

 quinonoid structure, It is clearly pjijugtifiahle to gxplain th? one case of colour 

 1920 Q 



