106 SPECTROSCOPIC INVESTIGATIONS 



spectra, overlap more or less with the absorption spectra. As it is 

 advantageous to have as high a light intensity as possible for the 

 study of the emission spectra, it is tempting to use highly concentrated 

 solutions. This, however, has the result that in the case of fluores- 

 cence practically the whole short-wavelength part of the emission 

 spectrum may be lost by self-absorption. In chemiluminescence, this 

 overlap may be even greater. Moreover, if the fluorescence yield of 

 the solution is not very low, the absorbed radiation is reemitted for an 

 important part as fluorescence. This means that in the examination of 

 fluorescence spectra the short-wavelength part of the emission spec- 

 trum will be suppressed, and in the examination of chemiluminescence 

 spectra one may arrive at the erroneous conclusion that the emission 

 spectrum is identical with the fluorescence spectrum of the starting 

 material. In this way, Kautsky (1943), who first studied the chemi- 

 luminescence spectrum of dimethylbisacridinium nitrate, may have 

 arrived at his conclusion that at low temperatures the emission spec- 

 trum is identical with the fluorescence spectrum. This conclusion 

 certainly is erroneous as is demonstrated by Fig. 6. In the experiments 

 of van der Burg, special attention has been paid to this source of 

 error, and it can be stated that the fluorescence of dimethylbisacridin- 

 ium nitrate certainly contributes less than 3% to the spectrum of Fig. 

 6. To avoid errors as those discussed above, it is therefore necessary 

 to use relatively dilute solutions and to apply a correction to the 

 emission spectrum for the absorption by the solution. 



In the case of the chemiluminescence of the derivatives of 2,3-dihy- 

 drophthalazinedione ( DPD ) 



O 



NH 

 .NH 



6 



the results were less clear-cut. We will discuss the most illustrative 

 examples, namely the spectra of 5-amino-DPD and 5-methylamino- 

 DPD, Fig. 7. It was found as a general rule that, as far as the wave- 

 lengths of maximum emission are concerned, the chemiluminescence 



