Pathways of Decomposition of Indole Derivatives 195 



spots and zones giving a yellow reaction with Ehrlich's reagent were 

 identified as being anthranilic acid, an o-aminobenzene derivative 

 which was found among the decomposition products of indole deriva- 

 tives by other workers (20). 



In contrast with substances reacting with Salkowski, Gordon- 

 Weber, and Ehrlich's reagents, fluorescent substances which do not 

 give those reactions were numerous. It was rarely possible to estab- 

 lish the correspondence of fluorescent spots of one indole derivative 

 with those of another one. Spot II of the chromatograms of lAA 

 (Figure 7) is an exception in that it was also apparently present in 

 chromatograms of TRPH, IPA, and lAAL. It would seem, therefore, 

 that most of the fluorescent substances originated separately from 

 products resulting from degradation or changes in the nucleus of 

 each derivative, rather than from degradation of the side-chain. In 

 this connection, the possibility of an opening of the benzene ring, 

 yielding a pyrrole derivative (5) which to our knowledge has not been 

 considered to date, is worthy of attention. 



Whether degradation of indole derivatives starts by the side-chain, 

 the pyrrole ring or the benzene ring, the final products must be ali- 

 phatic compounds. Meyer and Pohl (21) claim that the end product 

 of photolysis of lAA by riboflavin is anthranilic acid which, in ultra- 

 violet light, decomposes further into glycine and formic acid with 

 oxalic acid as an intermediate of the latter. It is possible that some 

 of our fluorescent substances are indeed aliphatic compounds and 

 represent final stages of decomposition of indole derivatives. 



The condensation of indole compounds with the formation of 

 colored dimers like indigo and its inducement by ultraviolet radia- 

 tion are well known (1). The possibility that dimerization occurs 

 during the oxidation of lAA has been discussed by Ray and Thimann 

 (25). Houff et al. (14) have shown that a colored dimer of lAA is an 

 intermediate in the formation of N-HIAA, the main product of the 

 Salkowski reaction. Evidence of polymerization in our chromato- 

 grams might be due to such dimers. In Figure 7, spot IX might be 

 the dimer of Houff et al. but we were not able to identify N-HIAA 

 among the decomposition products of lAA because that substance 

 diffuses extensively and quickly disappears in chromatograms. 



The case of lAAL and lA (Figures 5 and 6) is different in that 

 polymerization probably occurs as a result of the condensation of the 

 side-chains of the molecules involved, in a manner similar to that of 

 the corresponding aliphatic aldehydes. 



Summing up, we may say that spontaneous decomposition, or the 

 decomposition induced by ultraviolet radiation, of indole derivatives 

 may follow four main courses and their combinations: (1) oxidation 

 and other changes in the side-chain, (2) oxidation and opening of the 



