194 Bitancourt, Nogiieira, and Schwarz 



Other workers, however, have been more successful than we have 

 in demonstrating the presence of indolic products of decomposition 

 in chromatograms and ionograms of decomposed solutions of indole 

 derivatives. Mayr (19), von Denffer and his collaborators (10), espe- 

 cially Melchior who made the most extensive investigation in this 

 field (20), and recently Kaper and Veldstra (15) were able to identify 

 several indole derivatives or at least show the presence of several 

 Salkowski reacting substances which we have failed to demonstrate. 

 Except in the case of lAA and lAAL, we were not able to detect 

 the presence of skatole, and much less of indole, which several of 

 these authors have mentioned among the decomposition products 

 of many of the indole derivatives of the currently accepted scheme of 

 decomposition (Figure 1). With a much more refined sublimation 

 method than the test we have devised, Behrens and Fischer (2) were 

 able to demonstrate the presence of lA, skatole, and indole among 

 the photolytic decomposition products of lAA. 



All these authors, however, have made little if any use of ultra- 

 violet light to detect the presence of fluorescent substances among 

 the decomposition products of indole derivatives and have overlooked 

 those we find in our chromatograms and ionograms. 



Substances giving a yellow or orange-yellow color with Ehrlich's re- 

 agent but not Salkowski and Gordon-Weber reactions were often 

 found in our chromatograms and ionograms. Presumably they are 

 primary amines, especially o-aminobenzene derivatives, which could 

 result from cleavage of the pyrrole ring. Cleavage of that ring in 

 tryptophan with the production of kynurenine, one of those deriv- 

 atives, is well known and, at least in the case of enzymatic decompo- 

 sition, has been shown to occur also with other indolic substances 

 (18). Gordon and Weber (12) state that in excess of air inactivation 

 of lAA by X-radiation appears chiefly due to ring opening. 



Another evidence of ring opening, with the production of an elec- 

 tropositive amino group is given by the presence of zones on the 

 negative side of ionograms (pH 7) of practically all the derivatives 

 we studied. If the substances corresponding to those zones arc o-ami- 

 nobenzene derivatives, the group in the ortho position must be neu- 

 tral, which is suggested in the case of zone a of ionograms of lAA 

 by its strong reaction with DNPH (Figure 4). One possibility is that 

 zone a be o-amino])hcnylacetaldehyde or one of its hydroxyderiva- 

 tives. The o-aminobcnzene derivative resulting from the opening of 

 the pyrrole ring of indole in a reaction similar to that which pro- 

 duces kynurenine from tryptophan is o-aminobenzaldehyde, a sub- 

 stance which also gives a positive reaction with DNPH. None of the 



^Enzymatic decomposition and decomposition induced by visil^Ic li,>;lit in liie 

 |)icsence of fluorescent substances were beyond the scope of the present studies 

 .nid will not be discussed here. 



