Considering the increasing interest in indoles with unsaturated branched side chains2), we wanted to synthesize 6 by ring cleavage and decarboxylation of 5. This seemed to be promising, because heating above the melting point or with mineral acids usually causes 6-lactone cleavage to y, g u nsaturated acids (lacto-enoic-tauto~nerisrn)~) as well as decarboxylation of indole-2-carboxylic acids4).The results of our investigations were astonishing. When heated up to 350°C (metal bath), 5a -prepared according scheme 1 -remained unchanged. A reaction was observed in boiling mineral acids, but 6a newer could be isolated. Prolonged heating afforded the pentacyclic 9a. To facilitate the assignment of spectroscopic data, we also prepared 9b from 5b. 9a,b are alkyl derivatives of indolo[3,2-b]carbazol, which already was synthesized by cyclodehydrogenation of N,N'-diphenyl-p-phenylenediamine'), iindolisation of cyclohexan-1 ,Cdione bisphenylhydrazone6) and treatment of indole with formaldehyde in mineral acid7). The compounds should have a symmetrical structure with a planar ring system. In fact, the two identical parts of 9a.b give one set of signals in the 'H-and I3C-NMR spectra (400 MHz). Concerning the course of the reaction, we assume the initial formation of 6 as described above. Protonation gives the mesomer stabilized cation 7, which dimerises by intermolecular alkylation. The concluding dehydrogenation to 9 takes place without any sensitizers, which is in contrast to the investigations of Bergmann7). So 5 is transferred to 9 in a one-pot ring cleavage-decarboxylation-dimerisation-aromatisation-sequence just by heating with mineral acids.