Coding of open-ended carbon chains

IN a previous communication1, a method of coding open-ended carbon chains was suggested, which amounted, in effect, to forming the autocorrelation function of the net. This code had the advantage of being nearly unambiguous, but suffered from the drawback that, while the code of any given net could be determined straightforwardly, the net corresponding to a given code could not. A. Savitzky has suggested (private communication) that, nevertheless, this code could be useful as a check; that is, if added to the usual chemical name of a given compound, this code could serve to verify whether the topology of the compound reconstructed from the name is the correct one.