From this it would appear that the incongruity essential to the very existence of what is generally understood as Opera, as distinct from the early German Singspiel, was so strongly felt by Brahms as to seem to him incompatible with dramatic truth, and to be absolutely prohibitive in his own case of the dramatic exercise of his art. The matter is, however, susceptible of another explanation. It is clear that Brahms, when contemplating the composition of an opera, was bound by the necessities of his position to seek the attainment of dramatic truth in a direction other than that in which Wagner had led the way with such triumphant result. Every circumstance in the careers of the two men, and not least the representative position achieved by each in his own sphere, precluded the possibility that Brahms should run the risk of appearing to seek to emulate Wagner on his own ground, though it would be difficult to believe that he at no time cast longing thoughts towards the logical, consistent, rich means of artistic effect offered by the Melos.[55] No one can doubt that if he had been in a position, and had chosen, to use it, he would have employed it in his own way and for his own original purposes and effects. The skill with which he might have handled it in opera is to some small extent indicated in the Rhapsody (Goethe's 'Harzreise'), where the method of the two first sections is very much that of the Melos, whilst the prayer, affording an opportunity 'inherently suitable for musical expression,' reverts to the rhythmical melody of musical tradition. That Brahms had a respect almost amounting to veneration for Wagner's powers is matter of common knowledge. Though he was never present at a Bayreuth performance, he had studied Wagner's scores exhaustively, and, in the sense of his intimate acquaintance with them, was accustomed to call himself the 'best of all Wagnerians.' An anecdote related by Richard Heuberger,[56] to whom the master gave informal instruction in composition for a time from early in 1878, is highly illustrative in this connection. Heuberger says: