The Art of Fugue occupied Bach's mind intermittently, at least, during the last decade of his life. He left it incomplete and because he died before the engraving process had been finished we are left with imprecise knowledge about several important points. One of them concerns the order in which he intended the various pieces to be assembled. It's likely that the first 11 or so of the Contrapuncti are arranged in the correct order but there is much doubt about the sequence of the remaining pieces. Our knowledge of the instrument or instruments which Bach may have had in mind for performing The Art of Fugue is equally imprecise. He notated it in open score which was not uncommon in contrapuntal keyboard works; indeed, all but two sections of it can be played on a keyboard instrument by a single player and the fashion in recent years has been to perform it in this way. Nevertheless, since Wolfgang Graeser's pianeering orchestration of the work in the 1920s, performances using various instrumental combinations have taken place—I have even heard the entire work performed on saxophones; but for this recording Reinhardt Goebel, in a brief and interesting note on his approach, argues strongly for an interpretation on stringed instruments with harpsichords. Inclusive though such arguments must be, in the results which Goebel achieves are, for the most part, effective.

Listeners to this set will immediately be impressed by the clarity of texture which Goebel unfailingly draws for his ensemble (two violins, two violas, cello and two harpsichords in varying combinations). I was surprised by the fact that they did not double-dot the stretto fugue which Bach designated ''In Stylo Francese''. I have always imagined that it was his requirement of a sharply dotted rhythm which prompted him to provide the sub-title; but, generally I found Goebel's approach a very thoughtful and convincing one. The readings of the three-part mirror fugue, both ''Rectus'' and ''inversus'' (Contrapunctus 13), are absolutely splendid as indeed are those of Bach's own arrangement of it for two harpsichord with its striking rhythmic contrasts. But although this is playing which is likely to make considerable appeal to the general listener, a presentation of The Art of Fugue does seem to me to pose problems unless you have access to the score.

Without it there is a danger of finding the work merely didactic for the most part; indeed, that was how I was once taught to regard it. There is, as I have implied, an immediate and haunting beauty contained, above all, perhaps, in the dance-like mirror fugues, but the full ingenuity of Bach's scheme exerts its full power only when confronted by the music. I strongly recommend this painstakingly prepared version both to newcomers to the work and to those who are already familiar with it. For me it will go hand-in-hand with keyboard performances as a contrasting means of opening up a world of almost infinite musical riches achieved with breathtaking ingenuity. Perhaps I was a little sorry that Goebel decided to round off the unfinished quadruple fugue with a perfect D major cadence; of course that's a sensible thing to do, yet I find the very inconclusiveness of that tailing off an affecting moment. Sensibly he omits the organ chorale, Wenn wir in hochsten Noten sein (BWV668a) which has no true place in Bach's great scheme, having been added as a compensatory gesture for the unfinished fugue in the edition of 1751.

Fine recorded sound and good presentation contribute towards an outstanding issue. I doubt if there will be many new recordings of The Art of Fugue, even in 1985, and I cannot easily imagine this one being bettered.

Side 4, by the way, contains a selection of canons, including the 14 canons of BWV1087 based on the first eight notes of the aria ground of the Goldberg Variations. Also present is the Triplex a 6, BWV1076, which may be more familiar to you than you think since it is the one which Bach himself is holding in the imposing Haussmann portrait of the 1740s. A fine achievement all round.'