The Handel tercentenary has brought a predictable flood of recordings of the Water Music, no fewer than three new versions played on authentic instruments having been recorded in readiness during 1983. But the earliest of the seven versions under review appeared as long ago as 1972, and I well remember one respected critic noting at that time that the use of original instruments constituted 'an implied threat... I listened with trepidation' (unfounded, as it turned out). That few nowadays would harbour any such sentiments, and fewer still admit to doing so, is one measure of how far the proponents of 'authentic' performance have succeeded, not merely in rediscovering old techniques and performance practices, but also in fostering among listeners a wide acceptance of styles and sounds that were quite unfamiliar only a dozen years ago. The ultima Thule of authentic performance has, happily, yet to be explored. As Peter Williams, slightly tongue-in-cheek, has pointed out, in a Handel opera the audience chatted all the way through, and Mozart's players may not have rehearsed: he added, more seriously, that it is anomalous to collect old instruments and assemble the right kind of choir to perform Monteverdi's Vespers, only to give the work as a secular concert in the Albert Hall instead of liturgically in a suitably reverberant church. Pursuing this line of thought, a wholly authentic performance of the Water Music is something most of us would devoutly wish to avoid--involving, as it would, an erratically mobile venue formed of vessels notoriously clumsy and unmanageable in build wallowing on one of the world's trickiest tideways. Whimsy apart, however, the Water Music, with its colourful if imperfectly documented origins, does prompt certain questions about the proper limits of concern with authenticity in recorded performances. Stylish playing on instruments of suitabl construction being (one hopes) a sine qua non, four other aspects require consideration. The first is the size of the forces to be employed. We know that at the 1717 water party Handel had 50 players for the outdoor music: a large band, in which woodwind and possibly horns must have been doubled, producing an unusually rich and powerful tutti. Modern 'authentic' performances obviously need not replicate the exact forces any more than the alfresco acoustic in which they played, but they must certainly sound robust enough to be convincing as outdoor music, at least in the F major and D major suites; the more quietly scored G major suite, on the other hand, can well be played by a smaller body of strings, reflecting its likely origin as music to accompany King George's supper at Chelsea. Linked to this question is the subsidiary one of continuo instrumentation. Since Handel can scarcely have had a harpsichord on the barge, some devotees of authenticity argue that the continuo in the outdoor music should be supplied by theorbos-or even, in view of the fullness of sonority in these numbers, omitted altogether. This seems to me an extreme position. A performance surely need not, in order to claim authenticity, set out to reflect every special circumstance of the premiere; and since most of the many subsequent performances of the Water Music in Handel's lifetime must have been indoors, there is as much historical justification for using harpsichord or even organ as there is for using theorbos. Nor are my withers too much wrung by the question of whether or not to add kettledrums to the orchestra in the D major suite, even though there is no positive evidence, documentary or internal, that Handel himself used them there; but exactly where and how best to deploy them if they are included is perhaps more debatable. A third feature of the work which must be considered