Contrary to instrumental music, which was usually quite accurately notated, composers wrote their operas right from the beginning only in a sketchy form. In this way they propably wanted to allow conductors full scope for their creative autonomy and at the same time freedom to optimally adjust the work to whichever conditions prevailed. This is to be found in the Italian operas of Monteverdi and Cavalli and, too, in the stage works of Lully and his successors. Printed scores usually only show 2 or 3 instrumental voices while the solo vocal parts are written out fully. We know, however, that the orchestra had normally five voices and therefore the middle voices and the instrumentation had to be first worked out and written in. Now from Parisian performances there are many preserved orchestral parts so that many accurate scores can be reconstructed of various works. It is of course very difficult to see from this what part the composer played. In many cases the parts were probably written by the musicians themselves or by clever arrangers perhaps from the composer’s “workshop”.
Most of Rameau’s operas have also been handed down to us in this way. There are two printed scores of “Castor and Pollux", one of the original version of 1737 and one of the second version of 1754, both being only sketchily written. The overture is only written on two lines and there are only a few written remarks which give a clue to the instrumentation (in the Allegro there is occasionally written “violons”, “hautbois”, “bassons” or “tous”). The flute and oboe parts are now and then written out in full but the middle voices never. The dynamics are quite exactly marked even with intermediate values. - The preserved orchestral parts of the first performance allow a mainly five-voiced score with two violins to be constructed, a score which makes a very authentic impression. Judging from the quality I would like to assume that the middle voices and  the details of instrumentation come from Rameau or were at least supervised by hirn. The few directions in printed score-sketches have in these voices been exactly followed. There are, too, many arrangements, which diverge from this edition, which were obviously not sanctioned by Rameau (their instrumentation deviates too much from his directions) which show, however, how big the scope was the performers at that time had. (There is for example, one score in which apart from flutes, oboes and bassoons there appear also horns which certainly did not lie in the intention of the composer. In the same version bassoon solos written by Ramcau are given to the violas etc.)
Apart from notation and instrumentation there is posed in every Baroque opera the question of improvisation and ornamentation. In this respect the French were far stricter than the Italians and tolerated no free ornamentation whatsoever hut only the exactly coded “agréments” of instrumental music. This demanded an extremely refined and studied performance. In every case it must be decided which kind of ornament is intended (before or on the beat, long or short suspension, with or without termination etc.). Rameau said: “...even if an ornament is performed well and a “certain something", which makes all the difference, is missing... too much or too little, too early or too late, longer or shorter “belated notes” (suspensions), with notes that increase or decrease in volume, in one word that last bit of precision... if it is even missing only once,then every ornament becomes boring...” These ornaments were taken from instrumental music and used for vocal music, even in choral passages. This recorded performance has tried to find the spirit and sense of every ornament and to perform them with that “certain something” which was so important to Rameau.
In creating the recordings on this record we have tried to convey at least a minimum of operatic scenery. Of course we didn’t want to have the characters walking from one imaginary spot on the stage to the other but we did want to give them a definite place in their respective scenes even occasionally implying a necessary movement. In the “divertissements” in the Prologue and in the 2nd, 4th and 5th Acts we have included two small additional orchestras which as it were are positioned to the left and right of the stage and which play some of the dances but also accompany many of the choruses and solos. This practise corresponds, of course, to old theatrical usage and is intended to make the various levels of meaning of the main action and divertissement clearly audible.