According to some authorities ("The Dictionary of National Biography” amongst them), the first play handled by Purcell was Lee’s Sophonisba; or, The Overthrow of Hannibal; according to others, the first was Theodosius; or, The Force of Love.  Both, however, date not later than 1685, which is near enough for either when there is nothing like conclusive evidence as to which had the priority.  The music for the first plays is in no way bound up with the plays.  It consists of instrumental pieces and songs literally interpolated.  It is likely enough that tunes written for one play were often enough used for another.  The pieces were brief, but the unmistakable Purcellian mingling of strength and sweetness is to be found even in such trifles.  In 1690 and later Purcell took full advantage of masques which were inserted, the interpolations being sometimes as long as the rest of the play, and artistically of infinitely greater value.  For the present he confined himself to less imposing forms, which was certainly what he was engaged to do.

The finest example of the odes of the period is the so-called Yorkshire Feast Song (1689).  Many of the others are not, for Purcell, extraordinary.  They were written for such special occasions, for instance, as the King’s return all the way to London from Windsor, or even Newmarket, or the birthday of a Queen, and in one case the birthday of a six-year-old Duke.  They consist of overtures, songs, choruses, etc.  With one or two exceptions, the structure is Purcell’s ordinary.  What that structure was we shall see (once for all) in examining some of the later compositions, the only difference observable in the later works being, on the whole, an increased richness and greater breadth of scheme.  They are nearly always brilliant, often incisive; there are most lovely melodies; and there are numerous specimens of Purcell’s power of writing music, endless in its variety of outline and colour and changing sentiment, on a ground-bass—­i.e., a bass passage repeated over and over again until the piece is finished.  The instrumentation must have been largely dictated by the instruments placed at his disposal, though we must remember that in days when it was an everyday occurrence for, say, an oboist to play from the violin part save in certain passages, even an apparently complete score is no secure guide as to what the composer meant, and as to how the piece was given under his direction.  This remark applies to the scoring of much of the theatre music.  The Theatre Ayres contain only string parts, and it is nonsense to suppose that in the theatre of that time Purcell had only strings to write for.  Purcell wrote in all twenty-two sonatas—­twelve in three parts, ten in four.  So far as the number of parts is concerned, there is little real difference.  In the three-part works one stave serves for both the string bass-player and the harpsichordist; in the four-part ones there are two separate staves, with