Havergal Brian's second set of orchestral variations was written almost four years after the Burlesque Variations on an Original Theme in 1907. The Fantastic Variations on an Old Rhyme is, in its dimensions, a less ambitious work, while it shows a similar adventurousness in the handling of the orchestra, coupled with a greater professionalism in the writing and an increased subtlety in using the variation-form to articulate a large-scale structure. Though it isn't difficult to pick out the individual variations, these aren't the large, separately-numbered and titled, quasi-independent entities of the Burlesque Variations. They aren't numbered at all, most of them are short, and several may make up a larger structural unit within the work as a whole.

Nor is the basis of the work an extended theme, but a few short motives derived from the tune of the ‘Old Rhyme’ - which is, of course, Three Blind Mice. Moreover, Brian introduces a deliberately contrasting element, a theme underived from that source; and his key-scheme is more unified than the one employed in the earlier work - broadly speaking, the music oscillates between the gravitational poles of E and C. All of which may have to do with the fact that the Fantastic Variations was not originally intended as an independent work (though it fulfils that role admirably), but as the first movement of the first work to which Brian chose to give the title Symphony.

Perhaps with this ‘symphonic’ design Brian hoped to avoid direct competition with Josef Holbrooke's ‘Three Blind Mice’ Variations, which were already enjoying some popularity. But in fact the Fantastic Symphony as an overall conception seems to have died an exceedingly swift death (‘which didn't prevent Brian from publishing The Gothic as his Symphony No.2, or the defunct Fantastic lingering on as a mythical No.1 until as late as December 1967). Either because prospective publishers thought the idea uncommercial (according to Reginald Nettel in Havergal Brian: the Man and his Music p54, duplicating his earlier account in Ordeal by Music, p89), or because Granville Bantock did, and criticized the work on musical grounds also (Eastaugh, p107), the scherzo and slow movement were dropped from the scheme and, although not destroyed by Brian, were lost to him when he left Stoke-on-Trent: their subsequent fate is unknown.

Eastaugh (p74) suggests that Bantock ‘tore out’ these movements from the score, but this may simply be a misreading of a Brian letter quoted on p107, which speaks of Bantock ‘pulling the thing to pieces’ - an operation that might have been purely verbal. I have yet to see any conclusive evidence that Brian's original symphony score was ever physically broken up. However, in late 1912, Brian produced new and apparently substantially revised versions of the first movement and finale; and these were finally published by Breitkopf and Härtel in 1914 as two separate works - the Fantastic Variations and Festal Dance.

One wonders if in the original version there were more of these moves - whether Brian's revision involved formal reorganization to allow the work to stand more easily on its own as an independent concert piece. Eastaugh quotes a letter from Brian to Bantock, written while at work on the revision in November 1912, saying ‘I've purged the Variations on 'mice' of its worse crudities’ - which could mean anything; but also a letter of 1921 reminding Bantock ‘how I cut the thing down and rewrote it’. In this latter context Brian is referring to the Fantastic Symphony as a whole, but could the ‘purge’ have involved cutting and shortening of the Variations themselves -possibly through the removal of a ‘recapitulation’ of the contrasting counterscheme, which one still half expects in the score as it now stands?

The only known manuscript of the Fantastic Variations is clearly the basis of the printed score, ie the revised version of 1912; though the date on it - ‘Aug.1907, Stoke on Trent’ - obviously refers to the work's original composition. (Incidentally, this Ms doesn't carry the dedication to ‘my friend Herbert E. [sic] Robinson’ that appears on the title-page of the Breitkopf score: which leaves open the whole question of whether the original version was so dedicated, or whether the dedication was a late attempt by Brian to improve his credit with his erstwhile patron.)

Two years later it was given twice by the Bournemouth Municipal Orchestra under Sir Dan Godfrey, who has a fair claim to be considered Brian's principal champion in the 1920s. The first of these 1923 renderings was said to be ‘its actual first performance in the form that it was originally devised’. Which suggests, not that Godfrey was using the old Fantastic Symphony score, but that the Brighton performances had been of ‘a condensed version’ of the published work. Lewis Foreman (Havergal Brian and the performance of his Orchestral Music, pp40 and 45) guesses that Brian had had to reduce the instrumentation to bring the work within range of Lyell-Tayler's forces.

Although no detailed exegesis of the Fantastic Variations has ever been published (something I shall attempt in the second part of this article), the work has been quite well covered in previous Brian literature, Tovey's brief and whimsical programme-note, reprinted in Volume VI of his Essays in Musical Analysis, must have helped to keep Brian's name less than wholly obscure to successive generations of music-lovers.

Reginald Nettel gives a brief description of the piece in both his books Hamilton Law's programme-note for Godfrey's 1923 performance, more detailed than Tovey, has been reprinted in full by Lewis Foreman (op cit, p45) and is interesting in that the ‘programme’ it provides for the music sounds as if it may have been communicated to the writer by Brian himself. All three accounts are useful as an introduction to the work and make pertinent observations on it. Tovey makes the entirely just comparison with Dohnanyi's Variations on a Nursery Song - establishing, better than any other comment could, just what kind of a work the Brian is. He also identifies the ‘human feminine element’ in the Variations, the counter-theme, the Farmer's Wife - which provides his only music example:

example 1 to be added

What none of these writers has thought fit to mention is that the subject of Brian's variations - Ex 2, announced without preamble and plainly labelled ‘Thema’ in the score - is not, in the generally accepted sense, the tune of ‘Three Blind Mice’. It's a synopsis of that tune, shorn of the repetitions which the nursery-rhyme text requires of it, and tending to reduce the theme to two principal motives (marked a and b in ex 2).

example 2 to be added

It's as if Brian is saying ‘All right, all right, we all know what this tune is, there's no need actually to go through it note by note - let's get down to business straight away’. More than half a century later, he was to do something similar in The Jolly Miller overture. As a result, we are reminded musically of only the bare bones of the rhyme, as if it were ‘Three blind mice / see how they run / all ran after the Farmer's Wife / ever see such a thing in your life as / three blind mice’. Just that, and what appears to be the opening gesture of Strauss's Ein Heldenleben reduced to a giggle on solo oboe!