I have had the luxury of following Poulton and Lam’s transcription (and reconstruction) of Holmes’ MS Dd.9.33 67v–68r, rather than using the MS, which is is rather confusing.
 |
| Holmes’ MS Dd.9.33, p. 68r, bottom three lines. (Cambridge University Library.) You can see the state of the MS. All hail to Poulton & Lam for making their transcription. |
There are seven variations, of 12 bars each.
The first two variations keep quite closely to the melody, but later ones include loads of divisions (mostly in the form of scale fragments) which seem very workaday to my ears, with none of the unexpected twists and catchy repeated motifs of Dowland at his best.
Variation 5 is weird, with (in the transcription) four bars in 3/4 time with triplets, four bars in 9/8, and the final four in straight 3/4 with a totally different feeling and quite out of place. (I have tweaked bar 51 to tidy things a little, but it’s a bodge of a bodge job.)
Poulton, in her biography of Dowland, describes this setting of Walsingham as “far less satisfactory than the one in galliard form (P31)”. [You can see a transcription of P31 for ukulele here.] The MS is the only copy, so she says one can’t ascribe to blame to Dowland performing below par, or to Holmes using a corrupted source. Also, I believe that Holmes was quite old and becoming unwell at the time he wrote the MS. Poulton also describes extending the song from 8 to 12 bars as destroying the “beautiful balance” of the original.
It looks to me like a case of cut-and-paste, with some of the pastings in the wrong order.
Even so, it sounds pretty good when Nigel North plays it on the lute.
Still curious? You can download the arrangements in the following formats:
