VMs: Roger Bacon "gunpowder cipher"...?

[Nick Pelling <incoming@xxxxxxxxxxxxxxxxx>]

Hi everyone,

One other quick thing: while looking into Robert Steele (who donated the 
positive rotograph of the VMS to the British Museum/Library in 1931), I 
found this in Jim Reeds' bibliography:

	Steele, Robert. ``Luru Vopo Vir Can Utriet.'' Nature 121
	(11 Feb. 1928), pp.208-9.
	[About [Roger] Bacon ``gunpowder cipher,'' not VMS.]

For those that don't know about this, it is believed that Roger Bacon hid 
the recipe for gunpowder using an anagrammatic cipher: the anagram 
supposedly transforms...

	luru vopo vir can utriet

...into...

	R. VII PART. V NOV. CORUL. V ET

...which is allegedly short for...

	recipe VII partes, V novellae coruli, V et

...the central part of his recipe for gunpowder. This is certainly 
plausible from the words in the immediate context, though how any decoder 
would be certain to get the numbers (V and VII) in the right order escapes 
me (as well as VI and VI). Perhaps Bacon expected them to try all three and 
see which gave the biggest bang? :-/

Can anyone please tell me if there are more recent articles on this (or 
other similar anagrammatic ciphers) than Steele's 1928 article in Nature? I 
do know about Galileo & Huygens' use of these same kind of thing, but am 
hoping for rather earlier (medieval or very early modern) occurrences. :-)

Thanks, .....Nick Pelling.....