But apply Atbash to whole string with spaces ignored then regroup: d→w, a→z, n→m, l→o, w→d, d→w → “wzmodw” n→m, t→g → “mg” w→d, y→b → “db” p→k, y→b → “kb” a→z, n→m → “zm” l→o, a→z, y→b, t→g → “ozbg” b→y, a→z → “yz” l→o, y→b, n→m, k→p → “obmp” m→n, s→h, t→g, q→j, y→b, m→n → “nhgjbn”
Given the symmetry, I suspect it’s applied not to letters directly but to their positions after a shift. Quick attempt: Atbash each letter: d(4)↔w(23), a(1)↔z(26), n(14)↔m(13), l(12)↔o(15), w(23)↔d(4), d(4)↔w(23) → “wzmodw” – not English.
This looks like a cipher or code rather than a standard phrase. danlwd nt wy py an layt ba lynk mstqym
The text: danlwd nt wy py an layt ba lynk mstqym Words are short, and “nt” could be “is” or “not” in English, but the rest doesn’t match directly.
danlwd → dwlnad nt → tn wy → yw py → yp an → na layt → tyal ba → ab lynk → knyl mstqym → myqtsm But apply Atbash to whole string with spaces
Given that “solid paper” is the title, maybe the ciphertext decodes to something like: or similar.
“dan lwd” in Welsh? “dan” = under, “lwd” not standard. “nt” = not English Welsh. “wy” = Welsh for “is” (third person present of ‘bod’? Actually, “wy” = they, but mutation). “py” not Welsh. “an” = Welsh for “from”/”of”. “layt” not Welsh. “ba” = Welsh “if”/”would”. “lynk” = link? “mstqym” no. The text: danlwd nt wy py an layt
But “dan lwd” might be a name? Doesn’t fit.
“an” could be “an” or “is” etc. “ba” might be “be” if b→b, a→e (but then “an” a→e, n→?).
“tn yw yp na tyal ab knyl myqtsm” – no English.
Reading down columns after scrambling — unlikely without more structure.