Letters sit roughly by date, left to right. Scroll or pinch to zoom, drag to pan or to move a node. Hover a key to trace it; click to pin it and open its card. Turn on tried, did not fit to add the keys that were tested and rejected.
Keys tried and ruled out
Testing a key that does not fit is half of the work. These are the named keys the notes record as tried on a letter and rejected, with the reason.
About the data
Every link is taken from an explicit statement in the target's notes: that a letter was read with a named key, unchanged or
adapted, that a key was rebuilt from that letter's own material, or that a named key was tested on it and did not fit. Links are not
inferred from resemblance. A key counts as rebuilt here when this project recovered it, from the ciphertext, from glosses
or from a decipherment; keys recovered by other people (Lasry, Tomokiyo, Somogyi, and outside contributors) are scholars'
keys; keys of the period are those that survive in the archives or on DECODE. A scholar's key is placed at the date of the cipher,
not of its publication, which the card gives separately. Several of George Lasry's tables for the French correspondence of the 1520s
and 1530s appear as one key. Letters read straight from a contemporary decipherment, with no key rebuilt, are not shown. The file is
keys.json; the site build checks it.