3 The bibliographical description
Every rendering opens with a description of the edition in the form Fredson Bowers sets out in Principles of Bibliographical Description (1949), following the worked example at i. 128–9:
RT] |
| running-titles, with their variants grouped by the damage that identifies a skeleton |
Coll: |
| format, collational formula, leaf count, numbering — followed by the contents |
Sigs: |
| the signing statement in Bowers's $-notation: $2 signed A means the first two leaves of gathering A |
CW: |
| catchwords, with the following word bracketed where it does not answer |
Type: |
| lines to the page, type-page in mm., and the 20-line measurement |
States: |
| press variants by forme, with the point in the run at which each was corrected |
Copies: |
| how many were collated |
Notes: |
| compositor stints, and the pages showing strain from the casting off |
Two of these carry more weight here than the rest. The running-title paragraph is where the skeleton formes betray themselves — Bowers: the complete evidence of running-titles “would almost inevitably reveal simultaneous setting and printing of different portions of a book” (i. 125). And the catchword list is his “partial check for variant states of formes”: a catchword that does not answer the first word of the next page is evidence that something was reset or crowded after it was set, and the simulation produces exactly that.
The type-line is computed rather than invented. A pica em is 4.2175 mm and the type is set solid, so the 20-line measurement follows from the body — 84 mm for pica, written 84R in the trade’s shorthand.