Walton · The China Affair Evidence

The archive

Evidence

The page images that settled a specific question. Nothing here is illustrative — each is the thing that was actually looked at before a decision was made.

Page numbers are positions in the scanned PDF. The author’s own folio numbers, written at the foot of each sheet, run three lower.

How the page was read

The title page went through the scanner rotated ninety degrees, which is why its text had to be read separately from the
Scan page 1. The title page went through the scanner rotated ninety degrees, which is why its text had to be read separately from the rest.
The footnote rule is only about 250px wide. A first pass looked for rules of 400px or more and missed pages like this on
Scan page 8. The footnote rule is only about 250 px wide. A first pass looked for rules of 400 px or more and missed pages like this one.
Footnotes are single-spaced beneath a rule, while the body is double-spaced.
Scan page 154. Footnotes are single-spaced beneath a rule, while the body is double-spaced.
A quoted telegram: deeply indented but double-spaced, which ruled out line spacing as the signal for quoted matter.
Scan page 121. A quoted telegram: deeply indented but double-spaced, which ruled out line spacing as the signal for quoted matter.
The German childrens rhyme  centred and single-spaced, so it keeps its line breaks instead of being reflowed into a para
Scan page 52. The German children’s rhyme — centred and single-spaced, so it keeps its line breaks instead of being reflowed into a paragraph.
A short page ending in a centred continued. A production note from the typist, not part of the book, so it is dropped.
Scan page 80. A short page ending in a centred “continued”. A production note from the typist, not part of the book, so it is dropped.
The one running head that sits below the usual band. It slipped into the body text until headers were matched by their s
Scan page 52. The one running head that sits below the usual band. It slipped into the body text until headers were matched by their shape rather than their position.

Umlauts, checked name by name

Gppingen.
Scan page 27. Göppingen.
Schtte-Lanz, and Dr. Schtte.
Scan page 29. Schütte-Lanz, and Dr. Schütte.
Lt. Bdecker.
Scan page 105. Lt. Bödecker.
Gring, and Knigin Augustastrasse.
Scan page 134. Göring, and Königin Augustastrasse.
Brcker  the crew list that settled this spelling. The author also writes it Broeker elsewhere; both stand.
Scan page 225. Bröcker — the crew list that settled this spelling. The author also writes it “Broeker” elsewhere; both stand.
Sphkorb, the observation gondola.
Scan page 60. Spähkorb, the observation gondola.
Gbel, and his book Afrika zu unseren Fssen.
Scan page 183. Göbel, and his book Afrika zu unseren Füssen.
Der Feind hrt mit!  the enemy is listening.
Scan page 168. “Der Feind hört mit!” — the enemy is listening.

Passages read off the page

Only the last line failed, so this page was corrected rather than retyped: would require a ship of 1,765,500 displacemen
Scan page 49. Only the last line failed, so this page was corrected rather than retyped: “…would require a ship of 1,765,500 displacement”.
in the freezing temperatures, scanned as pErErnTER.
Scan page 125. “in the freezing temperatures”, scanned as pErErnTER.
as a spare elevator helmsman, scanned as sletiakel BeleGtiin.
Scan page 144. “as a spare elevator helmsman”, scanned as sletiakéel BeleGtiin.
A spy does more than steal information.
Scan page 158. “A spy does more than steal information.”
duck under it at a low altitude.
Scan page 191. “duck under it at a low altitude”.
already left. Mangelsdorf did not learn
Scan page 228. “already left. Mangelsdorf did not learn…”
giving it the benefit of what he had learned
Scan page 232. “giving it the benefit of what he had learned…”
leaving 29 dead and fifty injured.
Scan page 235. “leaving 29 dead and fifty injured”.
the people of Naples ran curiously to
Scan page 234. “the people of Naples ran curiously to…”

Checked, and deliberately left alone

mililtary is in the typescript. Not corrected.
Scan page 25. “mililtary” is in the typescript. Not corrected.
43 degress F. is typed that way, with degrees spelled correctly on the next line. Not corrected.
Scan page 150. “43 degress F.” is typed that way, with degrees spelled correctly on the next line. Not corrected.
seems to have been into circulation  the word put is genuinely absent from the page. An authorial slip, not a scanning l
Scan page 253. “seems to have been into circulation” — the word put is genuinely absent from the page. An authorial slip, not a scanning loss.
Here the opposite: the page reads code books. and the scanner read a comma. Corrected, because it is a misreading rather
Scan page 166. Here the opposite: the page reads “code books.” and the scanner read a comma. Corrected, because it is a misreading rather than the author’s punctuation.
The sentence the sky was solidly overcast. runs across a page break  it ends here
Scan page 190. The sentence “…the sky was solidly overcast.” runs across a page break — it ends here…
and completes here. The one-word line overcast. was faint enough that the noise filter was discarding it, silently trunc
Scan page 191. …and completes here. The one-word line “overcast.” was faint enough that the noise filter was discarding it, silently truncating the sentence. Finding this is why the filter now never drops a plausible word.
The single editorial judgment in the book: doing struck through by hand, the correction above it too faint to read. Set
Scan page 230. The single editorial judgment in the book: doing struck through by hand, the correction above it too faint to read. Set as “did”.

The faint pages

Pages 150 to 156 are carbon copies. The machine dropped whole lines on several of them, so these were read off the images and typed out in full.

Full scan of page 150 of the manuscript, a faint carbon copy.
Scan page 150. Mean confidence 89.6% — retyped by hand.
Full scan of page 151 of the manuscript, a faint carbon copy.
Scan page 151. Mean confidence 87.9% — retyped by hand.
Full scan of page 152 of the manuscript, a faint carbon copy.
Scan page 152. Mean confidence 91.1% — retyped by hand.
Full scan of page 153 of the manuscript, a faint carbon copy.
Scan page 153. Mean confidence 90.4% — retyped by hand.
Full scan of page 154 of the manuscript, a faint carbon copy.
Scan page 154. Mean confidence 84.4% — retyped by hand.
Full scan of page 155 of the manuscript, a faint carbon copy.
Scan page 155. Mean confidence 85.5% — retyped by hand.
Full scan of page 156 of the manuscript, a faint carbon copy.
Scan page 156. Mean confidence 86.6% — retyped by hand.
Full scan of page 49 of the manuscript; only the final line is faint.
Scan page 49. Mean confidence 91.1% — but only the last line failed, so this page was corrected rather than retyped. The boundary between the two treatments is a judgment about how much of a page has gone, not a fixed threshold.