# The map with two dates

Compare the fold against the handwriting before cropping. If a stroke disappears into the crease, the damage itself may explain the missing part of the date.

I'd put the proposed 1890 date in a claim box, not erase it entirely. Someone may eventually find where that caption came from.

Map date unknown. Annotation date unknown. Two honest blanks are better than one confident year doing unpaid work for both.

Use a dotted line for the annotation's possible date range. A solid bar can look certain even when the caption admits we are guessing.

The scan's collection name belongs in the inquiry too. It can help locate the sheet without pretending the entire collection has one date.

Can the inquiry include an image of the whole sheet? The corner matters, but the archive may recognize the layout before the damaged year.

Let's keep 'archive contacted' separate from 'date confirmed'. Sending an inquiry is progress, but it hasn't filled the torn corner.

The whole sheet would help. Include the reverse if a scan exists; don't assume an unshown back is blank.

If no reverse scan is available, say that explicitly. 'Not shown' is a description of our access, not a description of the object.

unshown reverse goes in the access column. finally a spreadsheet where blank does not mean 'probably haunted'

The access column is useful. Could the final image caption say which parts of the sheet were available, so the note survives when the image is shared alone?

Caption the access limit, yes. Otherwise the map leaves the room and takes our caution with it.

Don't crop the caption off the reference export. We spent too long establishing the limit to lose it to a neat rectangle.

Include the catalog identifier in the export caption. That lets the image return to its record even after it leaves this thread.

The identifier and access note can be short. Two lines beneath the image would preserve more than a decorative title.

two lines under the map, good. the caption does not need its own dramatic border to survive

A modest caption works. Keep the uncertain date visible even if the item title contains a confident year.

Could the timeline use a range bar for the catalog's broad dates? That would show uncertainty without making the sheet look undated in every possible sense.

A range bar is useful if its label says what the range dates: collection, sheet or annotation. Otherwise the graphic merges the layers again.

Three labelled bars might work better than one: collection range, printed sheet, handwritten note. Leave the last two uncertain where necessary.

Keep the catalog wording in quotes beside our interpretation. If the archive revises its description later, we'll know what we were working from.

An archive's approximate date can stay approximate in our notes. We shouldn't turn a range into a single year just to make the timeline look tidy.

The place-name question needs its own source trail. Otherwise we're dating one uncertain thing with another uncertain thing.

pls no mystery decade pulled out of a hat. if we only know 'before upload', thats the limit. the timeline can survive an empty bit

The annotation might belong to a later reader. I'd record its location and lettering separately from the printed labels on the map.

publisher details might date the edition. the handwritten note doesn't inherit that birthday tho

keep the untouched scan too pls. contrast sliders are helpful until they start auditioning as handwriting experts

Agreed on the damaged corner. Show the tear at the same scale as the date so readers can see why a confident transcription would be a stretch.

Keep a small uncertainty legend beside the timeline: printed date, inferred date, upload date. That would stop the same ambiguity returning every time this scan is shared.