I wonder if
in this particular case the profusion of quotes (which are
very loose there) was the main trigger instead.

No, a quick look at three <1870 samples from Google Books suggests it was general practice: 3-for-3 have "double" spaces between sentences.

1824

1840

1858

The reservation I have is the justification. Sure, if you're
going to insert extra spaces to justify a line, the first place
to do it is after a period. This has nothing to do with kerning.

There can be more than one reason to put extra space after a
period, and the looseness of fixed-width fonts could have been
one of the more compelling ones.

hhp

…the looseness of fixed-width fonts could have been one of the more compelling ones.

I agree, however, that was not always the case.
In the early days printers would often omit space after a period or comma, which is not without its own logic.
So when and where did the practice begin?

> that was not always the case.

Clearly that's true. But I think the kerning fallout
you suggested was even less often the case (if ever).

hhp

I for one would hesitate to call these ‘double’ spaces — it’s the old story about ems and ens. These spaces are ems, I think and maybe a shortcut used by typesetters that were paid by the line or page (a practice not uncommon).

[Explanation: if one used an em for a space after a full stop, that would be one wordspace less to justify (eg the distribution of the rest space in a line over all wordspaces) — it would look okay-ish & be a bit quicker to do.]

In the Monotype machine, Tolbert Lanston initially implemented a justification system that applied even spacing between all characters. However, nobody liked it and it was never put into production, they were happier with the familiar practice of justifying by adding to word spacing. It seems they were comfortable with the integrity and consistency of words in fixed-width fonts, and not bothered by wild variation in word spacing.

However, when Quark XPress launched with justification-by-character spacing as default, there was no comment. By that time it seems that variability of letterspace and evenness of overall “fit”, as effected by kerning, was preferred.

A somewhat different kerning effect than I had originally proposed (although I still think that plays a part). As Hrant says, it’s something to do with “the looseness of fixed width fonts”, but I would modify that to “openness” or “spaceyness”, because of the consistency of glyph fitting, which was precise and very tight in many combinations.

> It seems they were comfortable with the integrity
> and consistency of words in fixed-width fonts

I'm pretty sure the Monotype didn't use fixed-width fonts.

hhp

Reckon the 'silk' ribbon bookmark sewn to the spine qualifies as a nice quirk.

Fore-edge painting.

Setting decimal fractions with denominator or subscript figures.

Snails. Medieval monks had an obsession with drawing snails (the slow slippery little animal, or insect I don't quite know) in the margin of their carefully painted text works. Snails fighting knights. Rabbits riding snails. Women scared to death by a snail. Snails snails snails.

There are different interpretations of why monks did so. I take it as a critcism of the productivity level. The monks were experts at that. So if as a quirk in your book design, I'd recommend a snail! Because it is quirky, and historically booky.