It appears that "humble" was occasionally pronounced with a dropped H even
in the US at the time, especially if the speaker was from a family that
immigrated from the UK, or, like the Express's political editor, from
Canada. (Bob Hirst would know whether Larned's editorialzing was ever this
*sustainedly* humorous).

Or, as I think someone else has noted, "an humble" could have been in this
case, if it *was* written by MT, a typesetter's mistake.

I took a look at "an humble" in the NYS newspapers archive, and found an
interesting case of "a humble" vs "an humble" in the transcription of a
speech by an Illinois congressman on the effects of Republican tariffs on
farmers. If you compare these two versions, whose links will hopefully
preserve the highlighting, you'll see at least one case of humble treated
with a silent H and a voiced H in two reprints, suggesting an intervention
based on differences in dialect.
Geneva Gazette, 10 Jun 70
http://nyshistoricnewspapers.org/lccn/sn83031108/1870-06-10/ed-1/seq-4/#date1=01%2F01%2F1869&index=6&date2=01%2F31%2F1874&searchType=advanced&SearchType=phrase&sequence=0&words=humble&proxdistance=&to_year=1874&rows=20&ortext=&from_year=1869&proxtext=&phrasetext=an+humble&andtext=&dateFilterType=range&page=1
Herkimer Democrat, 3 Aug 70
http://nyshistoricnewspapers.org/lccn/sn83031101/1870-08-03/ed-1/seq-2/#date1=01%2F01%2F1869&index=2&date2=01%2F31%2F1874&searchType=advanced&SearchType=phrase&sequence=0&words=flannel+humble+shirt&proxdistance=&to_year=1874&rows=20&ortext=&from_year=1869&proxtext=&phrasetext=humble+flannel+shirt&andtext=&dateFilterType=range&page=1

A survey of the same speech in newspapers.com shows 44 cases of "an
humble," which is apparently how it was enunciated by Rep. Marshall, and
faithfully transmitted, vs 9 cases of "a humble." The typesetter's or
editor's intervention was apparently the dropping of the "n" in this case.
But this sort of intervention could go both ways, depending on a person's
dialect affinities

All to say, I suggest that "an humble" could just be a typo, and not
necessarily a viable data point -- or what I used to call, before I retired
from MTP, "a glitch."

I say data point, because in at least one branch of stylometry, articles,
conjunctions, and other words more unconsciously generated by a writer's
brain appear to make the best case for identification.

For Too Much Information on how stylometry works (yet you can cherry pick
really useful information from it) see
https://programminghistorian.org/en/lessons/introduction-to-stylometry-with-python
Leslie

On Wed, Feb 13, 2019 at 10:13 PM Clay Shannon <[log in to unmask]> wrote:

> Thanks, Barb! I've added it to my amazon shopping list - will purchase it
> later.
> - B. Clay Shannon
>
>     On Wednesday, February 13, 2019, 5:57:19 PM PST, Barbara Schmidt <
> [log in to unmask]> wrote:
>
>  Clay asked --  Has anybody compiled a list of Twain's "vocabulary" --
> Yes.
>
>  A MARK TWAIN LEXICON by Robert Ramsay and Frances Emberson. Published in
> 1963.
>
> Barb
>
>