The plot thickens. Through the decency of the HathiTrust Library, which has
given the UC Libraries emergency access to online texts, I've been able to
see In digital copies of the earliest printings of this lost letter.

In Cyril Clemens's *Mark Twain, the Letter-Writer *(1932), the dateline is
simply "*Michigan, Dec. 1884*." When Cyril Clemens printed this letter
again in the *Mark Twain Quarterly *in 1941, the dateline had grown to
read: "*Muskegon, Michigan, December 4, 1884*."

There are two possibilities. Either Cyril returned to the manuscript letter
and transcribed it more fully than he had before, OR [a strong nudge here]
he supplied "Muskegon" and "4" out of his own erratic brain. Since both
"December 4" and "Muskegon," as Scott points out, make no sense, I think
it's clear these details are mere invention. Later, no doubt, some well
meaning person made a "correction" from 4 to 14, on the grounds that at
least Clemens was *near *Muskegon on the latter date. Unluckily, they
"corrected" without getting all the facts. (I think it's right to emphasize
again that, for the letters of years we haven't edited yet, MTPO
information isn't as refined as it will be later.)

So much for Mark Twain's phantom side-excursion to Muskegon! (Rabbit hole?
. . . What rabbit hole?)

Ben