As I understand the "Silver Swan Rag" story, in the late 1960s, it came
to the attention of roll collectors (Mike Montgomery, among them) that,
among the pieces listed in QRS catalogs from 1914 and forward, TWO
pieces were found that that listed Scott Joplin as composer but were
not among Joplin's known body of works: "Silver Swan Rag" and
"Princeton Tiger March and Two Step".  A piece by the name "Princeton
Tiger March and Two Step" was known to exist, written by a Gerald Burke
and copyrighted in 1902, which bore no resemblance to a Joplin
composition.  It seemed that the "Princeton Tiger" listing could likely
be an error, giving Joplin's name instead of Burke's.  However, since
the actual QRS roll of "Princeton Tiger" had not yet been found, there
was a slim possibility that Joplin had written a piece with the same
title.

Eventually, a copy of the QRS roll of "Princeton Tiger" was found, and
it was indeed the Burke piece, and the catalog entry was confirmed to
be an error.  "Silver Swan," on the other hand, proved more enigmatic.
Copyright searches revealed no copyrighted piece by that name.  Until a
copy of the roll surfaced, there would be no way to determine if it
might be a Joplin composition.  Around 1970 (I don't have the exact
date in front of me) a copy of the QRS "Silver Swan" roll came into the
hands of Richard Zimmerman, who, on playing it, realized it did sound
very much as if it was composed by Joplin.  Excited (and exciting)
cross-country telephone calls to Michael Montgomery and Trebor Tichenor
followed, with Zimmerman playing the roll for them over the phone,
which confirmed Zimmerman's initial impression.  Indeed, the piece
really did sound like a Joplin work.

Over the next year or so, the roll was recut and issued to the public,
and in 1971 Zimmerman copyrighted and transcribed the piece for print
publication.  Much discussion about the piece at the time in ragtime
and music roll publications led to a general consensus that "Silver
Swan Rag" most likely was composed by Scott Joplin.  Since that initial
"find" by Zimmerman, a few other original copies have surfaced, but I
understand there are not many.  In addition to a few additional copies
of the QRS roll (number 31533), an identical roll, with the Master
Record label and manufactured by the National Music Roll Company
(number 1239) has also been found.  Given the ease with which player
rolls can be copied and the prevalence of pirated copies being
produced, it is not at all unlikely that the Master Record "Silver
Swan" was a pirated roll. 
 
But why was "Silver Swan" never copyrighted, and published as sheet
music until the Zimmerman transcription in 1971?  The answer lies in
copyright law at the time.  Originally, copyright for a piece of music
only covered print publication, not "mechanical rights" in sound
recordings such as player rolls, music box discs, or phonograph
records.  What Joplin seems to have done was, instead of taking the
piece to a sheet music publisher, (or, having failed to interest them
in the piece) he took it to QRS, who purchased it, and made a roll of
it for sale.  However, since QRS was not in the business of printing
sheet music for publication, they never copyrighted it and never issued
it as sheet music.  There may well have been a clause in the sale
agreement between QRS and Joplin that, upon purchase, all rights to the
piece would transfer to QRS, which would then have prevented Joplin 
from being able to get the piece published at a later date with a print
publisher.

As to the discussion of lost rolls, the case of "Silver Swan Rag" is an
interesting one, in that it not only was a roll that had not been
found, but a composition by a significant composer that was unknown to
modern audiences, prior to its discovery.  The finding of "Silver Swan
Rag," therefore, was more than the finding of a known roll that had not
yet surfaced, but also the discovery of a previously unknown
composition.

Bryan Cather
Board member, Friends of Scott Joplin, St. Louis, Mo., and employee of
the Scott Joplin House State Historic Site