Peter Wisse wrote:

>Not only modern composers, I think even Mozart and M. Haydn, wrote music
>to be included in works by others.  [...]

It was *very* common in the world of 18th-C Italian (and Italianate)
opera to write new arias (duets, etc.) to be included in existing operas
by other composers.  First, operas were written with specific singers in
mind, so when an opera was performed elsewhere (or at the same theater but
years later), the new leading singers often requested another arias, better
suited to display their vocal skills.  Or, on the contrary, the new singers
might have been not up to task and required easier arias.  Second, the
impresario might feel that this or that old aria is old-fashioned or not
very good.  Such new arias, duets, etc., could replace the original ones
or else they could be additional numbers.

Mozart wrote several such arias, a couple of ensemble scenes, and a
chorus (lost).  His "replacement" and "insertion" arias (as they are
referred to) are, rather incorrectly, lumped with his concert arias -
which were usually big dramatic scenes (recitativo accompagnato and
the aria proper) to be performed in a concert, although the text was
usually taken from an operatic libretto.  Michael Haydn didn't have much
opportunity to write any operatic music (for years there was no permanent
operatic theater in Salzburg), but Joseph Haydn did.  Actually, about any
operatic composer at that time wrote such new arias at some time or
another.

The practice of inserting arias by other composers and/or replacing
existing arias by new ones sometimes reached the point when an opera was
practically a "joint composition" by several composers, with little left
from the original, "one-composer" opera.  Such multi-composer works are
known as "pasticcio operas".

Less common was the practice of writing new movements to instrumental
works by other composers.  These, too, could be written because the
original movement was too difficult or too easy (in the case of concertos),
or for some less obvious reason.  Mozart added a slow introduction to a
symphony by Michael Haydn, and for years this symphony was regarded as
entirely Mozart's (K. 444).  There is also a (lost) slow movement for
violin and orchestra K. 470 which was written for an existing violin
concerto - unfortunately Mozart didn't specify whose concerto it was.
(It might have been his own or somebody else's.)

One more case of such "foreign" additions are additional instrumental
parts, usually trumpets and timpani parts.  When a symphony, concerto,
mass, or another work was going to be performed at a more prestigious or
spectacular occasion, it wasn't uncommon to add trumpets and timpani to
the existing orchestration.  Writing such parts was a simple task (mostly
doubling the parts or other instruments or stressing the rhythm), and it
could be done by any competent Kapellmeister - or by the composer himself,
as the case might be.  Other instrumental parts could also be added for
various reasons on various occasions.  Nowadays when scholars come across
various copies of the same symphony, concerto, etc., it may happen that one
copy (a set of orchestral parts) contains parts for certain instruments,
while another copy doesn't.  Often it's very difficult to determine which
orchestration is original, and which contains later additions by somebody
else.  (It can also happen that the larger orchestration is original and
the smaller is missing some parts, but the opposite is somewhat more
likely.)

-Margaret Mikulska