End of theorem marker placement

Question

I have a question about marker placement after theorems. I am using a style file provided by Oxford University Press, which I am not used to. It is available here:

http://www.oxfordjournals.org/our_journals/imrn/for_authors/tex_template.zip

This style file wants to put an \openbox at the end of the statement of each theorem, and a \filledbox at the end of every proof.

If one concludes a proof with an equation, I know to use \qedhere to position the end of proof marker correctly. But I don't know how to do something similar in the above situation (after the statement of a theorem). Many of the thoerems, propositions etc. end with equations, and if I use the above style sheet the box is placed too low.

How do I fix this?

Answer 1

You can redefine \qedsymbol just before using \qedhere as in:

\begin{proposition}
This is a funny equation
\begin{equation*}
a = b + c\,. \let\qedsymbol\openbox\qedhere
\end{equation*}
\end{proposition}

I don't know if it is possible, but you could maybe also suggest the people from the journal to fix their class file so that \qedhere works as expected?

Answer 2

Try this: in the math environment, put the \qedhere command inside of an \mbox, as in

\[ .... some numbers and equation. \mbox{\qedhere} \]

it won't work completely correctly, but now the mark inserted is the openbox, and not the filled box. I think this is a bug with amsthm (see below the cut), so you'd be best off just leaving well-enough alone and let the journal deal with it later.

---

The oupau class apparently uses amsthm and not ntheorem for its theorem needs. And this is how it defines the open-box symbol for the QED in the theorem environment:

\providecommand{\qedsymbolthm}{\openbox}
\DeclareRobustCommand{\qedthm}{%
  \ifmmode \mathqed
  \else
    \leavevmode\unskip\penalty9999 \hbox{}\nobreak\hfill
    \quad\hbox{\qedsymbolthm}%
  \fi
}
\def\@begintheorem#1#2[#3]{%
  \pushQED{\qedthm}\deferred@thm@head{\the\thm@headfont \thm@indent
    \@ifempty{#1}{\let\thmname\@gobble}{\let\thmname\@iden}%
    \@ifempty{#2}{\let\thmnumber\@gobble}{\let\thmnumber\@iden}%
    \@ifempty{#3}{\let\thmnote\@gobble}{\let\thmnote\@iden}%
    \thm@swap\swappedhead\thmhead{#1}{#2}{#3}%
    \the\thm@headpunct
    \thmheadnl % possibly a newline.
    \hskip\thm@headsep
  }%
  \ignorespaces}
\def\@endtheorem{\popQED\endtrivlist\@endpefalse }

I don't think the problem is actually with Oxford University Press! I think the problem lies in amsthm! See the definition there for \qedhere

\newcommand{\qedhere}{%
  \begingroup \let\mathqed\math@qedhere
    \let\qed@elt\setQED@elt \QED@stack\relax\relax \endgroup
}

and the definition for the proof environment

\providecommand{\qedsymbol}{\openbox}%
\newenvironment{proof}[1][\proofname]{\par
  \pushQED{\qed}%
  \normalfont \topsep6\p@\@plus6\p@\relax
  \trivlist
  \item[\hskip\labelsep
        \itshape
    #1\@addpunct{.}]\ignorespaces
}{%
  \popQED\endtrivlist\@endpefalse
}

I'm thinking that the pushQED and popQED commands are defined just so they can accomodate different end symbols! The problem, apparently, lies in the definition of \qedhere, which calls \math@qedhere when it sits in a math environment. And unfortunately, instead of the definitions used in \setQED@elt (which is called in text mode), which process the current qed symbol that's in the QED stack, \math@qedhere depends on

\newcommand{\mathqed}{\quad\hbox{\qedsymbol}}
\def\linebox@qed{\hfil\hbox{\qedsymbol}\hfilneg}

which explictly references the \qedsymbol, which is defined to be the filled box in oupau.cls. So in short, the amsthm package, uses two different ways of accessing the QED symbol depending on whether the environment ends naturally (with \popQED) or if you insert the symbol using \qedhere inside a math environment. This, I think, is a bug.