The most important thing is to avoid handcoding stylistic elements (e.g., fonts and spacing) wherever possible. Instead, you should use the standard latex environments wherever possible, e.g., for setting theorems, lists of any kind, equations, figures, sectional units, bibliographic references, etc. We will discuss these issues in detail below.
We ask this for a couple of reasons. First, we can then use the hyperref package to automatically set hypertext links to the various elements in your paper, "lighting up" the pdf version with crossreference links that enable the reader to jump to the statement of a theorem or equation whenever you refer to it in the text.
In fact, if you don't use the standard environments, we will have to go in and recode your article to use them.
An additional benefit of using the standard environments is that it will make it easier for us to reprocess your paper when new presentational formats become available. We did a complete retrospective conversion of nyjm papers to pdf format when that became available to us, and we hope to be able to make similar conversions later. The more modularity we have in our papers, the easier such conversions will be.
It is also extremely important that all fonts used be available in postscript type 1 form in the public domain. Otherwise, we can't use them in our pdf files.
Contents
 Papers should be prepared with the journal's latex class file, nyjm.cls. It automatically loads the AMS macros, so you can use AMS latex features like align, gather, etc.
 If the current version of nyjm.cls does not work on your system, please use amsart.cls instead. Any features in amsart will also work when we process your paper for publication.
 See [Lam94] or [MG04] for details on using latex. The latter gives a good account of the AMS macros.
 Please use the standard fields for \author, \title, \keywords, \subjclass, \address, \email and abstract. You may consult our template for topmatter for explanations, if necessary.
 All personal macros should be included in the paper (rather than included from an external file), and placed near the top of the file.
 Do not use \def in your personal macros. Use \newcommand or \renewcommand, instead.
 Please avoid using handcoded formatting commands, especially forced line or page breaks, but also forced vertical or horizontal spacing. Many of these things will need to be adjusted during copy editing, and premature fine tuning can make more work for the editor. Please use standard latex environments (e.g., theoremlike environments, subsectioning commands, enumerate or itemize environements) wherever possible.
 All sectional units should be designated by one of the standard sectioning commands: \part, \section, \subsection, \subsubsection, \paragraph, or \subparagraph.
 For a long paper, you may want to use \part{} as the topmost sectional unit, but most of the time, \section{} is appropriate.
 If you use \subsection{}, you must decide whether the subsections should be included in the table of contents. The default is to do so, but it is easily overridden. Just put \setcounter{tocdepth}{1} before the \tableofcontents declaration.
 The title of the paper and the titles of sectional units should be typeset with the first letter capitalized and no other caps except for proper names.
 If the contents entry for a section title is overly long, use the optional argument [] in the \section command to give a shorter version for the table of contents.
Theoremlike environments are declared using \newtheorem statements. These have two possible formats:
\newtheorem{thm}{Theorem}[section]
declares an environment whose header reads "Theorem 2.1", say, if it is the first {thm} environment used in Section 2. If you leave off "[section]" from the declaration, then the {thm} environments will be numbered consecutively throughout the document, with "Theorem 8" being the eighth such environment in the paper.
The second form looks like
\newtheorem{lem}[thm]{Lemma}.
In this form, the environment {lem} is numbered in consecutively with {thm} environments, using a common counter, "thm". So if "thm"s are numbered by section as above, and if you just set a {thm} environment labeled Theorem 2.1, then code like
\begin{lem} <stuff ...>\end{lem}
will result in an environment with header "Lemma 2.2".
If you want to use the same counter for both theorems and equations, you can use
\newtheorem{lem}[equation]{Lemma}
You can use a variant (starred) form of \newtheorem to set unnumbered environments. For instance,
\newtheorem*{conj}{Conjecture}
will produce a theoremline environment headed by "Conjecture."
If you have not declared a \theoremstyle prior to a given \newtheorem declaration, the default theoremstyle (\theoremstyle{plain}) is used. This sets the header in bold, and sets the body of the environment in italic type.
The plain theoremstyle should be used for theorems, corollaries, lemmas, conjectures, etc.
Certain elements look terrible in italic type, though, so you should include all parentheses, brackets, numerals, colons, and semicolons in {\rm } blocks, preceded by \/ commands when needed to add a little extra space.
If you precede your \newtheorem declaration by \theoremstyle{definition}, the theorem environment will produce headers in bold, and the body of the environment will be set in standard roman. The \theoremstyle declaration will remain in effect for all subsequent \newtheorem declarations unless it is overridded by a new \theoremstyle declaration (or has been delimited by { }).
\theoremstyle{definition} should be used for definitions, examples, statements of notation, remarks and formal discussions.
With the above specifications,
\begin{thm}[Hilbert] ...
will set an environment that begins with something like
Theorem 2.4 (Hilbert).
Proofs should be set between \begin{proof} and \end{proof}. An optional argument can be used as in \begin{proof}[Proof of Theorem 2.3] to give an alternate header to the environment.
The end of proof symbol (a hollow box) is automatically set by \end{proof}. But you can reposition it, say, to the end of a math display, but using the command \qedhere.
Please do not make lists by hand. The formatting generally looks terrible, and they have to be reset by the editors.
For instance, many theorems include listed features, and it's tempting to just write
(a) <stuff...>
(b) <stuff...>,
and so on. The result won't be as appealing as a list made using the enumerate environment.
So it might be useful to know how to modify the enumerate environment to get the labels you want. The counter for a top level enumerate list is enumi. The command that actually sets the label is \labelenumi. So if you want lists like (a), (b), ... you can make the declarations:
\renewcommand{\labelenumi}{(\alph{enumi})}
\renewcommand{\theenumi}{\alph{enumi}}
The former provides the proper output the the \item command, and the latter will give the proper output for \ref{stuff} (or \eqref{stuff}) if you use \item\label{stuff}.
It is advisable to number your equations by section (\numberwithin{equation}{section}) to avoid misunderstandings between the equations and lists numbered as (1), (2), ... .
Equation numbers should be set by the equation environment used, rather than by hand via \leqno, \tag, etc. If you use \begin{equation}<stuff ...>\end{equation}, a number will be set automatically, in parentheses at the left margin.
The purpose for this is to obtain automatic hyperlinks.
The various AMS equation environments (gather, multline, align, etc.) also provide numbers, with a common counter also used for the equation environment. The counter is called "equation". With align and some other multiline environments, there is a number given for each line. You can suppress this on selected lines by placing \notag prior to the \\ at the end of the line. See [MG04] for details.
The equation numbers can be suppressed for a particular environment by using the starred form, e.g., \begin{equation*} ...\end{equation*}.
Figures
Our production system uses pdflatex, which currently does not
tolerate
postscript. but accepts pdf and some graphical formats.
Figures should generally be allowed to float in the document, by enclosing them in the latex figure environment:
\begin{figure}[t]
<insertion code for figure>
\caption{<caption text>}
\end{figure}
will produce a float that will appear at the top of a page.
Note: The figure environment should not be used on either of the first two pages of your document. It will result in latex errors. So include the first figure on or after the third page.
References in the body of your paper should use the \cite command. The bibliography should be set using the standard latex "thebibilography" environment, the entries of which are set using the \bibitem command. Here is a sample that also illustrates some of the bibliographic style conventions discussed below:
\begin{thebibliography}{CS79}
\bibitem[CS79]{pseudofree} {\scshape Cappell, S.E.; Shaneson, J.L.} Pseudofree actions, I. Proc. {\em Algebraic topology, Aarhus 1978} (Proc. Sympos., Univ. Aarhus, Aarhus, 1978). Lect. Notes in Math., 763, {\em SpringerVerlag, Berlin}, 1979, 395447. MR0561231, Zbl 0416.57020.
\end{thebibliography}
Here, the optional argument, [CS79], is the text that appears in the body of the paper when you cite this reference with the command
\cite{pseudofree}.
The argument {CS79} after {thebibliography} should be the longest optional argument for all the \bibitems listed.
If you'd rather use numbered references, just don't use the optional arguments to \bibitem. The argument after {thebibliography} should then be {9...9}, where the number of 9's is equal to the number of digits in the total number of \bibitems you use.
In the pdf version of your paper, the citation in the text will be linked to the bibliographic reference at the end of the article.
For this reason, it is important never to code citations by hand. If you want to cite a particular part of a paper, you can use the optional argument to cite. For instance, \cite[p.~401]{pseudofree} will produce a reference of the form [CS79  p. 401] in the example above.
Authors' names should be given with last names before first. (This is important!) You may use initials in place of first names. Names of separate authors should be separated by semicolons (with no "and"). The author field should end with a period. The entire author field should be enclosed by {\scshape } so it appears in caps and small caps.
The title of the work referenced should be set in standard roman type. Only the first word and proper nouns should be capitalized (unless the title is in German). Journal names and publishers' names and locations should appear in italic. If you are citing a paper that appears in a book, the book title should be in italic.
In referencing a journal article, the volume number should be set in bold type and the year should be enclosed in parentheses. Thus, the entry for a paper in a journal should look like the entry for [Whi49] in the bibliography below.
The entries for most books should look like those for [ZS58]: The author, title, volume, edition, etc., should be followed by the name and location of the publisher and the date of publication.
If a book is part of a series of monographs, the series and number should be given, as in the entry for [Cha83] below.
When in doubt. Check the listing on MathSciNet.
Links to reviews
Published NYJM papers include links from the bibliographic entries to the reviews. We ask the authors to look up the reviewing numbers on both MathSciNet and Zentralblatt. We then insert macros to make the links. Here are some pointers on how to get the review numbers:
First open two browser windows, with the first one open to
http://www.ams.org/mrlookup
and the second to
http://euclid.library.cornell.edu/ZMATH/
Then start inputting author/title information for the papers in the reference list. For instance, consider the paper:
\bibitem{GW1} {\scshape Greene, R.E.; Wu, H.} Integrals of subharmonic functions on manifolds of nonnegative curvature. {\em Invent. Math.} {\bf 27} (1974) 265298.
Here, as is often the case, the title alone produces hits for both MathSciNet and Zentralblatt.
Just copy and paste the link numbers to the end of the reference.
Note that the MathSciNet number, MR0382723 (52 \#3605), is in a relatively new format. They replaced the old codes (still listed in parentheses) with a new numerical code. Please include both codes for the bibliographic entry.
If you have the correct title, you will usually get a hit on that alone. But if you get too many hits, you can add author, year, etc.
But titles are often not quite right. If you don't get a hit, try to shorten it to its most significant part. If you still don't get a hit, try using the author and date alone. And please do correct the citations if there is a problem.
Some pointers:
1. Umlauts and French accents are simply dropped by MathSciNet, so be sure to edit out the tex codes for making them. You should change K\"ahler to Kahler before sending the query to MathSciNet.
2. Zentralblatt edits out French accents, but replaces umlauts by a trailing e, so you should change K\"ahler to Kaehler for a Zentralblatt search.
3. You can leave in dollar signs, as in $p$group, but longer tex expressions tend not to produce hits, as your code may be different from theirs. So it's usually better to leave that part out of your query.
[CS79] S. E. Cappell, S.E.; Shaneson, J.L. Pseudofree actions, I. Algebraic topology, Aarhus 1978 (Proc. Sympos., Univ. Aarhus, Aarhus, 1978), 395447. Lect. Notes in Math., 763. SpringerVerlag, Berlin, 1979. MR 0561231 (81d:57034), Zbl 0416.57020.
[Cha83] Chapman, T.A. Controlled simple homotopy theory. Lect. Notes in Math., 1009. SpringerVerlag, Berlin, 1983. MR 85d:57016 (85d:57016), Zbl 0548.57001.
[Knu84] Knuth, Donald E. The TeX book. AddisonWesley, Reading, Massachusetts, 1984.
[Lam94] Lamport, Leslie. LaTeX: a document preparation system. Second edition. AddisonWesley, Reading, Massachusetts, 1994.
[MG04] Mittelbach, Frank; Goossens, Michel. The LaTeX companion. Second. edition. Addison Wesley, Reading, Massachusetts, 2004.
[Wei87] Weinberger, Shmuel. Continuous versus discrete symmetry. Geometry and Topology (Athens, Georgia, 1985), 319324. Marcel Dekker, New York, 1987,. MR 0873301 (88h:57034), Zbl 0629.57024
[Whi49] Whitehead, J.H.C. Combinatorial homotopy theory, I. Bull. Amer. Math. Soc. 55 (1949) 213245.
[ZS58] Zariski, Oscar; Samuel, Pierre. Commutative Algebra, vol. I. Van Nostrand, Princeton, N.J., 1958.
