Dear

Would you be willing to referee the attached paper


   by

for the Transactions of the AMS?  

Unfortunately, the backlog for TAMS has exceeded reasonable bounds,
and the editors have imposed a hard upper bound on the number of pages
that can be accepted.  Consequently, the competition for acceptance is
fierce, and only papers of the very highest quality should be
recommended for acceptance.  Please keep this in mind as you think
about this submission; be hard, but fair.

If by a preliminary reading you determine that even if the paper is
well written and correct, the results it contains do not meet the
strict standard mentioned above, then I would appreciate your
responding quickly and expressing this opinion.  By the way, I am not
trying to prejudice your reading, but just setting reasonable
guidelines.  If the paper merits deeper study, then you could look
over the AMS information for referees included below.

I hope that you are willing to accept this job, but if not then please
inform me as quickly as you can so that I can look for other possible
referees. In that case, any suggestions as to other suitable referees
would be very much welcome.

Sincerely,
Steffen Lempp

------------------------------------------------------
INFORMATION FOR EDITORS AND REFEREES
of Articles Submitted to the American Mathematical Society for Publication

A referee is asked to recommend acceptance, rejection, or revision of
papers; the final decision rests with the editor. Each recommendation
should be clear and well founded and should be based on a thorough
reading of the paper. Since the referee's report will be transmitted
(anonymously) to the author, reasons for a recommendation should be
carefully stated, preferably on a plain piece of paper without
identifying watermarks so that the report can be photocopied and
mailed to the author without revealing the identity of the referee.

REFEREEING TIME: Papers of less than 10 pages should not take more
than three weeks to referee.  Longer papers may require more time.  A
manuscript of 50 pages could take two months; no paper should require
more than four months.

CRITERIA FOR RECOMMENDING PUBLICATION
Although each journal should have its own statement of editorial
policy for acceptance of papers, the set of criteria listed below may
be taken as a standard for such statements.

1.  Papers must be correct. While this is basically the author's
responsibility, the referee must certainly be reasonably convinced of
the accuracy of the paper.

2.  Papers must be new in a nontrivial sense; e.g. a collection of new
definitions and deductions therefrom is not publishable unless the
author shows that (or unless it is clear that) it simplifies or solves
some existing problems of reasonable importance. Similarly, a new
theorem with an old proof may not be publishable.

The editorial committee should collectively strive to seek out papers
that will enable the journal to represent contemporary mathematical
research significantly in its many dynamic facets. There should be a
consensus of the committee as to the standards of excellence which
will ensure this objective, and the quality of a paper should be
judged on this basis.

3.  Papers must be of interest to an appreciable number of readers,
not just the author's students and a few colleagues.

4.  Papers must be written clearly; at least the referee should be
able to understand them without undue difficulty. In a paper which is
otherwise marginal, failure in this respect could be cause for
rejection.

ITEMS TO BE CHECKED
1. The title should be informative.

2. The 2000 Mathematics Subject Classification categories should
identify the fields of an article accurately.

3. The abstract should summarize the results of the paper.  The main
purpose of the abstract is to enable readers to take in the nature and
results of the paper quickly and to decide whether they wish to read
the entire paper. The abstract should contain no text references to
the bibliography except for the complete bibliographic
information. For example, [3] is meaningless when the abstract is
separated from the article.

4. If key words are included (they are optional), they should indicate
important topics considered. (Editors should be familiar with the
contents of the Information for Authors of Papers of AMS
publications.)

5.  Items listed in the bibliography should be relevant to the subject
of the paper. Referees need not be concerned about the style of the
bibliography.