A MANUAL FOR AUTHORS OF

MATHEMATICAL P A P E R S *
The purpose of this manual is to help mathematicians in preparing
papers for publication. A reasonable amount of attention on the part
of authors to the rules and suggestions given below will save the time
of editors, referees, printers, and readers, and will help keep the cost of
publication at a minimum. Although most of the discussion will
apply to mathematical papers in general, special rules will occasion-
ally be given for papers specifically intended for the Bulletin or
Transactions of the American Mathematical Society.
Style. In the course of writing a mathematical paper intended for
publication, the author must settle such questions of usage and ar-
rangement as (1) the organization of the paper into chapters, sections,
paragraphs, propositions, theorems, proofs, examples, remarks, and
the use of appropriate headings; (2) the adoption of a system of
enumeration for the sections, theorems; (3) the treatment of foot-
notes, references and bibliography; (4) the use of capital or small
initial letters in such adjectives as abelian, euclidean, jacobian; (5) the
use or non-use of hyphens in such technical terms as sub-harmonic,
fixed-point; (6) the questions of style which arise when words in the
text are replaced by symbols.
Concerning the first three questions—particularly the third—the
author's decisions must to some extent be governed by the routine
stylistic requirements of the journal for which his paper is intended.
For guidance, the author should consult current numbers of that
journal; if his paper is intended for the Bulletin or Transactions, he
may consult, instead, the Special Rules given below.
Concerning the remaining questions, perhaps the only general rule
that can be offered is the Rule of Uniformity: Whatever the usage
adopted by an author in a given paper, that usage should be main-
tained consistently throughout the paper. It may be remarked that the
reasons for urging a careful observance of this rule are economic as
well as stylistic. Every paper is carefully checked by an editorial
assistant before it is sent to the printer. An author who has written
"euclidean" part of the time and "Euclidean" the rest of the time
may have done so inadvertently. But the assistant cannot guess this.
* The writers of this manual have received invaluable advice from members of
the office staff of the American Mathematical Society and have consulted freely the
following excellent pamphlets: Notes on the preparation of mathematical papers (The
London Mathematical Society); Typographic suggestions to authors (George Banta
Publishing Co.); Author's manual (Duke Mathematical Journal).

License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use

 2 AMERICAN MATHEMATICAL SOCIETY

Nor can one guess whether the author prefers the capital or the small
letters. The correspondence between author and editorial staff which
must take place in order to settle the matter causes delay and ex-
pense which could easily have been avoided.
The use of mathematical symbols in the solid text is not only
permissible but is frequently encouraged by editors as a means for
securing greater economy of space on the printed page. Authors,
however, should consider carefully the questions of style which arise
from this practice, particularly in the use of " < " and " = ". Such
phrases as "assume x> y" and "assume t h a t x is >y" are in common
use and there appears to be no widely accepted ground for preferring
one to the other. On the other hand, no one would write "assume that
x and y are = " although the objection appears to be largely visual.
Some writers do not hesitate to begin a sentence with a mathematical
symbol, but others regard this as a barbarism. 1 In the complete ab-
sence of agreement on these questions, we offer only the following
general rules in addition to the Rule of Uniformity :
1. Symbols in the text should be used in such a way that their
translation into words or phrases will require no effort on the part
of the reader.
2. The resulting translation should conform to the standards of
good literary composition.
Rules 1 and 2 are admittedly not decisive in all cases. Thus "let
e ^ O b e a number such that • • • " would be objectionable on both
counts, but is probably preferable, in appearance at least, to "let e
be a number ^ 0 such t h a t • • • ." One could, of course, avoid both
forms by writing "let € be a non-negative real number • • • ." Again,
" • • • the expression on the left is < the smallest value of • • • "
satisfies rules 1 and 2. Yet the substitution of two words for " < "
would result in a far better appearance. This improvement is surely
worth the slight extra effort involved.
The printing of mathematical symbols. There are unfortunately
many useful arrangements of symbols which must be avoided in a
manuscript intended for publication, on account of the difficulties
which would have to be overcome by the compositor. A type con-
sists of a solid block or "body" with a raised character ön one of its
rectangular faces. The bodies are normally laid side by side by a
machine and form a row of uniform width. Besides the ordinary
1
There is a similar division of opinion concerning the use of the "hanging theorem"
ip which the word THEOREM is simultaneously the first word of a paragraph and the
last word of the sentence which precedes it.

License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use

 A MANUAL FOR AUTHORS 3

alphabets, most printers have on hand a number of "dashed letters"

(A, a, • • • ) and "tilde letters" (Â, • • • ). If a dashed letter, say 5,
is not in the printer's possession, it does not follow that this letter
cannot be printed. But the typesetting involved is much more
complicated and includes a number of operations by hand. If ö occurs
in the solid text, the body of an ordinary a must be reduced in size
by filing, to make room for a thin body carrying the bar. The two
bodies must then be placed in position by hand. The thin body may
or may not stay in place and thus the alignment of the bar may or
may not be correct. Moreover, the bars, when printed, may not
always be uniform. Thus to the extra expense involved in the forma-
tion and typesetting of the symbol a, it is necessary to add the cost
of re-alignment or re-setting in case of mishaps during the printing
of proof.
The typesetting of inferior and superior symbols (subscripts and
superscripts) in the solid text offers no difficulty when those symbols
are available as inferiors and superiors (see Appendix C) and when no
symbol is to be placed directly above or below another. Thus,
(a+bx)n/z> A2 can be set by machine, but #f, A% require hand work.2
Subscripts to superscripts and superscripts to subscripts, all belong-
ing to a single term, are extremely difficult to set, even by hand.
In displayed formulas, the insertion of bars, subscripts and the
like is somewhat less difficult since there is space above and below,
whereas in the solid text, the compositor has only a single line with
which to work. Thus the symbol A\ no longer requires hand work
since the superscript 2 can be set by machine in the subscript position
of the line above the line of the A. Note however that when set in
this manner, the superscript will be slightly higher : A3 rather than A%.
Special rules. Authors of papers intended for the Bulletin or Trans-
actions should give special attention to the following rules. A reason-
ably strict observance of these rules will help keep printing costs
down, prevent an uneconomical use of space, insure a certain uni-
formity of appearance, and will greatly simplify the task of final prep-
aration of manuscripts for the printer.
1. Use only such characters as the printer has in stock. The list
in Appendix C includes all special types now available for the Bul-
letin and Transactions. Additions to this list are made from time to
time. Note particularly what letters and numbers are available in
sub-superscript (inferior to superior) and super-subscript positions.
2
These remarks are not intended to imply that symbols such as A\ must never
be used in the solid text, rather that the author should exercise a measure of discretion.

License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use

 4 AMERICAN MATHEMATICAL SOCIETY

Note also that many symbols are not available in footnote size. For
this reason, if for no other, complicated formulas should not be put
into the footnotes. In this connection it might be remarked that an
excessive number of footnotes sometimes gives the undesirable im-
pression that the paper is being "written in the footnotes." Careful
organization, however, always makes it possible to reduce the number
and length of footnotes, even to eliminate them entirely.
2. Do not put into the solid text any formula which would require
spreading (uneven spacing) of lines.
3. In the solid text, write ]£?-1î II"-1î nmn-»ooï fa- In the first two
examples the n should be directly over the i. In the integral the b
should be directly over the a; it is not feasible to place the a nearer
the integral sign since the latter is on a rectangular face:[/]. In dis-
played formulas, the limits should occupy only one line. For ex-
ample, write

instead of

4. In the text, replace e( } by exp ( ) if the expression in the

parentheses is complicated.
5. Use Cm,n (binomial coefficient) instead of (™).
6. In the text, write ~ e , in display ~ .
7. Use G to mean "is a member of," € for epsilon. The symbol
for "is not a member of" is (£.
8. Use fractional exponents in place of radicals. A fraction occur-
ring as subscript or superscript should be formed with a solidus, for
example 2 / 3 . In display, the number of lines occupied by a fraction
should be as small as possible. Frequently the number of lines can
be reduced by a simple rearrangement. For example:
1
cos —
X # COS (1/ff)
COS \L/ X)
r»o -n Vw/i Tirriffûr» •
can be written •
(a + b/x)1'2
r %

9. Long formulas, or expressions which might break awkwardly

at the end of a line of text, should be displayed.
10. In numbering sections, theorems, displayed formulas, and so
on, the author should adopt a system which is simple and self-
explanatory. A system based on such symbols as II.3.6.la is unneces-
sarily complicated.

License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use

 A MANUAL FOR AUTHORS 5

11. Footnotes should be assembled at the end of the paper, typed

double spaced, and numbered with arabic numbers. The footnote
symbols currently in use are illustrated as follows: "It was proved 3
that • • • " (Bulletin); "It was proved (3) t h a t • • • » (Transactions).
A footnote symbol should not follow a mathematical symbol. Foot-
notes should be numbered in the order in which they appear, begin-
ning with 1 and continuing throughout the paper. Do not use any
symbol at all for the footnote "Presented to • • • " which frequently
appears on the title page.
12. Mathematical works which are referred to in the text may be
listed at the end of the paper under the heading "Bibliography" or
"References." The following examples illustrate the two methods of
arrangement and enumeration used in the Bulletin and Transactions:
Form I
BIBLIOGRAPHY for R E F E R E N C E S ]
1. R. R. Roe, Canonical systems, Amer. J. Math. vol. 60 (1938) pp. 1042-1100.
2. , An elementary solution of the fundamental problem of mathematics, J. of
Metamath. vol. 17 (1939) pp. 1-100.
3. D. D. Doe, The theory of topological hypergroups, New York, 1940.
Form II
BIBLIOGRAPHY [or R E F E R E N C E S ]
R. R. ROE
1. Canonical systems, Amer. J. Math. vol. 60 (1938) pp. 1042-1100.
2. An elementary solution of the fundamental problem of mathematics, J. of Meta-
math. vol. 17 (1939) pp. 1-100.
D. D. DOE
1. The theory of topological hypergroups, New York, 1940.
2. A generalization of the generalized Doe-differential, Duke Math. J. vol. 48 (1940)
pp. 1100-1200.

The following examples illustrate the various methods of referring

to the bibliography. The basic marks of enclosure are brackets: [ ].
With a small amount of care, the use of parentheses as additional
marks of enclosure can be kept at a minimum and the appearance of
the printed page thereby improved.
For form I: [l, Theorem 3.7]; • • • proved by Roe [2, Lemma
4] • • • ; (see [2, Lemma 4]).
For form I I : [Roe 1, Theorem 3.7]; • • • proved by Roe [2, p.
100] • • • ; (cf. Roe [2, p. 100]).
The use of abbreviations for the names of the more obscure scien-
tific journals often creates difficulties for readers and librarians. The
Bulletin and Transactions, therefore, make it a rule to print such

License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use

 6 AMERICAN MATHEMATICAL SOCIETY

titles without any abbreviations. Cooperation on the part of the authors

would diminish greatly the work involved in this connection. The names of
the standard journals, however, should be abbreviated. Appendix B
contains a list of abbreviations which have been adopted by the
Bulletin and Transactions.
Preparation of the manuscript. A compositor is not a trained
mathematician and cannot take the responsibility of altering so much
as the position of a single dot in the manuscript. It is obvious, there-
fore, that the manuscript, when it is sent to the printer, must be
exact in every detail. Since it is impossible to produce a manuscript
in which all characters are facsimiles of printed characters, there will
be many details which can be made clear to the printer only by spe-
cial indication such as underlining or marginal notes.
The task of preparing the manuscript for the printer is generally
shared between author and editorial staff. In the case of the Bulletin
and Transactions, indications concerning routine matters of style,
such as the spacing of material, the choice of type in titles, paragraph
headings, and so on, are inserted by an editorial assistant. All indica-
tions of a non-routine nature must be made by the author and must
be complete and explicit. The most important rules in this connec-
tion are the following :3
1. The manuscript should be typewritten. The text should be
double spaced and displayed material should not be crowded. Plenty
of space should be left for the characters which are to be filled in by
hand. It is well to make at least one carbon copy of the manuscript,
but the original—not the carbon copy—should be sent to the editor.
2. Subscripts or superscripts may be filled in by hand ; if typed they
should be a half-space above or below the line.
3. Handwritten characters should be made with great care and
each character must be distinct in shape from every other. In cases
where the author's intention may be in doubt, identification, such as
"Greek chi," should be written in pencil in the margin. Capital letters
may be indicated by "cap," small letters by "I.e." (lower case). Letters
and symbols which are particularly liable to confusion are :
oOO ymv pp sS5 i , l' hkicKK
zZ2 xPxx X 4 ell VU G<?
4. German and script letters may be handwritten if they can be
made with sufficient accuracy. A far more satisfactory procedure is to
type the corresponding Roman letters, then to underline or encircle
3
See also rules 3-8 in the section Special Rules.

License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use

 A MANUAL FOR AUTHORS 7

each letter with colored pencil, say green for German, blue for
script. In fact, all alphabets other than the Roman and Greek are
best treated in this manner. The meaning of the colored marks should
be stated at appropriate places in the margin or on a separate sheet
of "instructions for the printer."
5. No mark meaning "Greek alphabet" should be attached to the
symbols ^ (summation), U (product), d (in partial derivatives).
Notice the difference in the size and style between the first two sym-
bols and the Greek capital letters 2, II.
6. Italic type is indicated by underlining, preferably with a blue
pencil. In the Bulletin and Transactions, the text of a theorem or
corollary is always in italics and the necessary indications may be
left to an editorial assistant.
7. Omissions should be denoted by exactly three slightly raised
dots, such as ab • • • c. If commas are required, write ai, • • • , a n or
#i, #2, • • • , never ah • • • an or ah a2 • • • . A handwritten comma
must be carefully made so that it will not be taken for subscript 1.
8. The number one is ordinarily typed 1. The letter "ell" is denoted
by 1 (the dash may be written). Experience shows that the use of the
solidus (/) in place of the dash may lead to errors in typesetting.
9. Drawings and diagrams should be made with particular care
since they cannot be changed or corrected in galley proofs or page
proofs. Drawings should be made in black India ink on Bristol board
and should be two or three times larger than they are to be when
printed. When inserting letters or symbols, allowance should be
made for the reduction in size. For drawings which are to appear in
the Bulletin or Transactions, all letters and symbols should be drawn
lightly in pencil since they will eventually be replaced by type letters
and symbols. The reproduction of even the simplest drawing involves
considerable extra expense, and the author, therefore, should include
no drawing which is not really essential for a clear understanding of
his paper.
10. A displayed formula should be numbered only if referred to
elsewhere in the text. If displayed formulas are numbered indis-
criminately, it is impossible to combine several formulas into a single
line or to run an occasional formula into the text to facilitate the
breaking up of galleys into pages.

Proof reading. It is important that galley proofs be carefully read

and corrected by the author, since it is only the author who can detect
errors which are due to an imperfect manuscript.
Alterations in type which has been set are delicate and costly to

License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use

 8 AMERICAN MATHEMATICAL SOCIETY

make. The author should call for only such changes as are really
essential. The insertion or deletion of a comma, for example, should
be based on a compelling reason and should not be merely a matter
of whim. If a number of consecutive words or symbols must be de-
leted, they should if possible be replaced by words or symbols oc-
cupying the same amount of space. If this is not done, the com-
positor often must tear down and reset a number of preceding or
following lines in order to make the necessary adjustments. This is
not only costly but makes possible the introduction of new errors.
Galley proofs of papers for the Bulletin or Transactions are read
in the office of the American Mathematical Society before they are
sent to the author. During the reading, there may arise questions
which can be answered only by the author. Such questions are written
in the margins, and it would seem unnecessary to point out that they
should be answered were it not for the fact that they are frequently
ignored.
The list in Appendix A contains most of the standard signs used in
this country in correcting proof. A consistent use of these signs by
authors would simplify greatly the work of printer and editorial staff.
In any case, authors should make sure that their corrections are clear
and explicit and should refrain from using elaborate "systems" of
their own invention. Joint authors should be particularly careful not
to use two non-equivalent systems of signs in a single set of galley
The example in Appendix A illustrates the standard method of
using the proof reading signs. It is important to note that all indie-< -
tions for corrections must appear in margins and for each correction
a corresponding mark must appear in the text. The signs in t i e
margins should retain the order of the corresponding errors in the
line and should be separated from each other by |.

License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use

 APPENDIX A
SIGNS USED IN CORRECTING PROOFS
jjl Delete; take out y Change broken letter
O Close up stet Let it stand as set
A Insert Let it stand as set
7F= Insert space W.J Wrong font, size or style
r n
Raise l.C. Lower case, not capitals
i—i Lower À<rm. Use Roman letter
[ Move to left f Use black type letters
] Move to right O Period
II Straighten type line at side
^ Comma
of page
v* Apostrophe
/ / S t r a i g h t e n lines
(§, Paragraph V Superior figure

c e n t e r Put in middle of page or •/^Inferior figure

line
=/ Hyphen
c
<> Transpose
<AG> Use small capitals
M, Transpose
capsUse capitals
^ ) Turn inverted letter right
side up ital.Use italics

License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use

 10 AMERICAN MATHEMATICAL SOCIETY

EXAMPLE SHOWING THE USE OF PROOF READING SIGNS

of Bi and didj divides dxdpÇ^y. Hence B_ is reducible to the ftrjStm /f^JJMlJ
(11.5) with diagonal terms du d\di, • • • , d\d'p which proved (11.4).
$y£. 12. Groups with(a)finite number of generators. We shall discuss
certain properties of these groups culminating in the basic product

Ph decompsition (12.5).
(12.1) D E F I N I T I O N Let B= {gu • • . , gn}} B'= {g{, • • .A«» } be A.C.fî
*# two sets of elements/of G containing the same number n of elements. By
A a unimodulaf transformation T:B—>B' is meant a system of gelations Ac
Z (l|.2) g( = ^2 aagj> || a ùj| unimodular.
<§ /The following proposition shows in how natural a manner unimodular
transformations make their appearance in the theory of gro@ps with /lOm,
finite bases.
, (12.3) Let G be a group with a finite base B = {gi, • • • , g nA . In order \
wWt£\/ that Bf = {gi , • • • , gn} be a base for G it is necessary and sufficient
that B' be obtainable from B by a unimodulaj transformation. ^
For any given set B' — {g{, • • • , gn' } of elements of G there exist
relations
9*W(12.4) ü=?,ciSgh C = |klU O
A necessary and sufficient condition in order that A {g/ } be a^base is AnW ~ 4/
Xc% that the (pj be expressible as linear combinations of the g/, or that
there exist relations
(12.5) ]«*=E^«/, D = \\dij\\. I CZKlk
[_ L From this follows
gi = Z ) dijCjkgk/ O

v^ Hence since B is a base we must have ff C = U

n (This matrix relation yields | ^ | | T | | C | = 1 , and since the deter- A/Mvl/Hs
minants are integers we must have \C\ = ± 1. Thus in order that B'
be a base C must be unimodular, or the condition of (12.3) must be
F £ fulfilled. Conversely, if>fis fulfilled, C i s un imodular and (12.5) (li.l)
holds with D = CÇj), from which it follows readily thatj>_' is a base. ^lùU^

License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use

 A MANUAL FOR AUTHORS 11

THE PRECEDING PASSAGE PRINTED WITH ALL CORRECTIONS MADE

of Bi and didp divides d\dv+x. Hence B is reducible to the form (11.5)
with diagonal terms dh dxd{, • • • , d\d'p which proves (11.4).
12. Groups with a finite number of generators. We shall discuss
certain properties of these groups culminating in the basic product
decomposition (12.5).
(12.1) DEFINITION. Let B={gu • • • , g n } , B'= [gi 9 • • • , gi ) be
two sets of elements of G containing the same number n of elements. By
a unimodular transformation r:B-*Br is meant a system of relations
(12.2) gi = ^aijgjy \\<iij\\ unimodular.
The following proposition shows in how natural a manner unimodu-
lar transformations make their appearance in the theory of groups
with finite bases.
(12.3) Let G be a group with a finite base B = {gi, • • • , gn}. In order
that B' = {g{, • • • , gn } be a base for G it is necessary and sufficient
that B' be obtainable from B by a unimodular transformation.
For any given set B' — {g{, • • • , gn' } of elements of G there exist
relations.
(12.4) gl=Hciigi, C = \\c4.
A necessary and sufficient condition that Bf~{gl\ be a base is
that the gj be expressible as linear combinations of the gi, or that
there exist relations
(12.5) gi= Z,diigj, D = \\d4.
From this follows
gi = ] E dijCjkgk.
Hence since B is a base we must have DC=1. This matrix relation
yields \D\ - I C\ — 1, and since the determinants are integers we must
have | C\ = ± 1 . Thus in order that B' be a base C must be uni-
modular, or the condition of (12.3) must be fulfilled. Conversely, if
(12.3) is fulfilled, C is unimodular and (12.5) holds with D = C""1, from
which it follows readily that Bf is a base.

License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use

 APPENDIX B
ABBREVIATIONS USED IN
T H E BULLETIN A N D TRANSACTIONS
Abhandlungen aus dem Mathematischen Se- Abh. Math. Sem. Hansischen Univ.
minar der Hansischen Universitât
(formerly . . . Hamburgischen Universitât)
Academy of Sciences of the USSR. Journal of Acad. Sci. USSR. J. Phys.
Physics
Acta Universitatis Szegediensis. Acta Scientia- Acta Univ. Szeged.
rum Mathematicarum
Acta Mathematica Acta Math.
American Journal of Mathematics Amer. J. M a t h .
T h e American Mathematical Monthly Amer. Math. Monthly
American Mathematical Society Colloquium Amer. Math. Soc. Colloquium Pub-
Publications lications
American Statistical Association Bulletin Amer. Statist. Assoc. Bull.
Annales de la Faculté des Sciences de l'Univer- Ann. Fac. Sci. Univ. Toulouse
sité de Toulouse pour les Sciences Mathé-
matiques et les Sciences Physiques
Annales Scientifiques de l'École Normale Ann. École Norm.
Supérieure
T h e Annals of Mathematical Statistics Ann. Math. Statist.
Annals of Mathematics Ann. of M a t h .
Atti della Reale Accademia delle Scienze di Atti Accad. Sci. Torino
Torino. Classe di Scienze fisiche, Mate-
matiche e Naturali
II Bollettino di Matematica Boll. Mat.
Bulletin de l'Académie des Sciences de l'URSS. Bull. Acad. Sci. URSS. Sér. Math.
Série Mathématique
Bulletin de la Société Mathématique de France Bull. Soc. Math. France
Bulletin de la Société Physico-Mathématique Bull. Soc. Phys.-Math. Kazan
de Kazan et de l'Institut de Mathématiques
et Mécanique
Bulletin des Sciences Mathématiques Bull. Sci. Math.
Bulletin Mathématiques de l'Université de Bull. Math. Univ. Moscou
Moscou. Série Internationale
Bulletin of the American Mathematical Society Bull. Amer. Math. Soc.
Bulletin of the Calcutta Mathematical Society Bull. Calcutta M a t h . Soc.
Commentarii Mathematici Helvetici Comment. M a t h . Helv.
Compositio Mathematica Compositio M a t h .
Comptes Rendus (Doklady) de l'Académie des C. R. (Doklady) Acad. Sci. URSS.
Sciences de l'URSS
Comptes Rendus Hebdomadaires des Séances C. R. Acad. Sci. Paris
de l'Académie des Sciences
Doklady Akademii Nauk SSSR
See Comptes Rendus (Doklady) de l'Aca-
démie des Sciences de l'URSS
Duke Mathematical Journal Duke M a t h . J.
Fundamenta Mathematicae Fund. Math.
Izvestia Akademii Nauk SSSR
See Bulletin de l'Académie des Sciences de
l'URSS

12
License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use
 A MANUAL FOR AUTHORS 13

Jahresbericht der Deutschen Mathematiker Jber. Deutschen Math. Verein.

Vereinigung
Japanese Journal of Mathematics. Transactions Jap. J. M a t h .
and Abstracts
Journal de l'École Polytechnique J. École Polytech.
Journal de Mathématiques Pures et Appliquées J. Math. Pures Appl.
Journal fur die reine und angewandte Mathe- J. Reine Angew. Math.
matik
The Journal of Symbolic Logic J. Symbolic Logic
Journal of the Faculty of Science. Imperial Uni- J. Fac. Sci. Imp. Univ. Tokyo.
versity of Tokyo. Section I. Mathematics, Sect. I
Astronomy, Physics, Chemistry
The Journal of the Indian Mathematical J. Indian Math. Soc.
Society
T h e Journal of the London Mathematical J. London Math. Soc.
Society
Koninklijke Nederlandsche Akademie van Neder. Akad.. Wetensch.
Wetenschappen. Indagationes Mathematicae
ex Actis Quibus Titulus. Proceedings of the
Section of Sciences (sometimes referred to as
the "Amsterdam Proceedings")
Matematiëeskij Sbornik
See Recueil Mathématique
Mathematische Annalen M a t h . Ann.
Mathematische Zeitschrift M a t h . Zeit.
Nachrichten von der Gesellschaft der Wissen- Nachr. Ges. Wiss. Göttingen
schaften zu Göttingen. Mathematisch-Phy-
sikalische Klasse. Neue Folge. Fachgruppe I.
Nachrichten aus der Mathematik
Periodico di Matematiche Period. M a t h .
Philosophical Transactions of the Royal Society Philos. Trans. Roy. Soc. London.
of London. Series A. Mathematical and Phys- Ser. A.
ical Sciences
Proceedings of the Cambridge Philosophical Proc. Cambridge Philos. Soc.
Society
Proceedings of the Imperial Academy Proc. Imp. Acad. Tokyo
Proceedings of the London Mathematical Proc. London M a t h . Soc.
Society
Proceedings of the National Academy of Sci- Proc. Nat. Acad. Sci. U.S.A.
ences of the United States of America
Proceedings of the Royal Society. Series A. Proc. Roy. Soc. London. Ser. A.
Mathematical and Physical Sciences
T h e Quarterly Journal of Mathematics Oxford Quart. J. M a t h . Oxford Ser.
Series
Recueil Mathématique. Nouvelle Série Rec. M a t h . (Mat. Sbornik) N.S.
Rendiconti del Circolo Matematico di Palermo Rend. Circ. Mat. Palermo
Sitzungsberichte der Berliner Mathematischen Berlin M a t h . Ges. Sitzungsber.
Gesellschaft
Sitzungsberichte der Preussischen Akademie Preuss. Akad. Wiss. Sitzungsber.
der Wissenschaften
T h e Tôhoku Mathematical Journal Tôhoku Math. J.
Transactions of the American Mathematical Trans. Amer. Math. Soc.
Society
License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use
 APPENDIX C
A LIST OF SIGNS AND SPECIAL CHARACTERS
AVAILABLE FOR THE
BULLETIN AND TRANSACTIONS

Lightface Greek—a p y . . (all) A B T . . . (all).

* Lightface Greek Superiors—A and a
* y • • • (all except o).
* Lightface Greek Inferiors—A A s Q and « p y . . . (all except o).
* Boldface Greek—a g Y & e Ç n O K à y v Ç r c p d f c j x a A n and Q.
* Lightface German—SI 93 S . • . (all) a b c . . . (all). No superiors
or inferiors.
* Lightface dashed German—p â S8 Û ^ © ©
* Boldface German—b 21 » $
Script (special font) zA <B Q . . . (all). No lower case manu-
factured.
* Hebrew—SI ^ ^ 3 C troublesome to use. No superior or inferior.
• Dashed Italics—A â 2? b C c . . . (all)
• Tilde Italics—Â âeNnOfuysMTÊK
*k Tilde Greek— â q î ô û w
• Dashed Greek—a iHyorjeifiVpuT'kfaiple
ir Dotted Italic—à dèëgïmnqffùilxxy^zz
• Dotted Greek—ë rirjUOCtt^cb (single dotted P y d <t>; double
dotted 7 readily available).
H M
• Superiors—™
JÇ I n f e r i o r s ,» n n r à x y z Ï 2

Signs—Superior—- + - « ± < > M . . ' / i

Signs—Superior to Superior—""

Signs—Inferior— + « o ± n < > [ ] . . / i '

Signs—Inferior to Inferior— + _ ,

Special—Superior to Superior—l
Special—Inferior to Inferior—, f , 4 0 , h <i k l m n p r 9

* Additional characters readily available at small cost.

* Matrices for additional characters are made upon special orders. Nor-
mally they would necessitate a delay of from four to eight weeks and average
expense of $4.50 per matrix. However, because of wartime conditions delivery
dates are uncertain and prices subject t o change without notice.

14
License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use
 No. 6 8 10 12 No. 6 8 10 124

1. - V V V V 40. v V V
2. X V V V V 41. D V
3. X V 42. c V
4. -f- V V V V 43. * V
s. + V V V V 44., u V
6. + V 45. n V
7. ± V 46. .\ V
8. ± V V V v 47. A V
9. + V 48. A V
10. = V V V V 49. A V
11. = V V 50. A V
12. = V V 51. Z V V
13. ^ V V 52. A V
14. s V 53. • V
15. ^ V V V 54. H V
16. * V V 55. • V
17. s V V 56. • V
18. jk V V 57. GO V
19. + V 58. O V
20. oo V V V 59. BJ V
21. co V V 60. O V
2 2 . ex: V 61. O V V
23. =0= V V 62. © V V
24. ~ V V 63. © V
25. ~ V 64. © V
26. « V 65. Q V
27. < V V V 66. O V
28. ^ V 67. 0 V
29. > V V V 68. ® V V
30. ^ V 69. 0 V
31. ^ V V V 70. 25 V
32. ^ V V V 71. V V
33. > V V V 72. -* V V V
34. < V 73. <- V V
35. A V V 74. -»- V V V
36. V V V 75. <=± V
37. A V 76. ±3: V
38. « V 77. *~s V
39. <£ V 78. T± V

4
These figures refer to the size of type. T h e text of Bulletin and
Transactions is 10 point type and footnotes 8 point type.

License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use

 > >

> > > > > > > >> - > > > > > > > > > >

oo > > > > > > > > - > > >

vo > > > > > > >

. . .."v/s^*8o8«£-Ye & c ), „unuirwwcD

a O H C N O \ O H ( N f 0 ^ i O \ O N 0 0 O \ O H ( N C 0 ^ l O \ Ö N 0 0 a O H C S f 0 ^ N 0 0

> > > > > > > > " >

> > > > > > > > > > > > > > > > > > > > > > > > > > > > >
°
00 > > > > > > >> > > > > > > >>

vO > > > > > > > > > >

6
JK/ ^ ^ — = \ / / # > H ±11 mm>»<. ., sSSRWo,.
a d H C N ^ ^ ^ V O N O Ö O N O H O H N C O ^ ^ v d N O O O N Ó H C S ^ ^ ^ V O N o d

License or copyright restrictions may apply to redistribution; see https://www.ams.org/journal-terms-of-use