文档介绍：arXiv:hep-th/9609122 v2 18 Oct 1996
itational
The
Septem
action
1
ON
The
Researc
relation
of
b
er,
con
quan
M
h
FLUX
1996
-theory
tribution
of
supp
tization
AND
M
Scho
orted
-theory
,
including
ol
not
la
in
of
QUANTIZA
w
THE
Olden
part
to
kno
Natur
for
the
wn
the
b
one-lo
L
y
al
ane,
E
NSF
previously
an
EFFECTIVE
8
Scienc
×
tisymmetric
Edw
op
Princ
Gran
E
8
and
es,
ard
heterotic
t
eton,
.
PHY-9513835.
Institute
Chern-Simons
Witten
When
TION
tensor
NJ
string
it
1
hep-th/9609122,
08540,
for
is
field
included,
IN
A
greatly
dvanc
A
con
USA
of
CTION
M
M
tributions,
e
d
facilitates
-theory
the
-THEOR
Study
IASSNS-HEP-96/96
lo
w
con
energy
is
the
tains
w
ell-defined.
analysis.
effectiv
Y
a
gra
v-
e
1. Introduction
This paper is devoted to explaining some topological points concerning eleven-
dimensional M-theory. Roughly speaking, we will explain one new physical effect and
show how it enters in re-interpreting, or avoiding, several potential anomalies.
The new physical effect involves the three-form C of eleven-dimensional supergravity
and its field strength G = dC. It has been believed that G is constrained precisely by a
flux quantization law, which says that if G is correctly normalized its periods are integer
multiples of 2π. As we will see, this is not quite correct.
For the precise statement, recall first that in M-theory we can assume that the space-
time manifold Y is a spin manifold since spinors exist in the For Y a spin manifold,
the first Pontryagin class p1(Y ) is divisible by two in a canonical way. We set
p (Y )
λ(Y ) = 1 . ()
2
λ in turn may or may not be divisible by two, depending on the topology of Y .
It turns out that the object that must have integral periods is not G/2π but G/2π
−