The Arithmetic of Elliptic Curves : W 1. Introduction
The Arithmetic of Elliptic Curves : W 1. Introduction
The Arithmetic of Elliptic Curves : W 1. Introduction
23, 1 7 9 - 2 0 6 119741
9 by Springer-Verlag 1974
w 1. Introduction
After curves of genus 0 (e.g. lines and conics in the plane) come
curves of genus 1, or "elliptic" curves (e. g. plane cubics or intersections
of quadric surfaces in three-space). Elliptic curves are the first examples
of abelian varieties. Their points of finite order give the first non-trivial
examples of 6tale cohomology groups. The action of Galois groups on
these leads both to the classical theory of complex multiplication as
well as to systems of non-abelian extensions which may contain clues
to non-abelian class field theory. Elliptic curves are intimately connected
with the theory of modular forms, in more ways than one.
In the early sections I have tried to give a brief introduction to the
fundamentals of the subject, using explicit formulas to by-pass chunks
of general theory when possible. The later sections are a survey of recent
work with emphasis on three main topics: (1) The problem of rational
points, the Shafarevitch group, and the conjecture of Birch and Swinner-
ton-Dyer. (2) Modular curves and Weil's astounding idea that every
elliptic curve over the rational field is "'modular". (3) Serre's theorem
that the Galois groups obtained from points of finite order on elliptic
curves are "as big as possible". I hope to be able to convey some idea
of these advances here, illustrating them by numerical examples discussed
in the last section.
w 2. Weierstrass Models
In these lectures we will use the term elliptic curee to mean an abelian
variety of dimension 1, or, what is the same, an irreducible non-singular
projective algebraic curve of genus 1 furnished with a point 0, the
origin for the group law. Any such curve E, defined over a field K,
has a plane cubic model of the form
y 2 + a 1x y + a 3 y = x 3 + a 2 x2+a4x+a6 (1)
* Revised version of a survey, distributed in conjunction with the Colloquium Lectures
given at D a r t m o u t h College, Hannover, August 29-September 1, 1972 (77th summer
meelmg of the A.M.S.).
The author thanks N.S.F. for its support, and J.-P. Serre for his secretarial and editorial
help
13 invenUones math, Vol 23
t80 J.T. Tate
where x and y are coordinates in the affine plane and the coefficients a,
are in our g r o u n d field K. W e call (1) a Weierstrass equation because in
characteristics + 2 , 3 we can replace x and y by
a~ +4a2
~o=x+ 12 ' ~~
In the projective plane the curve (1) has a unique point at infinity which
we call 0 and take as the origin for the g r o u p law. It is a point of inflection
with the line at infinity as tangent; the other lines through 0 are the
"vertical" lines, x = const.
Given an Eq. (1), i.e., given five elements al, a2, a3, a,, a6 in K, we
define associated quantities b,, c,, A, and j by the following formulas.
T h e subscripts indicate weights. The quantity A, of weight 12, is called
the discriminant; its non-vanishing is necessary and sufficient for the
curve (1) to be non-singular, hence elliptic.
b2=aZ+4a2 c4--b~-24b~(=12g2)
b4=ala3+2a 4 %= -b3+3662b4-216b6(=21693)
b6=aZ+4a6 A=-b2bs-8b3-27b2+9bzb4b6(=g~-27g~ (2)
2 llZ2 9 C34 .
b 8 = bz a6 - al a3 a+ + a 2 a3 - , j= ~ = 1728 J).
and the transformation rules for the a's, b's, c's, A and j are:
uZ a ' 2 = a 2 - s a l + 3 r - s z u6 b'6=b6 + 2r b4 + r2 b2 + 4 r 3
u3a3=a3+ral+2t u4 b4=b4+rb2+6r 2
u6 a ' 6 = a 6 + r a 4 +rZ az + r 3 - - t a 3 - - r t a ~ - - t 2
If two elliptic curves E and E' are isomorphic, then j = j ' ; the
converse is true over an algebraically closed field K, as is not hard
to check using the formulas above.
If P is a point on E we denote the corresponding prime divisor
by (PI. We let D ~ D ' denote linear equivalence of divisors. The g r o u p
law on E, which is commutative and denoted by + , is determined
intrinsically by the rule
n n tl n
i= i i= 1 l= l i= 1
together with the fact that 0 is the identity element, i.e., O + P = P for
all P. On a Weierstrass model (in fact on any plane cubic model in
which 0 is a point of inflection) it follows that for three points P, Q, R
on E we have P + Q + R = O e * . ( P ) + ( Q ) + ( R ) ~ 3 ( O ) c * ( P ) + ( Q ) + ( R ) is
the intersection cycle of a line with E. (This means simply P, Q, and R
collinear if P + Q : t = R + P ; if P=Q:4=R, it means that the tangent to E
at P meets E also in R; and if P = Q = R it means that P is a point of
inflection.)
It is easy to m a k e the addition law on a Weierstrass model (1) very
explicit. Let P~= (x 1 , Yl) and P2 =(x2, Y21 be two points + 0 on the curve.
Then P1 + P2 = 0 ~ xl = x2 and y~ + Y2 + al x + a 3 = 0. Otherwise, we find
the sum P~+Pz=P3=(x3,Y3) as follows: Let y - - R x + v be the line
through P~ and Pz (tangent to E at Px if P~= Pz). Then x 1, x 2, x 3 are
the roots, with correct multiplicities, of the cubic equation in x obtained
by substituting 2 x + v for y in (1). Hence x3 can be calculated from
x I -Jvx 2 -t- x 3 = 2 2 -~-aa 2 -- a z , then 1'3 from - a t x3 - a3 - Y3 =)~ x3 + v. Of
course, if the coordinates of the points P~ and P2 lie in the g r o u n d
13"
182 J.T. Tate
y--o~ 1 X - - [ ] 1
P = (x, y) ~ 19)
y--o~2 X - - [ ~ 2
x 1 z 1
z=----, w=----, so x = - - , ). . . . . . . (11)
y y w w
= 2 3 ( 1 + A 1 z + A 2 z 2 + '' "),
dy/dz - 3 z- 4 +...
=
3x2+2azx+a,,--aly -3z-4+..
y, l _ p-1 (21)
s~i,..... g~l~ [AutE] 24
is indeed a free module of rank 2 over the ring Z/m Z. Moreover it has
a homological interpretation, as the 1-homology of E with coefficients
modm, because L can be identified with HI(C/L,Z). In view of
Theorem 1, if E is an elliptic curve over any field K and Ks a separable
algebraic closure of K, it is reasonable to view E,,,(Ks) as the 1-dimen-
sional homology of E with coefficients in Z/m Z, which, in fact. it is
in the 6tale cohomology theory of M. Artin and Grothendieck. Taking
m = l", 1 a prime 4: char(K), and passing to the projective limit as n-+ ~,
we get Weil's l-adic space
It- K=C,
which is a free module of rank 2 over the ring Z l of /-adic integers, and
plays the role of H1 (E, Zt).
Going back to the case K = C we note that the intersection pairing
of 1-cycles induces an alternating form on L=H1 with values on Z,
making A z L~_ Z. The algebraic analog is WeiFs '" e,,-pairing"
fe/k q ) (27)
~E/k(S)= (1--q-~)(1 _ql-~)"
where
fz/k(X)=detll - ~ X ) = 1 - ( T r n ) X + q X 2 = ( 1 -TtX)(1 - ~ X ) , (28)
and T h e o r e m 2 implies that its zeros are on the line Rels)= 89
188 J.T Tate
Then (27) holds in all cases, E singular or not. So does the relationship
w 6. Local Fields
The group E(C) of points on an elliptic curve over the complex field
is a connected compact complex Lie group of complex dimension 1, so
is isomorphic to the product of two circles.
The group E(R) of points on an elliptic curve over the real field is a
compact real Lie group of dimension 1 with one or two components
(according as d < 0 or A >0), and is therefore isomorphic to the circle
group or to its product with a group of order 2.
Suppose now that K is a field complete with respect to a discrete
valuation, v. Let R denote the ring of integers in K. p a prime element
in R, and k = R / p R the residue field. Assume v normalized so that
v (p)= 1. Let E be an elliptic curve over K. There exist Weierstrass Eqs. (1)
for E with coefficients aiER. A m o n g all such, choose one for which v (A)
is minimal. We call such an equation a minim~! Weierstrass equation for E.
Using the Eqs. (7) it is a simple matter to check that any two such minimal
equations are related by a transformation (5) with r, s, t~R and u invertible
in R. This will mean that the following considerations are essentially
independent of our choice of minimal Eq.(l). In particular, the Weier-
strass curve/~ over k got by reducing (1) modulo the prime in R is unique
up to a transformation of the form (5) over k.
A point P in projective n-space over K can be represented by a set of
coordinates (Xo . . . . . x,) such that xi~R all t and xi invertible in R for
some i (i.e., such that 0 = M i n v(xi)), and then on reducing the x~ mod pR
we get a point/5 = (Xo . . . . . ~,) in projective n-space. For n = 2 this "reduc-
tion m a p " P~--~P from P2(K) to Pz(k) obviously carries E(K) to /~(k).
We put
E o (K)= {PeE(K) IPeE.~ (k)}, (31)
E,(K)= {PeE(K) lP = 0 } . (32)
Recall (cf. w1) that/~,~ denotes the non-singular part of/~, and is a group.
The Arithmetic of Elhptic Curves 189
and
1
J-- - + 7 4 4 + 196884q+..-. (34)
A q
These make sense for qEK with 0 < v ( q ) < ~ , (i.e., l > l q l > 0 ) and, in
our non-archimedean field K, the relation (34) gives a bijection between
the set of all such q and the set o f j e K with v(j)<0 (i.e.,jCR). Any such q
generates an infinite cyclic discrete subgroup qZ of the multiplicative
group Gin. In the classical case, C*/q z is an elliptic curve with invariant i
given by (34). The same is true over K!
Theorem 5. For q as above, Gm/qZ=Eq is an elliptic curve over K.
It has a minimum Weierstrass equation with c4, A, and j as above. It is
characterized, up to K-isomorphism by the .fact that it has the given
j-invariant, together with the fact that its reduction is of split multiplicative
type (ie., Eq has a node with tangents rational over K).
I found this theorem in 1959 and would like to apologize for never
having officially published it. The most complete reference for it at
present is [51]. Mumford [38] has found a very non-obvious generaliza-
tion to curves of higher genus. For the generalization to abelian varieties,
see McCabe [88] and Raynaud [96].
The Arithmetic of Elhptlc Curves 191
is biadditive on E ( Q ) x E(Q).
Now it is straightforward to show that if, for some integer m > 2 ,
the points Pi represent all cosets of mE(Q) in E(Q), and if h o >h(P,) all i,
then E (Q) is generated by the set of points P such that h (P) < h o . Since that
set is finite for any h o, Theorem 1 will follow if we prove that E(Q)/mE(Q)
is finite; that is the second part of the proof.
To do this, one produces an exact sequence
E(Q)- '~ , E ( Q ) ~ S ~"~-,111,. - . 0 , 136)
in which S ~"~, the Selmer group for m, is finite and effectively computable,
and in which///,, is the set of elements of order dividing m in the Shafare-
The A r i t h m e t i c of Elhptic C u r v e s 193
fitch group 111 of E/Q. This takes care of the finiteness statement, but does
not yet give an effective method of constructing generators. The trouble
is that there is no known method of computing///,,! However, for each
integer n > 1 there is a commutative diagram
E(Q) , S~ - - + / / / , , , >0
any affine model for E there are only finitely many points with i~tegral
coordinates, which was non-constructive for so long, has recently been
made constructive by A. Baker and J. Coates ~3]. They prove
Theorem T. Let F(x,y) be an absolutely irreducible polyncmlial o[
degreee n with integer coefficients having absolute t,alues at most M such
that the curve F(x, y ) = 0 has genus 1. Then all integer solutions F(x, 1')=0
satisfy
Max(Ixt, lyl)<exp exp exp(2M) w"~~ (39)
Incidentally, their method of proof is to reduce to the Welerstrass
equation case, which had been treated earlier by Baker, with a somewhat
better bound. But this problem of integral points involves completely
different concepts from those we are discussing and we mention it only
in passing.
Let E (Q)~o,sdenote the torsion subgroup ofE (Q). In view of Theorem 6,
E(Q)to~~ is finite and
E (Q)'-- Z ~• E{Q)tor,, (40)
where r is a certain integer > 0 called the rank of E over Q. In all known
explicit examples the rank is quite small.
The carve ) , 2 = x 3 + A x 2 + B x has rank > 7 (very probably 7) for
A = - 3 . 5 . 11.13.17.19.23.29.31.37
with B---1 692602, B = 2 8 4 3 7 3 8 or B=2877338.
These curves were recently discovered by C. Pomerance and D. F!.
Penney [95] by computer search. I don't know of any explicit example
of rank >7. However, by spezialisation arguments, N6ron [94] has
shown there must be elliptic curves over Q with rank >11, and I would
guess that there is no bound on the rank. Shafarevitch and i [64] have
shown that the rank can be arbitrarily large if we take as ground field the
field of rational functions in one variable over a finite constant field
instead of the field of rational numbers.
Perhaps one reason that questions about E(Q) are so difficult is that
the adele methods which are so successful with linear algebraic groups
look like a mess in the case of elliptic curves. Since E is a projective
variety, the groups E(R) and E(Qp) are compact. Hence the map
w 8. L-Series
Let E be an elliptic curve over Q, and let (1) be a global minimal
Weierstrass equation for E. (In general, if K is the field of fractions of
a Dedekind ring R. there is for each E over K a certain ideal class in R,
2 A b o u t [12]. Cassels wrote m Math. Reviews (vol. 44, 1972, n ~ 2755). "" .. Unfortunately,
the e x p o s i t i o n is so obscure that the reviewer has ye! to meet s o m e o n e who would vouch
for the validity of the proof, on the o t h e r h a n d he has yet to be shown a m~stake that un-
a m b i g u o u s l y and irretrievably vitiates the a r g u m e n t . " (edit.)
14 Inxentloncs math., Vol.2~
196 J.T. Tate
Put
tp= I+p--Ap. (4l)
If p~/A, then tp is the "trace of Frobenius" and satisfies ]tpi<2l/p.
Ifp[A, then tp= 1, - 1 , or 0, according as/~(p) has a node with rational
tangents, a node with tangents quadratic over Fp, or a cusp.
One associates with E/Q an "L-function"
1 v1- 1
L~(s)= I1
plJ ( I - tpp -~) pl,~l
A 1 - t p p - ~ + p 1-z~"
(42)
E(R)
w 10. Examples
Here are some examples of curves over Q which illustrate various
points of the general theory we have discussed. WeJl's conjecture
predicts that there is no elliptic curve E over Q with conductor N < 11.
There are three known curves with N = 11 (cf. [78]), all isogenous as
they should be. Two of them are
y2q_ y = X 3 __ X2 (,~ = -- 11, j = --213/11)
and
yZ+y=x3-x2-1Ox-20 ( A = - l l S , j = -212313/115).
ql-[ll--q")2~(l--qlX")2= %q~,
n=l n=l n=!
then
and it follows from Igusa [19], [-20] that Eichler's exceptional set is
in fact empty (see also Deligne-Rapoport's paper in [84]].
Our next example is the curve
b/3 -1- U3 : W3 "
X = -3 .-' 9 [ u-v I
.+,,'
and it becomes
y2--3,=X3-- 7 (N=27, A= - - 3 9, j = 0 ) .
There are only three rational points (i.e., Fermat's last theorem is true
for exponent 3), and there is convincing numerical evidence that the
conjecture of Birch and Swinnerton-Dyer is true for it (cf. [71]). In the
Disquisitiones Gauss proved that the number of rational points on the
curve (modpl is
p+l, if p - - 1 (rood3),
:! /;
2-
7P
-2 /f / I N./~ ~ / 7 1 2 3 x
-p "~2P
-3 -t- ~, P
and is
p+l -iv, if p - 1 (mod3L
w h e r e tp is the u n i q u e i n t e g e r - - 1 ( r o o d 3 ) s u c h t h a t 4p=t~+27B 2
for s o m e i n t e g e r B. W e l l i n t e r p r e t e d Gauss" result as m e a n i n g t h a t
LE(s) was a c e r t a i n H e c k e L-series for the field Q ( ] / - 3 ) . T h e c u r v e
has c o m p l e x m u l t i p l i c a t i o n by the t h i r d r o o t s of unity, a n d the p o i n t s
o f finite o r d e r g e n e r a t e a b e l i a n e x t e n s i o n s o v e r t h a t field.
T h e g r o u p of r a t i o n a l p o i n t s o n the c u r v e
References
Most works cited below are post-1960, but no attempt at completeness has been
made even for this recent period. An excellent guide to the earlier literature is the btbliog-
raphy in Cassels' survey article [9]
I. Artin, M., Swinnerton-Dyer, H.P.F.: The Shafarevitch-Tate conjecture for pencils
of elliptic curves on K3 surfaces, lnventJones math. 20, 2~9-266 0973)
2. Atkln, A.O.L., Lehner, J.: Hecke operators on F0tin) Math. Ann. 185, 134- t60 (19701
3. Baker, A., Coates, J.: Integer points on curves of genus 1. Proc. Camb. Phil. Soc 67,
595-602 (1970)
4. Barsottl, I ' Analytical methods for abehan varieties in positive characteristic. Colloque
de Bruxelles, pp. 77-85, 1962
5. Birch, B.J., Swinnerton-Dyer, H.P.F.. Notes on elliptic curves (ill J remne u ange-
wandte Math. 218, 79-108 (1965)
6. Birch, B.J.: Elhptlc curves A progress report. Proceedings of the 1969 Summer institute
on Number Theory held at Stony Broek, New York, A.M.S pp. 396-400 (1971)
7. Cartier, P.: Groupes formels, fonctions automorphes et fonctions zeta des courbes
elliptiques Actes, Congres intern. Math. T. 2, 291-299 1197t0)
8. Cassels, J W S.: Arithmetic on curves of genus 1, IIV). Proof of the Hauptvermutung.
J. reme u. angewandte Math. 211, 95-112 (1962)
9. C assels, J W.S.: Diophantine equations with special reference to elliptic curves Survey
Article. Journ. London Math. Soc. 41, 193-291 (1966)
l0 Cassels, J W.S., Ellison, W J., Pfister, A ' On sums of squares and on eihptic curves
over functions fields. Journ. Number Theory 3, 125-149 (1971)
11. Coates, J.: An effective p-adlc analog of a theorem of Thue IIl. The Diophantine
Equation ),2 = x 3 + k. Acta Arithmetlca XV1,425-435 f1970)
12. Demjanenko, V.A.' On the torsion of elhptlc curves (in Russian). Isv. Acad Nauk
U.S.S.R. 35, 280-307 (19711 (English trans, in Math. of U.S.S R., isv., AMS)
The A m h m e t l c of Elhptic Curves 203
41. N e u m a n n , O. Zur Reduktion der elllptlschen Kurven. Math. Nach. 46, 285 310 ( 19701
42. Ogai, S V : On rational points on the curve y2=x(x2+ax+b) (Russian) Trudy,
Math. Inst. Steklov 80 (1965). Translated by A. M. S.
43. Ogg, A.P.: Cohomology of abelian varieties over function fields. Annals of Math.
pp. 185-212 (1962)
44 Ogg, A. P.: Abelian curves of small conductor. J. relne und angew. Math. 226, 204-
215 (1967)
45. Ogg, A. P.: Elliptic curves and wild ramification. Amer. J pp. 1-21 (1967)
46. Ogg, A.P.: Rational points of finite order on elliptic curves lsv. Math. 12, 105-111
(1971)
47. Ogg, A. P.: Rational pmnts on certain elhptic modular curves. [A talk given in St Lores
on March 29, 1972, at the A M S Symposium on Analytic N u m b e r Theory and Related
Parts of Analysis)
48. Rajwade, A.R.: Arithmetic on curves with complex multiplication by ] / - 2 Proc.
C a m b Phil. Soc. 64, 659-672 (1968)
49. Raynaud, M.: Caract6nstlque d ' E u l e r - P o i n c a r 6 d'un faisceau et cohomologle des
vari6t6s ab61iennes S6m. Bourbaki, 286, F6vner 1965
50. Razar, M.: The non-vanishing of L(II for certain elliptic curves. Harvard Ph. D Thesis,
June 1971
51. Roquette, P . Analytic theory of elliptic functions over local fields. Hamb. Math.
Einzelschriften Neue Folge, Heft 1, Gbttingen 1970
52 Serre, J-P.: Sur les groupes de congruences des vari6t~s ab611ennes lsv. Akad. Nauk,
Math Series 28, 3-20 (1964)
53. Serre, J-P.: Groupes de Lie l-adlques attach6s aux courbes elliptiques. Coll. lnternat.
du C N R.S., No. 143 a Clermont-Ferrand, Editions du C.N.R.S., Paris 1966
54 Serre, J-P.: Zeta- and L-functions, in Arithmetical Algebraic Geometry. New York:
Harper and Row 1966
55. Serre, J-P.: Abelian l-adic representations and elliptic curves. New York' BenJamin
1968
56 Serre, J-P., Tate, J.: Good reduction of abehan varieties. Annals of Math pp. 492-517
(1968)
57. Serre, J-P.: Facteurs locaux des fonctlons z6ta des vari6tes alg6briques (dSfimtlons et
conjectures), S6mlnalre Delange-Pisot-Poitou, 15 p., Mai 1970
58. Serre, J-P.: p-torsion des courbes elhptiques (d'apr6s Y Manan). Sere. Bourbaki, n ~ 380,
14 p., Mal-Juin 1970
59 Serre, J-P.: Propn6t6s galoisiennes des points d'ordre finl des courbes elliptlques.
inventiones math. 15, 259-331 (1972)
60 Serre, J-P.: Congruences et formes modulalres (d'apr+s H. P. F. Swinnerton-Dyer), S6m.
Bourbakl, No. 416, 20 p., Jura 1972
61. Setzer, C. B.: Elhptlc curves of prime conductor. Harvard Ph D. Thesis, J une 1972
62. Shafarevitch, I.R.. Algebraic number fields Proc. Int Cong. Math., Stockholm 1962,
pp. 163-176 (A M.S. Translation, Ser 2, Vol. 31, pp. 25-39)
63. Shafarevitch, t.R. Principal homogeneous spaces defined over a function-field. Amer.
Math. Soc. Transl. 37, 85-I 14 (1964)
64. Shafarevitch, I. R., Tare, J T.: The rank of elliptic curves. Dokl. Akad Nauk. USSR
175 (1967h pp. 770-773 (in Russian). (A.M S. Translations Vol. 8, 917-920 (19671)
65. Shlmura, G.: On the zeta functions ot the algebraic curves umformized by certain
automorphlc functmns Jour Math. Soc. Japan 13, 275-331 (1961)
66. Shimura, G.: A reciprocity law in non-solvable extensions. J. reine and angew. Math
221,209-220 (1966)
67. Shimura, G.: Introduction to the Arithmetic Theory of Automorphic Functmns.
Publ Math. Soc. Japan 11, lwanoml Shoten Publishers, and Princeton U m v Press,
267 p.(1971)
The Arithmetic of Elliptic Curves 205
68. Shimura, G.. On the zeta-function of an abehan variety with complex multiphcatlon
A n n Math 94, 504-533 (1971)
69. Shimura, G.: On elliptic curves with complex multiplication as factors of the jacoblans
of modular function fields. Nagoya Math. J. 43, 199-208 (1971J
70. Shioda, T. On rational points of the generic elliptic curve with level N structure over
the field of modular functions of level N (to appear)
71. Stephens, N. M ' The diophantine equation x J+ yJ= DZ 3 and the conjectures of Birch
and Swinnerton-Dyer J. Reine Angew Math. 231, 121-162 (1968)
72. Swinnerton-Dyer, H . P . F . The conjectures of Birch and Swinnerton-Dyer, and of
Tata. Proc. of a conference on local fields, pp 132-157. Berlin-Heidelberg-New York:
Springer 1967
73 Tate, J.:W.C. groups over P-adic fields S6mmalre Bourbakl, No. 156, 13 p., D~c. 1957
74 Tate, J." Algebraic cycles and poles of zeta functions. Arithmetical Algebraic Geometry
New York: Harper and Row 1966
75. Tate, J.: Endomorphisms of abehan varieties over finite fields. Inventiones math. 2,
134-144 (1966)
76 Tate, J.: On the conjecture of Birch and Swinnerton-Dyer and a geometric analog.
S+m. Bourbaki, No. 306, 26 p., F6v. 1966
77. Tate, J.: Classes d'isog6nie des vari6t6s ab61iennes sur un corps tim (d'apr6s T. Honda)
S6m Bourbakl, No. 352, 16 p., Nov. 1968
78. V61u, J.: Courbes elliptiques sur Q ayant bonne r6duction en dehors de (11 ). C. R Acad.
So. Paris, pp 73-75 (1971)
79. V61u, J.' Isog6nies entre courbes ellipttques. C. R. Acad. Sc Paris, pp 238-241 (1971)
80. Waterhouse, W.C., Milne, J.S.: Abelian varieties over fimte fields. A.M.S. S u m m e r
Institute on N u m b e r Theory, Stony Brook, 1969, Proc of Symposta m pure mathe-
matics, Vol. XX, AMS, 1971
81 Well, A.' Jacobi sums as ,,Grossencharaktere". Trans. Amer. Math. Soc. 75, 487-495
(1952)
82. Well, A.: Ober die Bestimmung Dirichletscher Relhen durch Funktionalglelchungen.
Math. Annalen 168, 149-156 (1967)
83. Well, A.: D~richlet Series and Automorphic Forms Lecture Notes in Mathematics 189.
Berhn-Heldelberg-New York. Springer 1971
Additional Bibliographv
84 Modular Functions of One Variable Vol. II. Lecture Notes in Mathematics 349.
Berlin-Heidelberg-New York: Springer 1973
85. Modular Functions of One Variable Vol. IV. Lecture Notes m Math. Berlin-
Heidelberg-New York: Springer 1974
86. Demjanenko, V.A. On the uniform boundedness of the torsion of elliptic curves over
algebraic number fields (in Russian). Izv. Akad. Nauk SSSR 36 (1972)
87. Kubert, D.: Universal bounds on the torsion and isogenies of elhptlc curves Harvard
P h . D . thesis, May 1973
88. McCabe, John. p-ad~c theta functions. Harvard P h . D . thesis, pt~. 1-222 (1968)
89. M a m n , Y.l : Cyclotomic fields and modular curves (in Russian). Usp Mat. Nauk 26,
7-71 (1971)
90. Manin, Y . l , Zarkin, Y G.: Heights on families of abehan varieties (in Russmn). Mat.
Sbornik 89, 171-181 (1972)
9l. Mazur, B.: Rational points of abehan varieties with values in towers of number fields.
Inventiones math. 18, 183-266 (1972)
92. Mazur, B, Swinnerton-Dyer, H.P.F.: Arithmetic of Weil curves. To appear m ln-
ventiones math.
93. Milne, J.: Weil-Chfitelet groups over local fields. A n n Sci. ENS, 3, 273-284 (1970);
Addendum, 1bid. 5, 261-264 (1972)
206 J.T. Tale