Abstract.
We call an order in a quadratic field odd (resp. even) if its discriminant is an odd (resp.even) integer.We call an elliptic curve over with CM odd (resp. even) if its endomorphism ring is an odd (resp. even)order in the imaginary quadratic field .
Suppose that and let us consider the set of all where is any elliptic curve that enjoys the following properties.
- β’
isisogenous to ;
- β’
;
- β’
has the same parity as .
We prove that the closure of in is the closed semi-infinite interval (resp. the whole ) if is odd (resp. even).
This paper was inspired by a questionof Jean-Louis Colliot-Thélène and Alena Pirutka about the distribution of -invariants of certain elliptic curves of CM type.
1. Orders in quadratic fields
As usual stand for the ring of integers and the fields of rational, real and complex numbers respectively.If is a positive integer then we write for the subring of generated by and . More generally,If is a field of characteristic zero and is a subring of (with ) then we write for the subringof generated by and ; clearly, is a -subalgebra of . We write for the subring of that consists of all fractions where and an odd integer.
Let be a quadratic field, its ring of integers, and
| | |
be the corresponding trace map attached to the field extension .We have
| | |
Let be an order in and let a positive integer be the conductor of . This means that
| | |
and coincides with the index of the additive subgroup in [2, Ch. 2, Sect. 7, Th.1].Since and ,
| | | (1) |
Recall that the discriminant of is the discriminant of the symmetric bilinear form
| | |
By definition, the discriminant of the field is . It is well known [2, Ch. 2, Sect. 7, Th.1] that
| | | (2) |
Definition 1.1 (Definition-Lemma).
An order is called odd if it enjoys the following equivalent properties.
- (i)
.
- (ii)
is an odd integer.
- (iii)
Both and the conductor are odd integers.
- (iv)
There exists an integer that is not a square such that
| | |
Definition 1.2 (Definition-Lemma).
An order in a quadratic field is called even if it enjoys the following equivalent properties.
- (i)
.
- (ii)
is an even integer.
- (iii)
Either or the conductor is an even integer.
Clearly, every order is either odd or even. We prove Lemmas 1.1 and 1.2 in Section 5.
Definition 1.3.
Let be an elliptic curve with complex multiplication (CM) over the field of complex numbers.Then its endomorphism ring is an order in the imaginary quadratic field
| | |
We say that is odd (resp. even) if the order is odd (resp. even).
Clearly, every elliptic curve over with CM is either odd or even. The following assertions will be proven in Section 6.
Proposition 1.4.
Let be an isogeny of complex elliptic curves with CM.Suppose that is an odd integer.Then is odd (resp. even) if and only if is odd (resp. even).In other words, and have the same parity.
Proposition 1.5.
Suppose that and are isogenous CM elliptic curves over . Suppose that there are elliptic curves and over the field of real numbers such that there are isomorphisms of complex elliptic curves
| | |
Then and are isogenous over .
Remark 1.6.
Let be an elliptic curve over . It is well known [5, Ch. 3, Sect. 2, Prop. 3.7] that if and only if may be defined over , i.e.,there is an elliptic curve over such that . (See also [11, Appendix A, Prop. 1.2(b)] and [9, Ch. 4, Sect. 4.1]).)
Examples 1.7.
- (e)
If is an elliptic curve then it is well known that the order has even discriminant and therefore is even.
- (o)
If is an elliptic curve then it is well known that the order has odd discriminant and therefore isodd.
2. -invariants of CM elliptic curves
In what follows, is an elliptic curve over .The following assertion will be proven in Section 3.
Proposition 2.1.
Let us considerthe subset of all where runs through the set of ellipticcurves over such that there exists an isogeny of odd degree.
Then is dense in .
Remark 2.2.
If has CM thenall that appear in Proposition 2.1 are elliptic curves with CMand the corresponding imaginary quadratic fields are isomorphic to .
Notice that in light of Examples 1.7, there exist an odd CM curve and an even CM curve. Now,combining Proposition 2.1 with Proposition 1.4, we obtain the following statement.
Corollary 2.3.
Let be the set of all where runs through the setof all odd elliptic curves over with CM.
Let be the set of all where runs through the setof all even elliptic curves over with CM.
Then each of two subsets and is dense in .
Our main result is the following assertion.
Theorem 2.4.
Suppose that is a CM curve with . Let us consider the set of all where is any elliptic curve over that enjoys the following properties.
- β’
is isogenousto ;
- β’
;
- β’
has the same parity as .
Then the closure of in is the closed semi-infinite interval (resp. the whole ) if is odd (resp. even).
We prove Theorem 2.4 in Section 4. Our proof is based on auxiliary results about isogenies of complex elliptic curves with CM that will bediscussed in Section 3.
Remark 2.5.
It follows from Remark 1.6 and Proposition 1.5 that Theorem 2.4.is equivalent to the following statement.
Theorem 2.6.
Let be an elliptic curve over such that its complexification is of CM type.Let us consider the set of all where is any elliptic curve over thatenjoys the following properties.
- β’
is -isogenous to ;
- β’
The complexification of has the same parity as .
Then the closure of in is the closed semi-infinite interval (resp. the whole ) if is odd (resp. even).
The paper is organized as follows. In Section 3 we discuss auxiliary resultsabout isogenies of complex elliptic curves. We prove our main results in Section 4,using properties of odd and even orders and parity preservation under isogenies of odd degrees that will be provenin Section 5 and 5 respectively. In Section 7 we use the techniques developed in Section3, in order to describe explicitly -torsion in the class group of proper -ideals where is an odd corderin an imaginary quadratic field. In Section 8 we βclassifyβ CM elliptic curve, whose -invariant is real andthe endomorphism ring is a given odd order.
3. The upper half-plane and elliptic curves
Definition 3.1.
If and are nonzero complex numbers such that then we write for the discrete lattice of rank in .
Let
| | |
be the upper half-plane. If then we write for the complex elliptic curvesuch that the complex tori and are isomorphic. Here
| | |
See [12, Ch. V, p. 408β411] for a Weierstrass equation of . (In the notation of [12, Ch. V], our is with .)
Remark 3.2.
If then the group of hom*omorphisms from can be be canonically identified with
| | | (3) |
Each that satisfies (3) corresponds to the hom*omorphism
| | |
of elliptic curves such that the corresponding action of on the complex points is as follows.
| | |
If then is an isogeny,whose degree coincides with the index of the subgroup in [6, p. 9]. In particular, if
| | |
(i.e., the index is ), is an isomorphism of elliptic curves and .
On the other hand, if w take and put
| | | (4) |
then the map
| | | (5) |
is a ring isomorphism.
Remark 3.3.
We have
| | |
i.e.,
| | | (6) |
It follows easily that if is a subring of then .
Remark 3.4.
- (i)
It is well known [9, Ch. 4, Sect. 4.4, Prop. 4.5] that has CM if and only if is an (imaginary) quadratic irrationality,i.e., is an imaginary quadratic field. If this is the case then is an order in the imaginary quadratic field ; in particular,
| | | (7) |
In addition, extending isomorphism (5) by -linearity,we get a canonical isomorphism of -algebras
| | |
between and the endomorphism algebra of .
- (ii)
Suppose that has CM, i.e., is an imaginary quadratic field. In light of (6),
| | | (8) |
It follows from Remark 3.3 that if is an order in then
| | |
- (ii)
Let and be two elliptic curves with complex multiplication. Then and are isogenous if and only if the corresponding imaginary quadratic fields coincide, i.e.,
| | |
(see [9, Ch. 4, Sect. 4.4, Prop. 4.9]).
The group of two-by-two real matrices with positive determinant acts transitively on by fractional-linear transformations. Namely, the continuous map
| | |
defines the transitive action of on .(Actually, even the action of the subgroup on is transitive.) We will mostly deal with the action of the subgroup
| | |
Lemma 3.5.
Suppose that a matrix
| | |
has positive determinant
| | |
i.e, .Then there exists an isogeny of odd degree.
Proof.
By definition of , there are integers ,and a positive odd integer such that
| | |
The conditions on imply that is a positive odd integer.Then
| | |
where the matrix
| | |
(Since
| | |
we get
| | |
We have
| | |
This means that
| | |
is a subgroup in of odd index .This implies that if we put
| | |
then is a subgroup of odd index in . This gives us the isogeny of odd degree . This ends the proof.β
Proof of Proposition 2.1.
Let . The weak approximation for the field [1, Th. 1] with respect to places implies the existence of a sequence of matrices with rational entries such that converges to in the real topology and to the identity matrixin the -adic topology. Removing first finite terms of the sequence, we may and will assume that for all
| | |
the latter inequality means that all and therefore
| | |
It follows that is the limit of the sequence
| | |
in the complex topology. Now the transitivity of implies thatevery orbit is dense in in the complex topology.
Recall that the classical holomorphic (hence, continuous) modular function
| | |
takes on every complex value. It follows that for every the set is dense in .
Let be an elliptic curve with complex multiplication. Then there exists such thatthe elliptic curves and are isomorphic and therefore
| | |
In light of Lemma 3.5, if then there is anisogeny of odd degree. Since
| | |
the set of complex numbers contains, which is dense in . Hence, is also dense in ,which ends the proof.β
Example 3.6.
Let . Let be positive odd integers.Then the matrix
| | |
satisfies the conditions of Lemma 3.5. We have
| | |
It follows from Lemma 3.5 that there exists an isogenyof odd degree.This implies that if is an imaginary quadratic irrationality then, by Proposition 1.4 (to be proven in Section 6),the CM elliptic curves and have the same parity.
Example 3.7.
Let where . Let be positive odd integers. Let us put
| | |
Notice that is even and therefore .Then the matrix
| | |
satisfies the conditions of Lemma 3.5. We have
| | |
It follows from Lemma 3.5 that there exists an isogenyof odd degree.By Proposition 1.4 (to be proven in Section 6), if is an imaginary quadratic irrationality then the CM elliptic curves and have the same parity.
In what follows, if is a negative real number then we write for .
Example 3.8.
If is a point in such that is an imaginary quadratic field and the lattice is an order in then . E.g., let be a negative integer that is congruent to modulo and
| | |
Then is an order in the imaginary quadratic field .By Remark 3.4(ii),
| | |
It follows from Lemma 1.1 thatis an odd order in the quadratic field . This means that the elliptic curve is odd.
Now let be any positive odd integers. Let us consider the complex numbers
| | | (9) |
It follows from Example 3.7 (applied to ) that all the elliptic curves are also odd.
Lemma 3.9 (Key Lemma).
Let such that is an imaginaryquadratic field, i.e., is an elliptic curve with CM.
Proof.
(i). Assume that . Suppose that is odd, i.e., the order is odd. It followsfrom Lemma 1.1that there is a negative integer
| | |
such that
| | |
Since and , there is a positive rational number such that
| | |
This implies that
| | |
It follows from the definition of (see (4)) that
| | |
This implies that there are integers such that
| | |
Taking the real parts of both sides, we get , which contradicts the integrality of .The obtained contradiction proves that is even, which proves (i).
(ii). Assume that . Recall that we are given thenegative integer such that
| | |
Then there is a positive rational number such that
| | |
It follows that
| | |
By definition of (see (4)), if and only if
| | | (11) |
However, since is a basis of the -vector space , there are certain rational numbers such that
| | | (12) |
The inclusion (11) is equivalent to the condition
| | | (13) |
Opening the brackets and collecting terms in (12), we get
| | |
Taking the real parts and βimaginaryβ parts, we get
| | | (14) |
The second equality of (14) implies that
| | | (15) |
Since is positive, is also positive.
It follows from first equality of (14) that is an integer if and only if is an odd integer.This implies that if is not an odd integer then (13) does not hold. So, in the course of the proof we may and will assume that
| | | (16) |
It follows from the last equality of (14) combined with (15) that
| | |
which means that
| | |
In light of (16),
| | | (17) |
In addition, is even (resp. odd) if (resp. if ). This means that
| | | (18) |
Now let us explore the integrality of .Combining (15) with third equality of (14), we get
| | | (19) |
Since both and are odd integers, their sum is an even integer.It follows that if the odd integer does not divide the odd integer then and therefore (13) does not hold. So, in the course of the proof we may and will assume that
| | | (20) |
It follows that the even integer is divisible by .
Taking into account that ,we conclude that if then is divisible by and therefore
| | |
Combining with (18), we get
| | |
This implies that
| | |
On the other hand, suppose that. Then the integer
| | |
and therefore the integer ,which implies that
| | |
In light of (18), and therefore
| | |
This implies that all are integers if and only if where is a positive odd integer dividing . This ends the proof of (ii), in light of (15).β
Example 3.10.
Fix a negative integer that is congruent to modulo (e.g., we may take ). If and are positive odd integers then is also a negative integer that iscongruent to modulo . If we put
| | |
then is a positive odd integer dividing . It followsthat if we put
| | | (21) |
then the elliptic curve is odd.
Recall that and could be any odd positive integers. Notice also that all CM elliptic curves are isogenous to each other, because the imaginary quadratic field
| | |
does not depend on .
Corollary 3.11.
Let be a square-free negative integer such that.If and are positive odd integers and
| | |
then is even.
Proof.
In light of Example 3.7 applied to , it suffices to check the case , i.e., we may (and will) assume that
| | |
Clearly,. Suppose that is odd, i.e., the order is odd.Let be the conductor of . By Lemma 1.1, is odd, hence,
| | |
It follows from [2, Ch. 2, Sect. 7, Th.1]that
| | |
So,
| | |
Notice that
| | |
where is a negative integer that is also congruent to modulo , because is odd.By Lemma 3.9, there is an odd positive integer such that
| | |
Since , we get, i.e., . Since is odd, we get the desired contradiction.β
In light of Theorem 7.4(ii) (see below), it is natural to βclassifyβ (count the number of) divisors of the discriminant that are relatively prime to . Let us start with the following definition (notation).
Definition 7.1 (Definition-Notation).
Let be a nonzero integer. We write for the (finite) set of prime divisors of . By Main Theorem of Arithmetic,
| | |
where are certain positive integers uniquely determined by .
Let us call a divisor of a saturated divisor if the integers and are relatively prime.
Clearly, is a saturated divisor of if and only if is one. It is also clear that is a saturated divisor of if and only if is also one.
The following assertion is an easy exercise in elementary number theory.
Proposition 7.2.
- (i)
Let be a subset of . Then both integers
| | |
and are saturated divisors of .(As usual, if is the empty set then .)Conversely, if is a saturated divisor of then is also a saturated divisor of and
| | |
- (ii)
The number of positive saturated divisors of is the number of subsets of , i.e., where is the cardinality of , i.e., the number of prime divisors of .
- (iii)
Let and be positive saturated divisors of . Then their greatest common divisor and least common multiple are also saturated divisors of . In addition, the integers and are also positive saturated divisors of .We also have
| | |
where
| | |
and
| | |
is the symmetric difference of and .
- (iv)
If and are disjoint subsets of then and are relatively prime and
| | |
- (v)
Suppose that is a negative integer that is congruent to modulo . Then the numberof positive saturated divisors of with equals .
Proof.
(i) Let . Then
| | |
If is a prime divisor of both and then it must belong to both sets and . However, and do not meet each other. Hence, and have no common prime factors and therefore are relatively prime. This implies that the divisor (and therefore ) of is saturated.
Conversely, suppose that is a saturated divisor of . Then and
| | |
I claim that
| | |
Indeed, suppose that for some . Then divides both and ,which contradicts the saturatedness of . This ends the proof.
(ii) It follows from (i) that the set of positive saturated divisors of coincides with
| | |
It is also clear that if and are distinct subsets of then ,because their sets of prime divisors (i.e., and ) do not coincide.So, the number of positive saturated divisors of equals the number of subsets of , which is .
(iii) Since
| | |
the GCD of and equals and the LCM of and equals .Now the saturatedness of both and follows from the already proven (i). In order to handle and , notice that
| | |
Again, it follows from (i) that both and are saturated divisors of .Since is the disjoint union of and ,
| | |
which is also saturated, in light of (i).
(iv) follows readily from (iii).
(v) Since , the positive odd integer is congruent to modulo and therefore is not a square. Hence, if is a positive saturated divisor of then is also a positive saturated divisor of while . Clearly,in the pair there is precisely one element that is strictly less than .In light of (ii), the number of positive saturated divisors of that satisfies this inequality is.β
Let be an odd order with discriminant in a quadratic field . It follows from Lemma 5.1 combined with Remark 5.4that the odd integer is not a square, and
| | |
If are elements of then we write for the additive subgroup of generated by the set . As usual, if and are additive subgroups of then we write for the additive subgroup of generated by all the products ().For example, if then
| | |
We write for the only nontrivial automorphism of the field defined by the formula
| | |
Clearly,
| | | (26) |
It is known [2, Ch. II, Sect. 7] that
| | |
In what follows, is always an odd integer. Let us put
| | | (27) |
We have
| | | (28) |
This implies that
| | | (29) |
We also have
| | | (30) |
in light of
| | |
combined with first equality of (29).
Let us consider the -lattice
| | | (31) |
in .It follows from (30) that is a subgroup of and the corresponding index
| | | (32) |
Taking into account that is a basis of the -vector space , we get
| | | (33) |
Examples 7.3.
- (i)
If then
| | |
- (ii)
If then, in light of (30),
| | |
- (iii)
| | |
because
| | |
So,
| | | (34) |
- (iv)
Since
| | |
we get
| | |
So,
| | | (35) |
Actually, we have checked that
| | |
This implies that does not depend on the choice of a square root of .
- (v)
We have
| | |
because
| | |
So,
| | | (36) |
- (iv)
Suppose that divides . Then
| | |
is an odd integer and
| | | (37) |
Indeed,
| | |
| | |
which proves (37).It follows that
| | |
| | |
This implies that and therefore
| | | (38) |
In the course of the proof we will needthe following multiplication table that will be proven later in this section.
Lemma 7.5 (Multiplication Table).
Let , and (of course) be odd integers. Then the following equalities and inclusions hold.
- (i)
- (ii)
- (iii)
In particular, if and only if .
- (iv)
| | | (42) |
| | |
- (v)
| | | (43) |
Proof of Theorem 7.4 (modulo Lemma 7.5.
(i) We need to find out for which
| | |
The first inclusion holds without any additional conditions on . Indeed,
| | |
| | |
Concerning the second inclusion, recall ((28) and Lemma 7.5(i,ii)) that
| | |
This implies that
| | |
| | |
Since obviously lies in , the product liesin if and only if the integer lies in . In light of (33), this means that is divisible by . Since and odd are relatively prime, the divisibility condition means that , i.e., divides . This ends the proof of (i).
(ii) Since divides , the ratio is an integer and
| | |
Case 1 Suppose that is relatively prime to . Then and are also relatively prime.
By Lemma 7.5(i) and Example 7.3(iii),
| | |
This implies that
| | |
| | |
| | |
Since and are relatively prime integers,
| | |
This implies that
| | |
| | |
(here we use (29)). So, we have proven that
| | | (44) |
According to the already proven (i), is an ideal in the ring and therefore is an -submodule in the field . The fractional -ideal
| | |
is also an -submodule in . It follows from (44) that
| | |
Hence, is an invertible -submodule of the field . By [3, Ch. 2, Sect. 5, n 6, Th. 4], is a projective -module of rank .
Now let us check that is a proper -ideal. Indeed,let be an element of such that . We need to check that . Indeed, we have
| | |
So, . Since , we get , i.e., .(See also [3, Ch. 2, Sect. 5. Ex. 9b].)
Case 2 Suppose that the integer is not relatively prime to . We are going to prove that the -ideal is not proper.
Let be the greatest common divisor of and . Let us consider the integers
| | |
Then is obviously an odd integer that divides . Let us stress that it follows from the very definition of that and are relatively prime.In addition, is obviously a square root of . On the other hand, is obviously odd, hence its square is congruent to modulo . Since , we conclude that is also congruent to modulo . Now let us consider the order
| | |
of discriminant in ,its element
| | |
and the subgroup
| | |
We have
| | |
It follows that
| | |
because the orders and have distinct discriminants. On the other hand,
| | |
and therefore
| | | (45) |
Now we are ready to apply to , , and (instead of , , and )all the already proven results of this section (including Case 1 of the proof of (ii)), taking into account that the odd integers and are relatively prime.We get that:
- (a)
is an ideal in the ring ;
- (b)
;
- (c)
is a projective -module of rank 1that is a proper ideal of - the latter means that
| | | (46) |
On the other hand, in light of (45), .Combining it with (b) and (46), we conclude that is an ideal in , and
| | | (47) |
| | |
Since , the ideal of is not proper.(Notice that (47) implies that is a proper ideal in.)
This ends the proof of (ii).
(iii) In light of Proposition 7.2(iii), is a saturated divisor of . Let us put
| | |
Then and are relatively prime and
| | |
We have
| | |
in particular,
| | |
In addition, in light of Lemma 7.5,
| | |
Since
| | |
contains both
| | |
Hence, contains both
| | |
and
| | |
Since the integers and are relatively prime,
| | |
Taking into account that with , we obtain that contains
| | |
This implies that
| | |
On the other hand, recall (Example 7.3(v)) that.In light of Example 7.3(v), in order to prove the desired equality, it suffices to check that. Let us do it.
Recall that .By Lemma 7.5,
| | |
Case 1 Suppose that . Then .This means that . It follows that
| | |
This implies that the integer is congruent to modulo , i.e., is odd.On the other hand, , which implies that is an odd integer. It follows that there are integers and such that
| | |
and therefore
| | |
| | |
Since, , we conclude that. So, we are done.
Case 2 Suppose that . Then . This means that . It follows that is divisible by (recall that both and are odd) and therefore the integer is even. It follows that there are integers and such that
| | |
Hence,
| | |
| | |
because . This ends the proof.
(v) Applying (iv) to
| | |
we get
| | |
ByExample 7.3, . This implies that
| | |
Multiplying both sides by , we get (in light of the already proven (40))
| | |
which means that
| | |
i.e.,
| | |
β
Proof.
(i) Recall (33) that If where are integersthen none of lies in while their product lies in . So, is not a prime idealif is not a prime number.
Suppose that is a prime number. In light of (32), the quotient ring consists of elements.Since is prime, is the prime finite field . Hence, is a maximal idealin ; hence it is a prime ideal in .
(ii) In light of Proposition 7.2 applied to , the positive integer is the GCD of and . In addition,if we put then coincides with the disjoint union of and ,every pair of integers fromis mutually prime and
| | |
In light of Theorem 7.4(ii, iii),
| | |
| | |
| | |
| | |
This ends the proof.β
Proof of Lemma 7.5.
(i) We have
| | |
| | |
Since is odd, . Since , the difference is divisible by , i.e., is an integer.
(ii)
| | |
By Vietaβs formulas,
| | |
(iii) Recall that . In light of already proven (i),
| | |
| | |
This proves the first assertion, of (ii). On the other hand, since , the product lies in if and only if .In light of (33), this is equivalent to the divisibility of , which ends the proof.
(iv)
| | |
| | |
So,
| | |
Since both and lie in , the same is true for . This proves the first formula from (iv).Interchanging and , we get the second formula from (iv).
(v)
| | |
| | |
| | |
β
Now and till the end of this section we assume that , i.e., is an imaginary quadratic field.We will view as the subfield of such that
| | |
Then the nontrivial automorphism coincides with the restriction of the complex conjugation to . This implies(in light of (26)) that
| | | (48) |
Recall [2, Ch. 2] that a full module in is a subgroup of that is a free abelian (sub)group of rank ;the multiplier ring of is defined as
| | |
It is well known that is an order in .
On the other hand, if is an order in then it is said that a full module belongs to if. Such a module is also often called a (fractional) proper -ideal.For example, is a proper -ideal for all ; such an ideal is called principal.Obviously, if is a proper -ideal then is also one for all . The proper -ideals and are called similar. Recall that . It follows that is also a proper -ideal. There is a nonzero rational number that is called the norm of such that
| | | (49) |
[2, Ch. 2, Sect. 7.4]. It follows that
| | |
which implies that is an invertible -submodule of the field . By [3, Ch. 2, Sect. 5, n 6, Th. 4], is a projective -module of rank .
If are two proper -ideals then is also a proper -ideal and the binary operation
| | |
defines the structure of a commutative group on the set of all proper -ideals where corresponds to the zeroof the group law while the inverse of is [2, Ch. 2, Sect. 7.4, Th. 2].(Notice thatthe hom*omorphism of -modules
| | |
is a surjective hom*omorphism of projective -modules of rank and therefore is an isomorphism.This implies that the -modules and are canonically isomorphic.)
The quotient of this group modulo the subgroup of principal ideals is called the class group of and denoted by ;its elements are similarity classes of proper -ideals. It is known [2, Ch. 2, Sect. 7.4, Th. 3] that is a finite commutative group. We write for the kernel of multiplication by in . In light of (49),elements of are precisely similarity classes of those proper -ideals such that is similar to .
When is a maximal order in , the structure of the elementary -group is well known ([8, Sect. V.1, Th. 39], [13, Teil II, $ 12]. The next assertion describes explicitly thestructure of when is any odd order (in an imaginary quadratic field).
Theorem 7.7.
Let be an imaginary quadratic field, an odd order in with discriminant .Let be the set of prime divisors of and the cardinality of .Let be the Booleanalgebra of subsets of where the sum of two (sub)sets is their symmetric difference. Let us consider the map
| | |
that assigns to each the similarity class of the ideal .
Then is a group hom*omorphism, whose image coincides with while the kernel consists of two elements, namely, and .
The following assertion is a natural generalization of a special case of([8, Sect. V.1, Cor. 1 to Th. 39], [13, p. 112, Korollar]).
Corollary 7.8.
Let be an odd order with discriminant in an imaginary quadratic field .Then the order of is where is the number of prime divisors of .
Proof of Corollary 7.8 (modulo Theorem 7.7).
One may view as the surjective group hom*omorphism ,whose kernel has order . Recall that the order of is .Then the order of is the order of divided by , i.e.,. This ends the proof.β
Proof of Theorem 7.7.
It follows from Corollary 7.6(ii) that is a group hom*omorphism. Since the order of every element of is either or , the image lies in .
If then
| | |
Now it follows from Examples 7.3(ii,iii)that is a principal ideal and therefore contains .Suppose that
| | |
We need to arrive to a contradiction.First,
| | |
Then
| | | (50) |
This implies that both and are divisors of odd ; hence, they are odd integers as well.In addition, it follows from (50) that either or .Replacing if necessary by , we may and will assume that
| | |
In addition, is an odd divisor of .
Second, the condition means that the ideal is principal, i.e., there is a nonzero such that. The latter means that, which implies that
| | |
where
| | |
This implies that (in the notation of Section 3)
| | | (51) |
Notice that
| | |
| | |
the latter inequality holds, because . On the other hand, since ,
| | |
This implies that and are distinct points of . In particular,
| | |
On the other hand, in light ofRemark 3.2, the equality (51) implies that the complex elliptic curves and are isomorphic, i.e.,
| | |
This implies that , which gives us the desired contradiction that proves our assertion about .
Now let us check the surjectiveness of
| | |
Let be a proper -ideal,whose similarity class has order 1 or 2 in , i.e., is a principal ideal for some nonzero . Notice that where is a basis of the -vector space . Then is similar to
| | |
where. Since , it follows from (48) that .Replacing if necessary by , we may and will assume that
| | |
So, is a proper -ideal, whose square is a principal -ideal.This means that (see [9, Ch. 5, Sect. 5.4, Claim (5.4.4)].Hence, there is
| | |
It follows from the basic properties of the modular function that there exists a unimodular matrix
| | |
such that
| | |
This implies that
| | |
It follows that is a proper -ideal that is similar to . In light of Key Lemma 3.9, there is a positive odd integer dividing such that
| | |
This implies that
| | |
This implies that is a proper -ideal that is similar to and therefore to .It follows from Theorem 7.4 that is a saturated divisor of . This implies that if is the set of prime divisors of then. This means that is the similarity class of , which coincides with the similarity class of .It follows that the similarity class of lies in the image of . This ends the proof.β