Past exam of the mathematics course of the University of Cambridge / 2005 / iii / Paper 32 / 4

Translate the rational point of order two to $(0,0)$ and complete the square. After clearing denominators by admissible scaling, writeThe companion curve is $E^{\prime }:Y^2=X(X^2-2aX+a^2-4b)$. The two-isogeny formula gives maps $\varphi :E\to E^{\prime }$ and $\psi :E^{\prime }\to E$ satisfying $\psi \varphi =[2]$.

For the Weak Mordell-Weil theorem in this setting, use the two-torsion square-class homomorphismand its counterpart ${\alpha }^{\prime }$ on $E^{\prime }$. These are homomorphisms: a nonvertical line intersecting the cubic in three points has the product of their $x$-coordinates a square, by comparison of the constant term of the intersection cubic. For a line through $(0,0)$, the product of the other two coordinates is $b$, explaining the special value there. The kernels are $\psi E^{\prime }(\mathbb{Q})$ and $\varphi E(\mathbb{Q})$, respectively, as follows by solving the elliptic isogeny preimage quadratic and testing its discriminant for a square.

The prime-support bound in two-isogeny descent makes both images finite. Indeed at a prime not dividing $b$, a negative valuation of $x$ makes the leading cubic term dominate, giving $2v(y)=3v(x)$; a positive valuation makes $x^2+ax+b$ a unit, giving $2v(y)=v(x)$. In either case $v(x)$ is even. Thus the square class of $x$ is supported only on primes dividing $b$, together with a possible sign. The same argument uses $b^{\prime }=a^2-4b$ on $E^{\prime }$. Since $2E=\psi \varphi E$, the two finite elliptic isogeny quotients imply that $E(\mathbb{Q})/2E(\mathbb{Q})$ is finite.

Finite generation still needs a descent on height. Let $h_x$ be the logarithmic height of the $x$-coordinate. The duplication formula has a degree-four rational $x$-coordinate map with coprime numerator and denominator. Polynomial height estimates, with the resultant supplying the lower bound, give the height growth under a morphism of the projective line relation $h_x(2P)=4h_x(P)+O(1)$. Thus the convergent telescoping limitdefines the canonical height of an elliptic curve, without assuming finite generation. It is nonnegative, satisfies $\hat{h}(2P)=4\hat{h}(P)$ and the height parallelogram identity, and differs from $h_x/2$ by a bounded amount. Consequently its bounded subsets of rational points are finite by the Northcott theorem.

Choose the finite list $R_i$ of representatives modulo $2E(\mathbb{Q})$ and let $H={\max }_i\hat{h}(R_i)$. For $P=2Q+R_i$, the canonical height pairing inequality givesRepeated halving brings the height below $H+1$. There are only finitely many such terminal points, and recovering $P$ expresses it as an integer combination of those points and the $R_i$. This proves the canonical-height proof of Mordell-Weil finite generation, so

Retain $E:y^2=x(x^2+ax+b)$ and put $b^{\prime }=a^2-4b$. The explicit two-isogeny formula iswith the maps extended at the kernel points. Direct substitution verifies the equations and $\psi \varphi =[2]$.

Here is a concrete kernel test for the descent homomorphisms. A preimage of $(X,Y)\in E^{\prime }$ has $x$ satisfying $x^2+(a-X)x+b=0$, whose discriminant is $(X-a{)}^2-4b$. The target equation says $Y^2=X((X-a{)}^2-4b)$, so for nonexceptional points this quadratic has a rational solution precisely when $X$ is a square. The resulting $y$ follows from the displayed elliptic isogeny formula. For a nonzero target two-torsion point, the preimage quadratic has a double root $x=(X-a)/2$, and the source equation gives $y^2=x^{2}X$, so the same square condition still applies. At $(0,0)\in E^{\prime }$, a rational preimage exists precisely when $b^{\prime }$ is a square, matching the special square-class value ${\alpha }^{\prime }((0,0))=[b^{\prime }]$. The dual argument gives the other kernel. ThusLet $\delta =[\ker \psi :\ker \psi \cap \varphi E(\mathbb{Q})]$. The quotient map induced by $\psi$ has this kernel, givingIf $b^{\prime }$ is a rational square, $E$ has four rational two-torsion points and $\delta =1$; otherwise it has two and $\delta =2$. In either case $\delta \#E(\mathbb{Q})[2]=4$. Combining with finite generation therefore proves the two-isogeny rank formulaThis derivation checks the factor four, including curves with full rational two-torsion.

To compute these images, enumerate the signed square-free divisors $d$ of $b$ and $b^{\prime }$. Substituting $x=d(U/V{)}^2$ yields the quartic covering in a two-isogeny descentThe class $d$ occurs precisely when this covering has a rational point, with the identity and two-torsion handled by the boundary cases. Test at the real place and at primes dividing $2b{b}^{\prime }$ to rule out candidates, and search for rational points to prove that surviving classes occur. The locally soluble classes form a finite elliptic isogeny Selmer group of an elliptic curve, giving an upper rank bound. rational points and independent height pairings give a lower bound. When the bounds agree the rank is proved. Local solubility alone does not prove a rational point exists: if a gap remains, a higher descent or further information about the covering obstructions is needed. This is why the procedure usually computes the rank but is not an unjustified universal termination claim.

For a rank-zero example, take $E_0:y^2=x^3+x$, with rational two-torsion $(0,0)$. Its two-isogeny formula companion is $E_0^{\prime }:y^2=x^3-4x$. The prime-support bound in two-isogeny descent allows only $\pm 1$ on $E_0$; all nonzero affine $x$ are positive because $x(x^2+1)$ is a square. Hence $A_0=\{1\}$. For $E_0^{\prime }$ the possible classes are $\{1,-1,2,-2\}$, all realized: $O$ gives $1$, $(0,0)$ gives $[-4]=-1$, and $(2,0),(-2,0)$ give $2,-2$. ThereforeThis is the rank-zero elliptic curve with cubic x cubed plus x certificate; no unresolved locally soluble covering is being counted as soluble.

For a positive-rank example, take $E_1:y^2=x^3-2x$. Its companion is $E_1^{\prime }:y^2=x^3+8x$. The first image is contained in $\{1,-1,2,-2\}$. The identity gives $1$, $(0,0)$ gives $-2$, and $(2,2)$ gives $2$; multiplicativity of the square-class group homomorphism also gives $-1$. Thus $A_1$ has all four classes. On $E_1^{\prime }$, the same prime-support bound allows $\pm 1,\pm 2$, but $x(x^2+8)$ is nonnegative only for $x\ge 0$, eliminating the negative classes. The identity gives $1$ and $(0,0)$ gives $[8]=2$, so $B_1=\{1,2\}$. ConsequentlyThe rank-one elliptic curve with cubic x cubed minus twice x therefore supplies the required positive-rank example. The exhibited point $(2,2)$ is itself nontorsion: doubling gives $(9/4,-21/8)$, whose nonintegral coordinates contradict the Nagell–Lutz theorem if it were torsion. No assertion that this point generates the full free part is needed for the rank calculation.