After the discussion in the previous elliptic curve question, we know that the elliptic curve $y^2+4xy=x^3-6x^2-24x+144$ (Given by @YongHaoNg), which can be transform into $C:y^2=x^3-20x^2+108x$, has rank $1$, and the only non-identity torsion point in $C$ is $(0,0)$, which is of order two. Now I'm working on finding the generators of the infinite subgroup in $C(\mathbb Q)$, the group of all rational solutions of $C$, which is isomorphic to $\mathbb Z$.

We know that $$\Gamma=C(\mathbb Q)\cong \mathbb Z_2\times \mathbb Z$$ So I define the subgroup $S=\{\mathcal O, (0,0)\}$ of $\Gamma$ to be the subgroup of all rational points of finite order. I also define $\Gamma_1$ to be the subgroup of $\Gamma$ isomorphic to the $\mathbb Z$ part.

My approach is: We first find out the index $\Gamma:\Gamma_1$, which I believe is $2$, and then since $(6,12)$ is in $\Gamma$, by group theory $(6,12)$ is either in $\Gamma_1$ or in $(0,0)+\Gamma_1$, in which case $(6,12)-(0,0)=(18,-36)$ is in $\Gamma_1$.

Now let $Q_0$ be the generator of $\Gamma_1$, in the book "Rational Points on Elliptic Curves" by Silverman and Tate, page 49, it has been shown that any point $(x,y)$ in $\Gamma$ is of the form $x=m/e^2,y=n/e^3$ for some integers $m,n,e$ with $\gcd(m,e)=\gcd(n,e)=1$. If $P,Q$ are two points in $\Gamma$ such that $x(P)=m_1/e_1^2,x(Q)=m_2/e_2^2$ with $e_1,e_2$ having nontrivial common divisor, then $x(P+Q)$ mustn't be an integer.

Combining the few facts above, we know that there is an integer $n$ such that $nQ=(6,12)$, if $(6,12)\in \Gamma_1$, or $nQ=(18,-36)$ if $(18,-36)$ is in $\Gamma_1$, therefore I could show that $Q_0$ is also an $\textbf{integer point}$. Finally, I will use my result, which says that $\Gamma$ is generated by all points in $\Gamma$ having height not larger than $568000$, then I just need to check $1136000$ cases by computer, $Q_0$ is ready to be found!

But I'm skeptical at one point, this problem may sounds silly, but I should keep asking until I can convince myself. Given $\Gamma\cong \mathbb Z_2\times \mathbb Z$ is true, is it really true that the index $\Gamma:\Gamma_1$ is equal to $2$? I'm having this problem because I knew that although $\mathbb Z$ is properly contains $2\mathbb Z$, but they two are actually isomorphic, so I'm wondering can $\Gamma:\Gamma_1$ be something larger than $2$?


Your index is $2$.

If we have an abelian group decomposition $G=T\oplus L$, then the first isomorphism theorem applied to projection on $T$ yields $G/L\simeq T$, and $[G:L]= \vert T\vert$ (provided that $T$ is finite).

Now take $G=\Gamma.$ Since $G$ is a finitely generated abelian group, we have a decomposition $\Gamma=T\oplus \Gamma_1$, where $T$ is the torsion subgroup of $G$ (the subgroup of points of finite order) and $\Gamma_1$ is a free abelian group of finite rank (the rank of the elliptic curve). In your example, $T$ has order two.

| cite | improve this answer | |
  • $\begingroup$ Thanks for your verification! $\endgroup$ – kelvin hong 方 Nov 21 '19 at 0:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.