Pete L. Clark
Abstract: We show that there exist genus one curves of every index over the rational numbers, answering affirmatively a question of Lang and Tate. The proof is "elementary" in the sense that it does not assume the finiteness of any Shafarevich-Tate group. On the other hand, using Kolyvagin's construction of a rational elliptic curve whose Mordell-Weil and Shafarevich-Tate groups are both trivial, we show that there are infinitely many curves of every index over every number field.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: **Does there exist genus one curves (i.e., curves of genus one) of arbitrary index over every number field?**
Specifically, Lang and Tate proposed a relatively simple question nearly half a century ago: Does there exist genus one curves of arbitrary index over the field of rational numbers \(\mathbb{Q}\). The importance of this question lies in the fact that although the period can be directly dealt with by the method of Galois cohomology, the index is more geometrically significant. However, the period and the index are not always equal, which makes the problem more complicated.
### Background and Motivation
For a genus one curve \(C/K\) defined over a number field \(K\), there are two numerical invariants that can quantify the degree to which \(C\) lacks \(K\)-rational points:
- **Index**: The degree of the smallest extension \(L/K\) such that \(C\) has an \(L\)-rational point; or equivalently, the degree of the smallest positive - degree \(K\)-rational divisor on \(C\).
- **Period**: The order of the cohomology class in the Weil - Châtelet group \(H^1(K,\text{Jac}(C))\) corresponding to \(C\); or equivalently, the degree of the smallest positive - degree \(K\)-rational divisor class on \(C\).
It is known that the period always divides the index, and they have the same prime factors. When \(K\) is a number field, the index can be strictly greater than the period, which makes the problem more complicated.
### Main Results
The author Pete L. Clark gives a complete affirmative answer to the Lang and Tate problem through the following theorems:
#### Theorem 1
Let \(E/K\) be an elliptic curve defined over a number field \(K\), and \(E(K) = 0\) (i.e., \(E\) has no non - trivial rational points over \(K\)). For every positive integer \(n\), there exists an element \(\eta\in H^1(K, E)\) whose index is \(n\).
#### Corollary 2
For every positive integer \(n\), there exist infinitely many genus one curves \(C/\mathbb{Q}\) defined over the field of rational numbers \(\mathbb{Q}\) whose index is \(n\).
Furthermore, through Kolyvagin's construction, the author further proves:
#### Corollary 4
For every number field \(K\) and every positive integer \(n\), there exist infinitely many genus one curves with index \(n\).
These results not only answer the Lang and Tate question but also show the richness of genus one curves over a wider range of number fields.
### Methods and Techniques
To prove these results, the author uses some key techniques and tools, including:
- **Kolyvagin classes**: A special class of cohomology classes that ensure the equality of the period and the index.
- **Tate duality theorem**: Used to study the relationship between local and global cohomology groups.
- **The finiteness assumption of the Shafarevich - Tate group**: Although it is not necessary to assume the finiteness of the Shafarevich - Tate group, in some cases, this assumption simplifies the proof process.
In conclusion, this paper solves a long - standing unsolved problem regarding the index of genus one curves through strict mathematical proofs and provides new ideas and methods for subsequent research.