Every congruence class modulo except is represented in this list, which incorrectly suggests that if then is not a congruent number. Though no with is a congruent number, is a congruent number congruent and .
Deciding whether an integer is a congruent number can be subtle since the simplest triangle with area can be very complicated. For example, as Zagier pointed out, the number is a congruent number, and the ``simplest'' rational right triangle with area has side lengths
This solution would be difficult to find by a brute force search.
We call congruent numbers ``congruent'' because of the following proposition, which asserts that any congruent number is the common ``congruence'' between three perfect squares.
are all perfect squares of rational numbers.
The main motivating open problem related to congruent numbers, is to give a systematic way to recognize them.
Fortunately, the vast theory developed about elliptic curves has something to say about the above problem. In order to understand this connection, we begin with an elementary algebraic proposition that establishes a link between elliptic curves and the congruent number problem.
and
given explicitly by the maps
and
For , let be the elliptic curve .
Multiplying through by yields the side lengths of a rational right triangle with area . Are there any others?
Observe that we can apply to any point in with . Using the group law we find that , and
Example 6.5.9 foreshadows the following theorem.
Tunnell has proved that the Birch and Swinnerton-Dyer conjecture (alluded to above), implies the existence of an elementary way to decide whether or not an integer is a congruent number. We state Tunnell's elementary way in the form of a conjecture.
If is odd and square free then is a congruent number if and only if
Enough of the Birch and Swinnerton-Dyer conjecture is known to prove one direction of Conjecture 6.5.12. In particular, it is a very deep theorem that if we do not have equality of the displayed cardinalities, then is not a congruent number. For example, when ,
The even more difficult (and still open!) part of Conjecture 6.5.12 is the converse: If one has equality of the displayed cardinalities, prove that is a congruent number. The difficulty in this direction, which appears to be very deep, is that we must somehow construct (or prove the existence of) elements of . This has been accomplished in some cases do to groundbreaking work of Gross and Zagier ([#!gross-zagier!#]) but much work remains to be done.
The excellent book [#!koblitz:cong!#] is about congruent numbers and Conjecture 6.5.12, and we encourage the reader to consult it. The Birch and Swinnerton-Dyer conjecture is a Clay Math Institute million dollar millennium prize problem (see [#!cmi!#,#!wiles:cmi!#]).
William 2007-06-01