Find all arithmetic progressions where
are all squares.
So we need to find integers such that
ie.,
Hence we get a rational point on the circle
is a point on the circle and now using a pencil of lines passing though
parametrized by slope
, we see that all the point are given by
What if you fix the common difference to be ?
Congruent Numbers, Elliptic Curves:
Find rational triples such that
where
So we have the map
But
Numbers which have rational solutions on this elliptic curve with
are called congruent numbers. By the above, they exactly correspond to arithmetic progressions in squares with common difference
gives a map to right triangle
, who area
Fact: The only points of finite order on are
and
Hence we can use the group law on the curve to generate infinitely many rational points on this curve. And therefore if there is one arithmetic progression we can find infinitely many arithmetic progressions in squares with common difference
corresponds
and
on
Computing
we get the progressions
Open: How do we classify the congruent numbers ? And what about the rational solutions (what is the rank, generators of the rational points — what are all progressions in squares?
For instance, the following family of curves have rank at least 2 for almost all
INTEGERS: (Congruum Problem)
The above discussion is about rational progressions. What if we ask for the progressions to be in integers? That is we need integer solutions for
Fibonacci: All the integer solutions to these equations are given by
Proof Sketch: Observe that
hence above parametrization of the solutions comes from parametrization of Pythagorean triples.
FOUR TERMS?
What about progressions of length ?
Fermat’s four squares theorem: There are no four distinct rational squares in arithmetic progression.
We can suppose the the progression looks like , all of them squares.
is the common difference. So we have an integer point on
is a Pythagorean triple.
So we have for some which implies
and
have to be squares.
Again using Pythagoren parametrization we get and
and
and
implies that there are integers
so that
gives us that
and
are squares. Using the above argument, in reverse gets you
squares in progression with common difference
which is smaller than
the common difference we started with.
By descent by are done!
In the background of the above elementary argument is the curve– the argument amount to showing that there no rational points which corresponds to 4 term progressions in squares.(We need points of infinite order on this curve which don’t exist.)
https://kconrad.math.uconn.edu/blurbs/ugradnumthy/4squarearithprog.pdf
https://kconrad.math.uconn.edu/blurbs/ugradnumthy/3squarearithprog.pdf
https://arxiv.org/pdf/0712.3850.pdf