1
$\begingroup$

A lattice point is a point $(x, y)$ in the plane, both of whose coordinates are integers.It is easy to see that every lattice point can be surrounded by a small circle which excludes all other lattice points from its interior. It is not much harder to see that it is possible to draw a circle that has exactly two lattice points in its interior, or exactly 3, or exactly 4. How to prove that it satisfies for n $\forall n\in\mathbb{N}$? I have tried to generalize it. If I take a the centre on the line segment joining 2 latice points ( not the diagonal). Then by increasing the radius I can draw with all the n which are even by keeping it internally within the circle. If I take the centre on the diagonal of 2 lattice points then we can easily draw for all the odd numbers. If i take x as the radius using greatest integer function we can find the no. of points in 1 quadrant. In this way I have tried it but is there a more clear solution to this problem.

$\endgroup$
1
  • $\begingroup$ You should add some more details on what you have tried. $\endgroup$
    – Debashish
    Mar 4 '17 at 8:15
5
$\begingroup$

For example you can take point $(\frac13, \pi)$ as a centre. Then distances to all lattice points from it are pairwise distinct.

Suppose this is not true. Then there are $x_1, y_1, x_2, y_2 \in \mathbb{Z}$ such that $x_1 \ne x_2 \lor y_1 \ne y_2$ and $$(x_1 - \frac13)^2 + (y_1 - \pi)^2 = (x_2 - \frac13)^2 + (y_2 - \pi)^2.$$

If $y_1 = y_2$ then $(x_1 - \frac13)^2 = (x_2 - \frac13)^2 \land x_1 \ne x_2$. That is either $x_1 - \frac13 = x_2 - \frac13$ or $x_1 - \frac13 = -x_2 + \frac13$. The first equation leads to $x_1 = x_2$ and the second one leads to that at least one of $x_1$ and $x_2$ is not integer.

So $y_1 \ne y_2$. Then $\pi = \frac{x_1^2 + y_1^2 - x_2^2 - y_2^2 - \frac23 x_1 + \frac23 x_2}{2(y_2 - y_1)}$ that is not true because $\pi$ is not rational. This contradiction finishs proof.

Now it is easy to see that increasing radius or circle with centre in $(\frac13, \pi)$ we will cover lattice points one by one.

$\endgroup$
1
  • 1
    $\begingroup$ P. S. It is much more hard to prove that for any $n$ there is a circle containing exactly $n$ lattice poins on its border. $\endgroup$
    – Smylic
    Mar 3 '17 at 19:29

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.