So I have spent some time thinking about Fermat's Last Theorem and about how to come up with a proof for certain cases of n. To begin with I took n = 3.

$$A^3 + B^3 = C^3 (A,B,C \in \mathbb {Z})( A,B,C ≠ 0)$$ We can assume that $A<B<C$. Because of this we can rewrite this as: $$A^3 + xA^3 = C^3$$ We now have three possibilities: $x$ is an integer, $x$ is an irrational, or $x$ is a fraction.

If $x$ is an integer then $x$+1 and $x$ must be cubes. The only numbers which satisfy this are $x$ = 0. However, this means $B^3 = 0$.

$x$ cannot be irrational because then $C^3$ is irrational. Therefore $x$ is a fraction.

We can represent this now as: $$A^3 + \frac{p}{q}B^3 = C^3 (p,q \in \mathbb {Z})(GCD(p,q) = 1)$$ This means $\frac{p}{q}$ is actually a cube over a cube. Therefore $p$ and $q$ are cube numbers.

However, $1 + \frac{p}{q}$ is also a cube. Therefore $\frac{p+q}{q}$ is a cube over a cube. Hence: $$\sqrt[3]{p} ,\sqrt[3]{q}, \sqrt[3]{p+q} \in \mathbb {Z}$$

We can now create another variable $v$ for $p+q$. Therefore, $$p+q = v$$ (Where $\sqrt[3]{v} \in \mathbb {Z}$)

Now we must deal with $\frac{p}{q}+1$. Since this is a cube we can write this as: $$\frac{p}{q} + 1 = \lambda^3$$ Since $p$ and $q$ are co-prime and in simplest form, $\lambda$ and $\lambda^3$ are also fractions in simplest form. Rearranging, we get: $$p = q(\lambda^3 - 1)$$

Substituting this into $p+q=v$, we get: $$q+q(\lambda^3 - 1) = v$$ $$q\lambda^3 = v$$

Since $\lambda^3$ is a fraction, we can express it as $\frac{D^3}{E^3}$. Hence: $$qD^3 = vE^3$$

However, because $v$ is a cube, $vE^3$ is also a cube. Therefore $\lambda^3$ doesn't have to be a fraction. This is a contradiction, and so there can be no integer solutions for $p$ and $q$. Subsequently, there can be no integer solutions for A,B,C in: $$A^3 + B^3 = C^3$$.

If anyone could point out the problem in this and what you think of it it would be greatly appreciated. Thank you.

  • $\begingroup$ How do you justify the sentence “This means $\frac{p}{q}$ is actually a cube over a cube”? $\endgroup$ – José Carlos Santos Jul 22 '18 at 12:24
  • 2
    $\begingroup$ The very definition of $x$ implies that $x=\frac{B^3}{A^3}$ $\endgroup$ – random Jul 22 '18 at 12:27
  • 2
    $\begingroup$ "Therefore $\lambda^3$ doesn't have to be a fraction" ? Formulated this way, it could be a fraction, so no contradiction arises. If you mean "$\lambda^3$ cannot be a fraction", then justify why this is the case ! $\endgroup$ – Peter Jul 22 '18 at 12:30
  • $\begingroup$ Why is "$1+\frac pq$ also a cube"? $\endgroup$ – Arnaud Mortier Jul 22 '18 at 12:55
  • 2
    $\begingroup$ After the conclusion that $x$ is a fraction, shouldn't the next formula be $A^3+\frac p q A^3 = C^3$ ? $\endgroup$ – random Jul 22 '18 at 13:14

You have $$qD^3=v E^3.\tag{1}$$ But if you've followed carefully, you'd have noticed that $D=C$, $E=B$, $p=A^3$, $q=B^3$ and $v=C^3$. The conclusion given is then that "$\lambda^3$ doesn't have to be a fraction". I presume this is intended to mean that $D^3/E^3\notin\Bbb Z$. But (1) says that $B^3C^3=C^3B^3$. From this tautology, why can we deduce anything non-trivial at all, let alone $D^3/E^3\notin\Bbb Z$?


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.