5
$\begingroup$

I have a book that has a brief history of the complex numbers and it covers de Moivre's formula:

$(\cos(x) + i\sin(x))^n = \cos(nx) + i\sin(nx)$.

I am very curious as to how this result was originally found, or derived, BEFORE Euler's formula was around. Also, what was the original proof of this?

$\endgroup$
  • 3
    $\begingroup$ Probably by induction. $\endgroup$ – dustin Feb 6 '15 at 1:46
  • 2
    $\begingroup$ Induction indeed...and basic trigonometric formulas, of course. $\endgroup$ – Timbuc Feb 6 '15 at 1:52
  • 2
    $\begingroup$ @Christopher have answer working on putting it together. $\endgroup$ – dustin Feb 6 '15 at 2:06
  • $\begingroup$ I was puzzled by this result when I saw it because the way I was introduced to multiplication of complex numbers was that one multiplies the absolute values and adds the angles, so this formula seemed like a complete triviality. $\endgroup$ – Michael Hardy Feb 6 '15 at 2:41
6
$\begingroup$

Euler was the first to write down de Moirve's formula which I found in The Elementary Mathematical Works of Leonhard Euler starting on page 29. Euler found $$ (\cos(x)\pm i\sin(x))(\cos(y)\pm i\sin(y))(\cos(z)\pm i\sin(z)) = \cos(x \pm y \pm z)\pm i\sin(x\pm y\pm z) $$ from here he deduce what we call de Moirve's theorem $$ (\cos(x)\pm i\sin(x))^n = \cos(nx)\pm i\sin(nx) $$ Then it follows that \begin{align} \cos(nx) &= 1/2[(\cos(x)+ i\sin(x))^n+(\cos(x) - i\sin(x))^n]\tag{1}\\ \sin(nx) &= 1/2[(\cos(x)+ i\sin(x))^n-(\cos(x) - i\sin(x))^n]\tag{2} \end{align} Euler than expanded equations $(1)$ and $(2)$ by the binomial theorem. He then let $x$ be a small angle approximation and $n$ be infinitely large so that $xn$ was finite. Let $xn=v$. Then $\sin(x) = x = v/n$ and we get \begin{align} \cos(v) &= 1 - \frac{v^2}{2!} + \cdots\\ \sin(v) &= v - \frac{v^3}{3!} + \cdots \end{align} Using equations $(1)$ and $(2)$ and letting $j = n$ be the infinitely large number so that $jx=v$ so that $nx = v = v/j$. Now we have $\sin(x) = v/j$ and $\cos(x)=1$. With these substitutions, we get \begin{align} \cos(v) &=\frac{(1+iv/j)^j+(1-iv/j)^j}{2}\\ \sin(v) &=\frac{(1+iv/j)^j-(1-iv/j)^j}{2i} \end{align} In the linked document, the author who compiled the document (Euler) previously has shown that $(1+z/j)^j = e^z$. Letting $z=\pm iv$, we get \begin{align} e^{iv} &= \cos(v) + i\sin(v)\\ e^{-iv} &= \cos(v) - i\sin(v) \end{align}

$\endgroup$
  • 1
    $\begingroup$ Great answer and great link! Thank you so much. $\endgroup$ – Christopher Feb 6 '15 at 2:39
  • $\begingroup$ @Christopher you probably want to go through the link since I left out the some details as well as the binomial expansion of $(1)$ and $(2)$. $\endgroup$ – dustin Feb 6 '15 at 2:39
  • 1
    $\begingroup$ @dustin Very nice approach! $\endgroup$ – Math-fun Feb 16 '15 at 20:49
  • $\begingroup$ @Mehdi unfortunately, Euler gets all the credit here. I have no idea why this was named after de Moirve though. It mentions he was working on some equations but I don't see the relation or why he gets the credit. $\endgroup$ – dustin Feb 16 '15 at 20:50
  • $\begingroup$ @dustin well he has been great (and is). By the way I just found an interesting youtube video on Euler: youtube.com/watch?v=h-DV26x6n_Q $\endgroup$ – Math-fun Feb 16 '15 at 20:54

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.