In lecture we were told to think about the following:
Kummer's Lemma: Let $p$ be an odd prime and let $\zeta := e^{2\pi i / p}$. Every unit of $\mathbb{Z}[\zeta]$ is of the form $r\zeta^g$, where $r$ is real and $g$ is an integer.
This theorem says that the units of $\mathbb{Z}[\zeta]$ can be thougth as the points of the lines that pass through the vertices of an equally spaced $p$-gon centered at the origin. (Said vertices are of the form $\zeta^g$.)
However, we know that $1+\zeta$ is a unit. But this number does not lie on one of those lines. What is wrong here?
Edit: This question seems to originate from Stewart & Tall's "Algebraic Number Theory and Fermat's Last Theorem" (4th edition), where it appears as Ex. 11.6 (p. 199).
Let the odd prime number be $p=2k+1$. Consider the number $$ w = \zeta^k(1+\zeta)\ . $$ Then $$ \begin{aligned} \bar w &=\overline{\zeta^k(1+\zeta)} =\bar\zeta^k(1+\bar\zeta) =\zeta^{-k}(1+\zeta^{-1}) \\ & =\zeta^{p-k}\frac 1\zeta(\zeta+1) =\zeta^{k+1}\frac 1\zeta(\zeta+1) =\zeta^k(\zeta+1)=w\ . \end{aligned} $$ So $w$ is a real number, and $1+\zeta=w\;\zeta^{-k}=w\;\zeta^{k+1}$.