Remark 23.5.4. With $A \supset I$ as in Lemma 23.5.3, let $\delta : I \to I$ be a map satisfying (2) and (3) of the lemma. Then (1) holds too. First, we prove $\delta (nx) = n\delta (x) + \frac{n^ p - n}{p!}x^ p$ for $n \geq 1$ by induction on $n$. The case $n = 1$ is OK. Assume that it is OK for some $n$, then $\delta ((n + 1)x) = \delta (nx) + \delta (x) + (\sum _{i = 1}^{p - 1} \frac{n^ i}{i!(p - i)!})x^ p = (n + 1)\delta (x) + \left( \frac{n^ p - n}{p!} + \frac{(n + 1)^ p}{p!} - \frac{n^ p + 1}{p!} \right)x^ p = (n + 1)\delta (x) + \frac{(n + 1)^ p - (n + 1)}{p!}x^ p$, so it is also OK for $n + 1$. On the other hand, we have $\delta (nx) = n^ p\delta (x)$, hence $(n^ p - n)\delta (x) = \frac{n^ p - n}{p!}x^ p$. But for any $p$ there exists $n$ such that $p^2 \not\mid n^ p - n$. Namely, we have $(\mathbb {Z}/p^2\mathbb {Z})^\times \cong \mathbb {Z}/p(p-1)\mathbb {Z}$. So there exists $n$ such that $n^{p-1} \not\equiv 1 \mod p^2$. Hence we get $p!\delta (x)=x^ p$
Post a comment
Your email address will not be published. Required fields are marked.
In your comment you can use Markdown and LaTeX style mathematics (enclose it like $\pi$). A preview option is available if you wish to see how it works out (just click on the eye in the toolbar).
All contributions are licensed under the GNU Free Documentation License.
Comments (0)