1

Let $m_1 ,...,m_k$ be $k$ distinct squarefree integers $\neq 1$ which are pairwise coprime.

Then is $[\mathbb{Q}(\sqrt{m_1},...,\sqrt{m_k}):\mathbb{Q}]=2^k$?

Intuitively, I guess that the statement is true, by observing the result when $k=1,2$, but I have trouble actually proving this for the general case.

What I know is that $\mathbb{Q}(\sqrt{m_1},...,\sqrt{m_k})$ is Galois over $\mathbb{Q}$, so $[\mathbb{Q}(\sqrt{m_1},...,\sqrt{m_k}):\mathbb{Q}]=|\mathrm{Gal}(\mathbb{Q}(\sqrt{m_1},...,\sqrt{m_k})/\mathbb{Q})|$ and every element of $G=\mathrm{Gal}(\mathbb{Q}(\sqrt{m_1},...,\sqrt{m_k})/\mathbb{Q})$ sends $\sqrt{m_i}$ to either itself or $-\sqrt{m_i}$ for each $i=1,...,k$, so $G$ has at most $2^k$ elements.

I think the key of the proof is to prove that if $\phi(\sqrt{m_i})$ is given, then there is always an automorphism $\phi$ of $\mathbb{Q}(\sqrt{m_1},...,\sqrt{m_k})$ having such value of $\phi(\sqrt{m_i})$, and here is where I stuck.

Does anyone have ideas? Any advice or comments will be helpful!

If it is hard to give the answer here directly, then any bibliography for reference is also acceptable.

bellcircle
  • 2,939
  • See this question: https://math.stackexchange.com/questions/1230173/elementary-proof-for-sqrtp-n1-notin-mathbbq-sqrtp-1-sqrtp-2 and the question which is linked there. – MooS Oct 29 '17 at 16:24
  • For square roots, I think that the most simple and direct proof consists in using Kummer theory, see https://math.stackexchange.com/a/1609061/300700 – nguyen quang do Nov 08 '17 at 07:57

0 Answers0