\documentclass{article}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\usepackage{amsmath}
\usepackage{amssymb}
\usepackage{amsfonts}
\setcounter{MaxMatrixCols}{10}
%TCIDATA{OutputFilter=LATEX.DLL}
%TCIDATA{Version=5.00.0.2552}
%TCIDATA{}
%TCIDATA{Created=Thursday, May 06, 2004 09:06:20}
%TCIDATA{LastRevised=Monday, May 31, 2004 13:27:24}
%TCIDATA{}
%TCIDATA{}
%TCIDATA{Language=French}
%TCIDATA{CSTFile=40 LaTeX article.cst}
\newtheorem{theorem}{Theorem}
\newtheorem{acknowledgement}[theorem]{Acknowledgement}
\newtheorem{algorithm}[theorem]{Algorithm}
\newtheorem{axiom}[theorem]{Axiom}
\newtheorem{case}[theorem]{Case}
\newtheorem{claim}[theorem]{Claim}
\newtheorem{conclusion}[theorem]{Conclusion}
\newtheorem{condition}[theorem]{Condition}
\newtheorem{conjecture}[theorem]{Conjecture}
\newtheorem{corollary}[theorem]{Corollary}
\newtheorem{criterion}[theorem]{Criterion}
\newtheorem{definition}[theorem]{Definition}
\newtheorem{example}[theorem]{Example}
\newtheorem{exercise}[theorem]{Exercise}
\newtheorem{lemma}[theorem]{Lemma}
\newtheorem{notation}[theorem]{Notation}
\newtheorem{problem}[theorem]{Problem}
\newtheorem{proposition}[theorem]{Proposition}
\newtheorem{remark}[theorem]{Remark}
\newtheorem{solution}[theorem]{Solution}
\newtheorem{summary}[theorem]{Summary}
\newenvironment{proof}[1][Proof]{\noindent\textbf{#1.} }{\ \rule{0.5em}{0.5em}}
\setlength{\headsep}{5mm}
\setlength{\topmargin}{-15mm}
\setlength{\oddsidemargin}{-10mm}
\setlength{\textwidth}{190mm}
\setlength{\textheight}{245mm}
\input{tcilatex}
\begin{document}
\begin{verbatim}
ROBERT FERREOL (robert.ferreol@prepas.org)
\end{verbatim}
\begin{center}
CONCOURS CENTRALE SUPELEC MATHS 2 2004
CORRIG\'{E} POUR LE CONTR\^{O}LE A POSTERIORI
\end{center}
\bigskip
I. A. $\mathcal{M}_{2}\left( \mathbb{Z}\right) $ est un anneau car il
contient $-I_{2}$ et est stable pour $+$ et $\times .$
\bigskip
I.B.1) Dans un anneau, l'ensemble des \'{e}l\'{e}ments inversibles est
toujours un groupe multiplicatif.
\bigskip
I.B.2) Si $A$ et $A^{-1}\in \mathcal{M}_{2}\left( \mathbb{Z}\right) ,$ $\det
\left( A\right) .\det \left( A^{-1}\right) =1$ donc $D=\det \left( A\right) $
divise 1 et $\left\vert D\right\vert =1$ ; la r\'{e}ciproque se voit par la
formule $\left[
\begin{array}{cc}
a & b \\
c & d%
\end{array}%
\right] ^{-1}=\dfrac{1}{D}\left[
\begin{array}{cc}
d & -b \\
-c & a%
\end{array}%
\right] $.
\bigskip
I.C.1) $SL_{2}\left( \mathbb{Z}\right) $ est un groupe car il contient $%
I_{2} $ et $\left( A,B\in SL_{2}\left( \mathbb{Z}\right) \Rightarrow
AB^{-1}\in SL_{2}\left( \mathbb{Z}\right) \right) .$
\bigskip
I.C.2) et 3) \ $\left[
\begin{array}{cc}
3 & 5 \\
c & d%
\end{array}%
\right] \in SL_{2}\left( \mathbb{Z}\right) \Leftrightarrow \left\{
\begin{array}{c}
3d-5c=1 \\
3.2-5.1=1%
\end{array}%
\right. \Leftrightarrow 3\left( d-2\right) =5\left( c-1\right)
\Leftrightarrow \exists k\in \mathbb{Z}$ $/$ \ $\left\{
\begin{array}{c}
c=1+3k \\
d=2+5k%
\end{array}%
\right. $
\ $\left[
\begin{array}{cc}
3 & 5 \\
c & d%
\end{array}%
\right] \in GL_{2}\left( \mathbb{Z}\right) \backslash SL_{2}\left( \mathbb{Z}%
\right) \Leftrightarrow \left\{
\begin{array}{c}
3d-5c=-1 \\
3.3-5.2=-1%
\end{array}%
\right. \Leftrightarrow 3\left( d-3\right) =5\left( c-2\right)
\Leftrightarrow \exists k\in \mathbb{Z}$ $/$ \ $\left\{
\begin{array}{c}
c=2+3k \\
d=3+5k%
\end{array}%
\right. $
\bigskip
I.C.4) $\exists \left( c,d\right) \in \mathbb{Z}^{2}\ /\ \left[
\begin{array}{cc}
a & b \\
c & d%
\end{array}%
\right] \in GL_{2}\left( \mathbb{Z}\right) \Leftrightarrow PGCD\left(
a,b\right) =1$ : c'est exactement le th\'{e}or\`{e}me de B\'{e}zout.
\bigskip
I.D.1) et 2) $T$ est non diagonalisable quel que soit le corps car elle poss%
\`{e}de 1 pour valeur propre double et le sous-espace propre correspondant
n'est pas de dimension 2 ; elle est trigonalisable puisque triangulaire...
$S$ poss\`{e}de deux valeurs propres non r\'{e}elles $i$ et $-i$, elle est
donc diagonalisable dans $\mathcal{M}_{2}\left( \mathbb{C}\right) $, mais
non trigonalisable dans $\mathcal{M}_{2}\left( \mathbb{R}\right) $ ; une
forme r\'{e}duite est $D=\left[
\begin{array}{cc}
i & 0 \\
0 & -i%
\end{array}%
\right] $ et la matrice de passage $P$, telle que $S=PDP^{-1},$ est $P=\left[
\begin{array}{cc}
1 & 1 \\
i & -i%
\end{array}%
\right] $
$TS=\left[
\begin{array}{cc}
1 & -1 \\
1 & 0%
\end{array}%
\right] $ a pour polyn\^{o}me caract\'{e}ristique $X^{2}-X+1$ ; ayant deux
racines non r\'{e}elles $-j$ et $-\overline{j},$ $TS$ est diagonalisable
dans $\mathcal{M}_{2}\left( \mathbb{C}\right) $ mais non trigonalisable dans
$\mathcal{M}_{2}\left( \mathbb{R}\right) $ ; une forme r\'{e}duite est $E=%
\left[
\begin{array}{cc}
-j & 0 \\
0 & -\overline{j}%
\end{array}%
\right] $ et la matrice de passage $Q$, telle que $TS=QEQ^{-1},$ est $Q=%
\left[
\begin{array}{cc}
1 & 1 \\
-\overline{j} & -j%
\end{array}%
\right] .$
\bigskip I.E.1) et 2) Si $A^{2}=I_{2},$ $A$ est une matrice de sym\'{e}trie,
donc semblable \`{a} $\left[
\begin{array}{cc}
1 & 0 \\
0 & 1%
\end{array}%
\right] ,$ $\left[
\begin{array}{cc}
-1 & 0 \\
0 & -1%
\end{array}%
\right] $ ou $\left[
\begin{array}{cc}
1 & 0 \\
0 & -1%
\end{array}%
\right] $ ; la derni\`{e}re forme est impossible car le d\'{e}terminant vaut
-1, donc $A=\pm I_{2}$ ; la r\'{e}ciproque est facile.
\bigskip
I.F.1) si $A^{2}=-I,$ le polyn\^{o}me scind\'{e} \`{a} racines simples $%
\left( X-i\right) \left( X+i\right) $ est annulateur de $A$ donc $A$ est
diagonalisable dans $\mathcal{M}_{2}\left( \mathbb{C}\right) $ \ et donc
semblable \`{a} $\left[
\begin{array}{cc}
i & 0 \\
0 & i%
\end{array}%
\right] ,$ $\left[
\begin{array}{cc}
-i & 0 \\
0 & -i%
\end{array}%
\right] $ ou $\left[
\begin{array}{cc}
i & 0 \\
0 & -i%
\end{array}%
\right] $ \ ; la seule forme o\`{u} le d\'{e}terminant est 1 est la troisi%
\`{e}me, et l'on constate qu'alors la trace de $A$ est nulle.
\bigskip
I.F.2) D'apr\`{e}s ce qui pr\'{e}c\`{e}de,
\begin{equation*}
\left\{
\begin{array}{c}
A^{2}=-I_{2} \\
A\in SL_{2}\left( \mathbb{Z}\right)%
\end{array}%
\right. \Leftrightarrow \exists a,b,c\in \mathbb{Z}\ /\ \ A=\left[
\begin{array}{cc}
a & b \\
c & -a%
\end{array}%
\right] ,\text{avec }a^{2}+bc=-1
\end{equation*}
\bigskip
I.G.1) Exercice classique : on \'{e}crit $PU=VP$ avec $P=R+iS\in
GL_{2}\left( \mathbb{C}\right) ,$ et on montre qu'il existe $\lambda $ r\'{e}%
el tel que $Q=R+\lambda S\in GL_{2}\left( \mathbb{R}\right) ,$ d'o\`{u} $%
QU=VQ$ et $U$ et $V$ sont semblables dans $\mathcal{M}_{2}\left( \mathbb{R}%
\right) .$
\bigskip
I.G.2) Si $A\in SL_{2}\left( \mathbb{Z}\right) $ est solution de $%
A^{2}=-I_{2},$ on vu qu'elle \'{e}tait semblable \`{a} $\left[
\begin{array}{cc}
i & 0 \\
0 & -i%
\end{array}%
\right] ,$ comme $S,$ donc $A$ et $S$ sont semblables dans $\mathcal{M}%
_{2}\left( \mathbb{C}\right) $ ; elle le sont donc aussi dans $\mathcal{M}%
_{2}\left( \mathbb{R}\right) $ par ce qui pr\'{e}c\`{e}de.
\bigskip
II. A. 1) $\Lambda $ contient $0$ et est stable par soustraction : c'est
donc un groupe additif. Ce peut-\^{e}tre un anneau (par exemple pour $\alpha
=1$ et $\beta =i,$ mais pas forc\'{e}ment, par exemple pour $\alpha =1$ et $%
\beta =\pi $ (en effet $\pi ^{2}$ ne peut s'\'{e}crire $u+v\pi $ avec $u$ et
$v$ entier car $\pi $ n'est pas alg\'{e}brique).
\bigskip
II. A. 2) Remarquons d\'{e}j\`{a} que $\func{Im}\left( \dfrac{\alpha }{\beta
}\right) \neq 0$ car sinon $\left( \alpha ,\beta \right) $ serait $\mathbb{R}
$-li\'{e}e. Si $\func{Im}\left( \dfrac{\alpha }{\beta }\right) <0,$ on
change $\beta $ en $-\beta $ : le r\'{e}seau engendr\'{e} est identique : on
peut donc toujours supposer $\dfrac{\alpha }{\beta }\in \mathcal{H}.$
\bigskip
II. A. 3) $\func{Im}\left( \dfrac{az+b}{cz+d}\right) =\dfrac{1}{2i}\left(
\dfrac{az+b}{cz+d}-\dfrac{a\overline{z}+b}{c\overline{z}+d}\right) =\dfrac{1%
}{2i}\left( \dfrac{\left( ad-bc\right) \left( z-\overline{z}\right) }{%
\left\vert cz+d\right\vert ^{2}}\right) =\dfrac{ad-bc}{\left\vert
cz+d\right\vert ^{2}}\func{Im}\left( z\right) .$
\bigskip
II. B. I) \ $\Lambda $ est engendr\'{e} par $\mathcal{B=}\left( \omega
_{1},\omega _{2}\right) $ et par $\mathcal{B}^{\prime }\mathcal{=}\left(
\omega _{1}^{\prime },\omega _{2}^{\prime }\right) $ avec $\dfrac{\omega _{1}%
}{\omega _{2}}$ et $\dfrac{\omega _{1}^{\prime }}{\omega _{2}^{\prime }}\in
\mathcal{H}.$ il existe donc 8 entiers $a,b,c,d,e,f,g,h$ tels que $\left[
\begin{array}{c}
\omega _{1}^{\prime } \\
\omega _{2}^{\prime }%
\end{array}%
\right] $ =$\left[
\begin{array}{cc}
a & b \\
c & d%
\end{array}%
\right] \left[
\begin{array}{c}
\omega _{1} \\
\omega _{2}%
\end{array}%
\right] $ et $\left[
\begin{array}{c}
\omega _{1} \\
\omega _{2}%
\end{array}%
\right] $ =$\left[
\begin{array}{cc}
e & f \\
g & h%
\end{array}%
\right] \left[
\begin{array}{c}
\omega _{1}^{\prime } \\
\omega _{2}^{\prime }%
\end{array}%
\right] $ ; mais $\left[
\begin{array}{cc}
a & c \\
b & d%
\end{array}%
\right] $ et $\left[
\begin{array}{cc}
e & g \\
f & h%
\end{array}%
\right] $ sont les matrices de passage respectives de $\mathcal{B}$ \`{a} $%
\mathcal{B}^{\prime }$ et de $\mathcal{B}^{\prime }$ \`{a} $\mathcal{B},$
donc sont inverses l'une de l'autre ; $\left[
\begin{array}{cc}
a & b \\
c & d%
\end{array}%
\right] $ appartient donc \`{a} $GL_{2}\left( \mathbb{Z}\right) $ ; de plus $%
\func{Im}\left( \dfrac{\omega _{1}^{\prime }}{\omega _{2}^{\prime }}\right) =%
\dfrac{ad-bc}{\left\vert c\dfrac{\omega _{1}}{\omega _{2}}+d\right\vert ^{2}}%
\func{Im}\left( \dfrac{\omega _{1}}{\omega _{2}}\right) ,$ donc $ad-bc>0$ et
$\left[
\begin{array}{cc}
a & b \\
c & d%
\end{array}%
\right] $ appartient \`{a} $SL_{2}\left( \mathbb{Z}\right) .$
\bigskip
II. B. 2) On montre facilement la r\'{e}ciproque en prouvant que les r\'{e}%
seaux engendr\'{e}s par $\mathcal{B}$ et $\mathcal{B}^{\prime }$ sont inclus
l'un dans l'autre.
\bigskip
II. C. C'est le m\^{e}me probl\`{e}me que I. C. 3)
II. D. Si $\Lambda _{\tau }=\Lambda _{\tau ^{\prime }},$ il existe $\left[
\begin{array}{cc}
a & b \\
c & d%
\end{array}%
\right] $ $\in SL_{2}\left( \mathbb{Z}\right) $ telle que $\left\{
\begin{array}{c}
\tau ^{\prime }=a\tau +b \\
1=c\tau +d%
\end{array}%
\right. $ , or $\tau $ est non r\'{e}el donc $c=0$ et $d=1$ ; de $ad-bc=1$
on tire $a=1$ : donc $\tau ^{\prime }=\tau +b$ ; r\'{e}ciproquement, on
montre facilement que pour tout $b$ entier $\Lambda _{\tau }=\Lambda _{\tau
+b}.$
\bigskip
III. A. 1) Par II. A. 2), on sait qu'un r\'{e}seau $\Lambda $ est engendr%
\'{e} par $\left( \alpha ,\beta \right) $ avec $\tau =\dfrac{\alpha }{\beta }%
\in \mathcal{H}$ ; il est clair que $\Lambda =\beta \Lambda _{\tau },$ donc $%
\Lambda $ est semblable \`{a} $\Lambda _{\tau }.$
\bigskip
III. A. 2) \ Si \ $\Lambda _{\tau }$ et $\Lambda _{\tau ^{\prime }}$ sont
semblables, $\Lambda _{\tau ^{\prime }}=\lambda \Lambda _{\tau }$ avec $%
\lambda \in \mathbb{C}^{\ast }$ et il existe $\left[
\begin{array}{cc}
a & b \\
c & d%
\end{array}%
\right] $ $\in SL_{2}\left( \mathbb{Z}\right) $ telle que $\left[
\begin{array}{c}
\tau ^{\prime } \\
1%
\end{array}%
\right] $ =$\left[
\begin{array}{cc}
a & b \\
c & d%
\end{array}%
\right] \left[
\begin{array}{c}
\lambda \tau \\
\lambda%
\end{array}%
\right] $ ; on en d\'{e}duit $\tau ^{\prime }=\dfrac{a\tau +b}{c\tau +d}$ ;
la r\'{e}ciproque se montre en en remontant les calculs, en utilisant II. B.
2).
REM : ceci fait le lien avec la partie IV.
\bigskip
III. B. 1) \ L'ensemble $S\left( \Lambda \right) $ est l'ensemble des
coefficients $a$ des similitudes directes de centre $0$ : $z\mapsto az$
laissant stable le r\'{e}seau $\Lambda $ (plus le nombre 0, car
l'application nulle n'est pas une similitude).
\bigskip
III. B. 2) Si $\lambda $ est le rapport d'une homoth\'{e}tie de centre 0
laissant stable le r\'{e}seau $\Lambda $ de base $\left( \alpha ,\beta
\right) ,$ $\lambda \alpha =u\alpha +v\beta $ avec $u$ et$\ v$ entiers donc $%
\lambda =u$ et $\lambda $ est entier ; r\'{e}ciproquement, il est clair que
si $\lambda $ est entier, $\lambda \Lambda \subset \Lambda $ ; nous avons
montr\'{e} que $S\left( \Lambda \right) \cap \mathbb{R}=\mathbb{Z}.$
\bigskip
III. B. 3) $S\left( \Lambda \right) $ est un sous-anneau de $\mathbb{C}$ ;
en effet $-1\in S\left( \Lambda \right) ,$(car $-\Lambda =\Lambda )$ et si $%
z,z^{\prime }\in S\left( \Lambda \right) ,$
\begin{equation*}
\left( z+z^{\prime }\right) \Lambda \subset z\Lambda +z^{\prime }\Lambda
\subset \Lambda +\Lambda \subset \Lambda
\end{equation*}
et
\begin{equation*}
\left( zz^{\prime }\right) \Lambda =z\left( z^{\prime }\Lambda \right)
\subset z\Lambda \subset \Lambda
\end{equation*}
donc $z+z^{\prime }$ et $zz^{\prime }\in \Lambda .$
\bigskip
III. B. 4) $S\left( \Lambda _{\mathcal{B}}\right) =S\left( \Lambda _{\tau
}\right) $ car
\begin{equation*}
z\Lambda _{\mathcal{B}}\subset \Lambda _{\mathcal{B}}\Leftrightarrow z\omega
_{2}\Lambda _{\tau }\subset \omega _{2}\Lambda _{\tau }\Leftrightarrow
z\Lambda _{\tau }\subset \Lambda _{\tau }
\end{equation*}
\bigskip
III. B. 5) Si $z\in S\left( \Lambda _{t}\right) ,$ comme 1$\in \Lambda
_{\tau },$ $z=z\times 1\in \Lambda _{t},$ donc $S\left( \Lambda _{t}\right)
\subset \Lambda _{t}.$
\bigskip
III. C. 1) Soit $z$ un \'{e}l\'{e}ment non entier de $S\left( \Lambda
_{t}\right) $ ; comme $z$ et $\tau z$ appartiennent \`{a} $\Lambda _{\tau }$
, il existe 4 entiers $a,b,c,d$ tels que $z=a\tau +b$ et $\tau z=c\tau +d,$
donc $\tau \left( a\tau +b\right) =c\tau +d,$ et $a\tau ^{2}+\left(
b-c\right) \tau -d=0$ ; $z$ n'\'{e}tant pas entier, $a$ est non nul et $\tau
$ est racine d'un polyn\^{o}me du second degr\'{e} \`{a} coefficients
entiers.
\bigskip
III. C. 2) a) On suppose que $\tau $ v\'{e}rifie $u\tau ^{2}+v\tau +w=0$
avec $u,v,w$ entiers, $u\neq 0$ ; alors $\left( u\tau +v\right) \tau =-w$
appartient \`{a} $\Lambda _{\tau }$ donc $z=u\tau +v$ appartient \`{a} $%
S\left( \Lambda _{\tau }\right) $ ; $u$ \'{e}tant non nul, $z$ n'est pas r%
\'{e}el, donc $S\left( \Lambda _{\tau }\right) $ n'est pas contenu dans $%
\mathbb{R}$.
\bigskip
III. C. 2) b) Si $u=1,$ $\tau +v$ appartient \`{a} $S\left( \Lambda
_{t}\right) ,$ donc $\tau $ \'{e}galement ainsi que tout $a\tau +b$ avec $a$
et $\ b$ entiers : $\Lambda _{t}$ est donc inclus dans $S\left( \Lambda
_{t}\right) $ et lui est donc \'{e}gal par III. B. 5).
REM : $\Lambda _{\tau }$ est donc dans ce cas un sous-anneau de $\mathbb{C}$.
\bigskip
IV. A. 1) $\func{Im}\left( g\left( z\right) \right) =\dfrac{1}{\left\vert
cz+d\right\vert ^{2}}\func{Im}\left( z\right) $, donc $\func{Im}\left(
z\right) >0\Rightarrow \func{Im}\left( g\left( z\right) \right) >0$ et $%
g\left( \mathcal{H}\right) \subset \mathcal{H}.$
\bigskip
IV. A. 2) On montre par un calcul simple que $\left( \Phi \left( A\right)
\circ \Phi \left( A^{\prime }\right) \right) \left( z\right) =\Phi \left(
AA^{\prime }\right) \left( z\right) .$ $\Gamma $ est donc stable pour la loi
$\circ .$
\bigskip
IV. A. 3) $\ \Phi \left( A\right) \circ \Phi \left( A^{-1}\right) =\Phi
\left( I_{2}\right) =id_{\mathcal{H}}$ donc $\Phi \left( A\right) $ est
bijective et $\left( \Phi \left( A\right) \right) ^{-1}=\Phi \left(
A^{-1}\right) .$ $\Phi $ est donc un morphisme du groupe $SL_{2}\left(
\mathbb{Z}\right) $ dans le groupe de bijections de $\mathcal{H}$, et son
image $\Gamma $ est donc un groupe.
\bigskip
IV. A. 4)
\begin{eqnarray*}
\Phi \left( A\right) &=&id_{\mathcal{H}}\Leftrightarrow \forall z\in
\mathcal{H}\ \ \ z=\dfrac{az+b}{cz+d}\Leftrightarrow \forall z\in \mathcal{H}%
\ \ \ cz^{2}+dz=az+b\Leftrightarrow \left\{
\begin{array}{c}
c=0 \\
d=a \\
b=0%
\end{array}%
\right. \\
&\Leftrightarrow &\left\{
\begin{array}{c}
b=c=0 \\
a=d=\pm 1%
\end{array}%
\right. \Leftrightarrow A=\pm I_{2}
\end{eqnarray*}
\bigskip
IV. A. 5)
a) $\Phi \left( A\right) =\Phi \left( A^{\prime }\right) \Leftrightarrow
\Phi \left( AA^{\prime -1}\right) =id_{\mathcal{H}}\Leftrightarrow
AA^{\prime -1}=\pm I_{2}\Leftrightarrow \fbox{$A^{\prime }=\pm A$}.$
b) $ST=\left[
\begin{array}{cc}
0 & -1 \\
1 & 1%
\end{array}%
\right] $ et $TS=\left[
\begin{array}{cc}
1 & -1 \\
1 & 0%
\end{array}%
\right] $ donc $ST\neq \pm TS$ d'o\`{u} $\Phi \left( S\right) \circ \Phi
\left( T\right) \neq \Phi \left( T\right) \circ \Phi \left( S\right) $ et $%
\Gamma $ n'est pas commutatif.
\bigskip
IV. B. 1) L'\'{e}quation demand\'{e}e est une traduction de $\left\vert
z-\omega \right\vert ^{2}=R^{2}$ ; $\mathcal{C}\left( \omega ,R\right)
\subset H\Leftrightarrow R<\func{Im}\omega .$
\bigskip
IV. B. 2) On a $s\left( z\right) =s^{-1}\left( z\right) =-\dfrac{1}{z}$ ;
donc
\begin{equation*}
z\in s\left( \mathcal{C}\left( \omega ,R\right) \right) \Leftrightarrow -%
\dfrac{1}{z}\in \mathcal{C}\left( \omega ,R\right) \Leftrightarrow
\left\vert z\right\vert ^{2}+\dfrac{\overline{\omega }\overline{z}+\omega z}{%
\left\vert \omega \right\vert ^{2}-R^{2}}+\dfrac{\left\vert \omega
\right\vert ^{2}}{\left( \left\vert \omega \right\vert ^{2}-R^{2}\right) ^{2}%
}=\dfrac{R^{2}}{\left( \left\vert \omega \right\vert ^{2}-R^{2}\right) ^{2}}
\end{equation*}
donc $s\left( \mathcal{C}\left( \omega ,R\right) \right) =\mathcal{C}\left( -%
\dfrac{\overline{\omega }}{\left\vert \omega \right\vert ^{2}-R^{2}},\dfrac{R%
}{\left\vert \omega \right\vert ^{2}-R^{2}}\right) $
\bigskip
IV. C. 1) On a $z\in \mathcal{D}\Leftrightarrow \func{Im}z=\beta ,$ d'o\`{u}
\begin{equation*}
z=x+iy\in s\left( \mathcal{D}\right) \Leftrightarrow s^{-1}\left( z\right)
\in \mathcal{D}\Leftrightarrow \func{Im}\left( -\dfrac{1}{z}\right) =\beta
\Leftrightarrow \dfrac{y}{x^{2}+y^{2}}=\beta \Leftrightarrow \left\{
\begin{array}{c}
x^{2}+\left( y-\dfrac{1}{2\beta }\right) ^{2}=\dfrac{1}{\left( 2\beta
\right) ^{2}} \\
x^{2}+y^{2}\neq 0%
\end{array}%
\right.
\end{equation*}
donc $s\left( D\right) =\mathcal{C}\left( \dfrac{i}{2\beta },\dfrac{1}{%
2\beta }\right) \backslash \left\{ 0\right\} $
\bigskip
IV. C. 2) On a $z\in \mathcal{D}\Leftrightarrow \left\{
\begin{array}{c}
\func{Re}z=\alpha \\
\func{Im}z>0%
\end{array}%
\right. ,$ d'o\`{u}
\begin{eqnarray*}
z &=&x+iy\in s\left( \mathcal{D}\right) \Leftrightarrow s^{-1}\left(
z\right) \in \mathcal{D}\Leftrightarrow \left\{
\begin{array}{c}
\func{Re}\left( -\dfrac{1}{z}\right) =\alpha \\
\func{Im}\left( -\dfrac{1}{z}\right) >0%
\end{array}%
\right. \Leftrightarrow \left\{
\begin{array}{c}
\dfrac{x}{x^{2}+y^{2}}=-\alpha \\
y>0%
\end{array}%
\right. \\
&\Leftrightarrow &\left\{
\begin{array}{c}
\left( x+\dfrac{1}{2\alpha }\right) ^{2}+y^{2}=\dfrac{1}{\left( 2\alpha
\right) ^{2}}\text{ si }\alpha \neq 0 \\
y>0%
\end{array}%
\right.
\end{eqnarray*}
donc $s\left( \mathcal{D}\right) =\mathcal{C}\left( -\dfrac{1}{2\alpha },%
\dfrac{1}{2\alpha }\right) \cap \mathcal{H}$ si $\alpha \neq 0,$ la
demi-droite $Oy$ sinon.
\bigskip
IV. D) $\mathcal{F}$ est la portion de bande verticale ferm\'{e}e de largeur
1 centr\'{e}e sur $Oy$ limit\'{e}e inf\'{e}rieurement par le cercle trigonom%
\'{e}trique ; comme $t\left( z\right) =z+1,t\left( \mathcal{F}\right) $ et $%
t^{-1}\left( \mathcal{F}\right) $ sont des translat\'{e}s de $\mathcal{F}$
d'une unit\'{e} vers la droite et vers la gauche.
\bigskip
IV. E. 1) Toute partie born\'{e}e, contient un nombre fini de noeuds du r%
\'{e}seau $\Lambda _{\tau }$ (par exemple parce qu'on peut placer en chaque
noeud du r\'{e}seau un disque centr\'{e} en ce noeud de rayon fix\'{e} ne
contenant pas d'autre noeud que ce noeud) ; donc l'ensemble $\left\{
\left\vert c\tau +d\right\vert ^{2}\ \ /\ \ \left( c,d\right) \in \mathbb{Z}%
^{2}\ \right\} $ a une intersection finie avec tout segment $[0,U],U>0,$
donc a fortiori \'{e}galement l'ensemble $\left\{ \left\vert c\tau
+d\right\vert ^{2}\ /\ \ A=\left[
\begin{array}{cc}
a & b \\
c & d%
\end{array}%
\right] \in SL_{2}\left( \mathbb{Z}\right) \ \text{\ et }\Phi \left(
A\right) \in G\right\} $; cet ensemble poss\`{e}de donc un minimum, atteint
pour un couple $\left( c_{0},d_{0}\right) $ ; notons $A_{0}=\left[
\begin{array}{cc}
a_{0} & b_{0} \\
c_{0} & d_{0}%
\end{array}%
\right] $ $\ $la matrice correspondante et $g_{0}=\Phi \left( A_{0}\right) $
; soit alors $g\in G$ ; on a bien%
\begin{equation*}
\func{Im}\left( g\left( \tau \right) \right) =\dfrac{1}{\left\vert c\tau
+d\right\vert ^{2}}\func{Im}\left( z\right) \leqslant \dfrac{1}{\left\vert
c_{0}\tau +d_{0}\right\vert ^{2}}\func{Im}\left( \tau \right) =\func{Im}%
\left( g_{0}\left( \tau \right) \right)
\end{equation*}
\bigskip
IV.E.2) $\tau ^{\prime }=g_{0}\left( \tau \right) $ ; remarquons que $\func{%
Re}\left( t^{m}\left( \tau ^{\prime }\right) \right) =\func{Re}\tau ^{\prime
}+m$ ; or \'{e}tant donn\'{e} un r\'{e}el $x$ il existe toujours un entier $%
m $ (l'entier le plus proche de $-x)$ tel que $\left\vert x+m\right\vert
\leqslant \dfrac{1}{2}$ ; donc il existe un entier $m$ et que $\left\vert
\func{Re}\left( t^{m}\left( \tau ^{\prime }\right) \right) \right\vert
\leqslant \dfrac{1}{2}.$
\bigskip
\QTP{Body Math}
IV. E. 3) Soit $\tau ^{\prime \prime }=t^{m}\left( \tau ^{\prime }\right) $
;\ $\func{Im}\tau ^{\prime \prime }=\func{Im}\tau ^{\prime }=\func{Im}%
g_{0}\left( \tau \right) \geqslant \func{Im}g\left( \tau \right) $ pour tout
$g\in G$ d'apr\`{e}s la d\'{e}finition de $g_{0}$ ; or $s\left( \tau
^{\prime \prime }\right) =g\left( \tau \right) $ avec $g=s\circ t^{m}\circ
g_{0}\in G$ ; donc $\func{Im}\tau ^{\prime \prime }\geqslant \func{Im}%
g\left( \tau \right) =\func{Im}s\left( \tau ^{\prime \prime }\right) =\dfrac{%
\func{Im}\tau ^{\prime \prime }}{\left\vert \tau ^{\prime \prime
}\right\vert ^{2}}$ ; on en d\'{e}duit bien $\left\vert \tau ^{\prime \prime
}\right\vert ^{2}\geqslant 1,$ soit $\left\vert \tau ^{\prime \prime
}\right\vert \geqslant 1$ \ et $\tau ^{\prime \prime }\in \mathcal{F}.$
\QTP{Body Math}
$\bigskip $
\QTP{Body Math}
IV. F. 4) Soit $g$ un \'{e}l\'{e}ment de $\Gamma $ ; si l'on prouve que $%
g\in G,$ on aura prouv\'{e} $G=\Gamma .$
\QTP{Body Math}
Soit $\tau $ un point int\'{e}rieur \`{a} $\mathcal{F}$ et consid\'{e}rons
un \'{e}l\'{e}ment $g_{0}\in G$ d\'{e}fini dans IV. E. 2) associ\'{e} \`{a} $%
g\left( \tau \right) $ au lieu de $\tau $ ; soit $m$ d\'{e}fini dans IV. E.
2) tel que $t^{m}\circ g_{0}\left( g\left( \tau \right) \right) \in \mathcal{%
F}$ ; d'apr\`{e}s le r\'{e}sultat admis dans l'\'{e}nonc\'{e} (qui est
"visuellement" \'{e}vident...), il est impossible que $t^{m}\circ g_{0}\circ
g$ soit diff\'{e}rent de $id_{H}$ car sinon $\tau $ ne serait pas un point
int\'{e}rieur \`{a} $\mathcal{F}$. Donc $t^{m}\circ g_{0}\circ g=id_{H}$ et $%
g=g_{0}^{-1}\circ \tau ^{-m}\in G$ \ \ C.Q.F.D.
\QTP{Body Math}
$\bigskip $
\end{document}