1: 

2: \documentclass{amsart}

3: \usepackage{amsmath, amsthm, amsopn, amsfonts}

4: %\usepackage{showkeys}

5: \usepackage{graphicx}

6: \usepackage{graphpap}

7: %\usepackage[english, francais]{babel}

8: 

9: %\textwidth= 14.5 cm

10: %\textheight= 20.5cm

11: %\hoffset = -1cm

12: % UNE MACRO POUR PLACER DU TEXTE SUR UNE BOITE. 

13: % SYNTAXE: \ecriture{boite}{

14: % \aat{coordhoriz}{coordvert}{texte}} la macro trace un quadrillage 

15: %de 50

16: % pour faciliter le placement du texte. On le supprime apres mise au

17: % point par la commande \pasdegrille en debut de fichier

18: 

19: 

20: \def\finpreuve{\hfill $\Box$\par}

21: \def\pasdegrille{\let\grille = \pasgrille}

22: \def\ecriture#1#2{\setbox1=\hbox{#1}

23: \dimen1= \wd1

24: \dimen2=\ht1

25: \dimen3=\dp1

26: \grille #2 \box1 }

27: \def\aat#1#2#3{

28: \divide \dimen1 by 48

29: \dimen3=\dimen1

30: \multiply \dimen1 by #1

31: \advance \dimen1 by -\dimen3

32: \divide \dimen1 by 101

33: \multiply \dimen1 by 100

34: \divide \dimen2 by \count11

35: \multiply \dimen2 by #2 

36: \setbox0=\hbox{#3}\ht0=0pt\dp0=0pt

37:   \rlap{\kern\dimen1 \vbox to0pt{\kern-\dimen2\box0\vss}}\dimen1= \wd1

38: \dimen2=\ht1}

39: \def\pasgrille{

40: \count12= \dimen1 

41: \divide \count12 by 50

42: \divide \dimen2 by \count12

43: \count11 =\dimen2

44: \ 

45: \divide \dimen1 by 48

46: \setlength{\unitlength}{\dimen1}

47: \smash{\rlap{\ }}

48: \dimen1= \wd1

49: \dimen2=\ht1

50: }

51: \def\grille{

52: \count12= \dimen1 

53: \divide \count12 by 50

54: \divide \dimen2 by \count12

55: \count11 =\dimen2

56: \ 

57: \divide \dimen1 by 48

58: \setlength{\unitlength}{\dimen1}

59: \smash{\rlap{\graphpaper[1](0,0)(50, \count11)}}

60: \dimen1= \wd1

61: \dimen2=\ht1

62: }

63: 

64: \pasdegrille

65: %%%%

66: %%%

67: % FIN DE KLA

68: %MACRO POUR INCLURE DU TEXTE DANS UNE FIGURE

69: %

70: %

71: 

72: 

73: \usepackage{amssymb}

74: %\usepackage{draftcopy}

75: %\ifx\pdfoutput\undefined

76: %\usepackage{graphicx}  \DeclareGraphicsExtensions{.ps}

77: %\def\Dessin#1{\begin{center}

78: %\includegraphics{#1.mps}

79: %\end{center}

80: %}

81: %\else

82: \usepackage{amsfonts}

83: \usepackage{graphicx}  

84: \DeclareGraphicsExtensions{.pdf,.mps}

85: %\def\Dessin#1{\begin{center}

86: %\includegraphics{#1.mps}

87: %\end{center}

88: %}

89: %\fi

90: 

91: \newcommand{\PSbox}[1]{\includegraphics[0in,0in][3in,3in]{#1}}

92: \usepackage[T1]{fontenc}

93: 

94: % Added by JW

95: \setlength{\marginparwidth}{.8in}

96: \newcommand{\mar}[1]{{\marginpar{\sffamily{\scriptsize #1}}}} 

97: \usepackage{epic, eepic}

98: % Have to temporarily widen margins to make marginal notes fit!

99: \setlength{\textheight}{8in} \setlength{\oddsidemargin}{0.35in}

100: \setlength{\evensidemargin}{0.35in} \setlength{\textwidth}{5.8in}

101: \setlength{\topmargin}{0.18in} \setlength{\headheight}{0.18in}

102: \setlength{\marginparwidth}{1.0in}

103: 

104: \newcommand{\abs}[1]{{\left\lvert{#1}\right\rvert}}

105: \newcommand{\norm}[1]{{\left\lVert{#1}\right\rVert}}

106: \newcommand{\bignorm}[1]{{\Bigl\lVert{#1}\Bigr\rVert}}

107: \newcommand{\pa}{{\partial}}

108: \newcommand{\h}{{\frac 12}}

109: \newcommand{\cinf}{{\mathcal{C}}^\infty}

110: 

111: \def\mathbbone{{\mathchoice {1\mskip-4mu \text{l}} {1\mskip-4mu \text{l}}

112: { 1\mskip-4.5mu \text{l}} { 1\mskip-5mu \text{l}}}}

113: 

114: \newcommand{\bc}{\mathbb C}

115: \newcommand{\br}{\mathbb R}

116: \newcommand{\bz}{\mathbb Z}

117: \newcommand{\im}{\operatorname{\rm Im}}

118: \newcommand{\mn}[1]{\Vert#1\Vert}

119: \newcommand{\op}{\operatorname{Op}}

120: \newcommand{\re}{\operatorname{\rm Re}}

121: \newcommand{\set}[1]{\left\{\,#1\,\right\}}

122: \newcommand{\w}[1]{\langle #1\rangle }

123: 

124: 

125: \def\squarebox#1{\hbox to #1{\hfill\vbox to #1{\vfill}}} 

126: \newcommand{\stopthm}{\hfill\hfill\vbox{\hrule\hbox{\vrule\squarebox 

127:                  {.667em}\vrule}\hrule}\smallskip} 

128: 

129: \pagestyle{headings}

130: 

131: 

132: 

133: 

134: 

135: %\pagestyle{plain}

136: 

137: \newcommand{\1}{{\bold 1}} 

138: \newcommand{\F}{{\mathcal F}} 

139: \newcommand{\CC}{{\mathbb C}}

140: \newcommand{\CI}{{\mathcal C}^\infty }

141: \newcommand{\HH}{{\mathcal H}}

142: \newcommand{\CIc}{{\mathcal C}^\infty_{\rm{c}} }

143: \newcommand{\CIb}{{\mathcal C}^\infty_{\rm{b}} }

144: \newcommand{\Op}{{\operatorname{Op}^{{w}}_h}}

145: \newcommand{\Oo}{{\mathcal O}} 

146: \newcommand{\Hh}{{\mathcal H}}

147: \newcommand{\pic}{{\mbox{Pic}}}

148: \newcommand{\Z}{{\mathbb Z}}

149: \newcommand{\Q}{{\mathbb Q}}

150: \newcommand{\RR}{{\mathbb R}}

151: \newcommand{\SP}{{\mathbb S}}

152: \newcommand{\ZZ}{{\mathbb Z}}

153: \newcommand{\NN}{{\mathbb N}}

154: \newcommand{\MO}{{\mathcal M}}

155: \newcommand{\U}{{\mathcal U}}

156: \newcommand{\N}{{\mathbb N}}

157: \newcommand{\vol}{\operatorname{vol}}

158: \newcommand{\rank}{\operatorname{rank}}

159: \newcommand{\itt}{\operatorname{it}}

160: \newcommand{\KKer}{\operatorname{ker}}

161: \newcommand{\supp}{\operatorname{supp}}

162: \newcommand{\Ran}{\operatorname{Ran}}

163: \newcommand{\UC}{U_{\circlearrowright}}

164: \newcommand{\QC}{Q_{\circlearrowright}}

165: \newcommand{\comp}{\operatorname{comp}}

166: \newcommand{\loc}{\operatorname{loc}}

167: \newcommand{\tr}{\operatorname{tr}}

168: \newcommand{\rest}{\!\!\restriction}

169: \newcommand{\ttt}{|\hspace{-0.25mm}|\hspace{-0.25mm}|}

170: \renewcommand{\Re}{\mathop{\rm Re}\nolimits}

171: \renewcommand{\Im}{\mathop{\rm Im}\nolimits}

172: 

173: % JW messed with the following:

174: \theoremstyle{plain}

175: \def\Rm#1{{\rm#1}}

176: \newtheorem{thm}{Theorem}

177: %   \renewcommand{\thethm}{\arabic{thm}}

178: \newtheorem{prop}{Proposition}[section]

179: %   \renewcommand{\theprop}{\arabic{prop}}

180: \newtheorem{cor}[prop]{Corollary}

181: %   \renewcommand{\thecor}{}

182: \newtheorem{lem}{Lemma}[section]

183: %   \renewcommand{\thelem}{\arabic{lem}}

184: %\numberwithin{lem}{section}

185: 

186: %\numberwithin{prop}{section}

187: \theoremstyle{definition}

188: \newtheorem{ex}{EXAMPLE}[section]

189: \newtheorem{exmple}{Example}[section]

190: \newtheorem{rem}{Remark}

191: \newtheorem{defn}[prop]{Definition} 

192: 

193: 

194: 

195: \newcommand{\thmref}[1]{Theorem~\ref{#1}}

196: \newcommand{\secref}[1]{Section~\ref{#1}}

197: \newcommand{\lemref}[1]{Lemma~\ref{#1}}

198: \newcommand{\exref}[1]{Example~\ref{#1}}

199: \newcommand{\corref}[1]{Corollary~\ref{#1}}

200: \newcommand{\propref}[1]{Proposition~\ref{#1}}

201: 

202: \def\bbbone{{\mathchoice {1\mskip-4mu \rm{l}} {1\mskip-4mu \rm{l}}

203: { 1\mskip-4.5mu \rm{l}} { 1\mskip-5mu \rm{l}}}}

204: 

205: 

206: \title

207: {Bouncing ball modes and quantum chaos}

208: \author[N. Burq]{Nicolas Burq}

209: \address{Universit{\'e} Paris Sud,

210: Math{\'e}matiques,

211: B{\^a}t 425, 91405 Orsay Cedex}

212: \email{Nicolas.burq@math.u-psud.fr}

213: \author[M. Zworski]{Maciej Zworski}

214: \address{Mathematics Department, University of California \\

215: Evans Hall, Berkeley, CA 94720, USA}

216: \email{zworski@math.berkeley.edu}

217: 

218: 

219: \usepackage{amssymb}

220: \usepackage{amsmath, amsthm, amsopn, amsfonts}

221: %\usepackage[dvips]{graphicx}

222: 

223: 

224: 

225: 

226: 

227: 

228: 

229: 

230: \def\Im{\textrm{Im}}

231: \def\Re{\textrm{Re}} 

232: \def\11{{\rm 1~\hspace{-1.4ex}l} }

233: \def\R{\mathbb R}

234: \def\C{\mathbb C}

235: \def\Z{\mathbb Z}

236: \def\N{\mathbb N}

237: \def\E{\mathbb E}

238: \def\T{\mathbb T}

239: \def\O{\mathbb O}

240: \def\K{\mathbb K}

241: \def\cal{\mathcal}

242: 

243: 

244: \begin{document}    

245: \begin{abstract}

246: Quantum ergodicity of classically chaotic systems has been 

247: studied extensively both theoretically and experimentally,

248: in mathematics, and in physics. Despite this long tradition

249: we are able to present a new rigorous result using only elementary 

250: calculus. In the case of the famous Bunimovich billiard table shown in 

251: Fig.1 we prove that the wave functions have to spread into any neighbourhood of the

252: wings.

253: \end{abstract}

254: 

255: \maketitle   

256: 

257: 

258: The quantum/classical correspondence is a puzzling issue that has 

259: been with us since the advent of quantum mechanics hundred years ago.

260: Many aspects of it go back to the Newton/Huyghens debate over

261: the wave vs. corpuscular theories of light.

262: 

263: On the surface of our existence we live in a

264: world governed by laws of classical physics. That does not

265: mean that we know precisely how fluids flow or solids move. They are 

266: described by highly non-linear rules which are hard to unravel mathematically.

267: Even the simplest classical motion, that of a ball bouncing elastically

268: from confining walls poses many unanswered questions -- 

269: see {\small \tt http://www.dynamical\- systems.org/billiard/} for a fun 

270: introduction. 

271: 

272: If we investigate deeper, or if we simply use any modern technical device,

273: we come in contact with quantum mechanics. It is governed by a different

274: set of rules which mix wave and matter. The simplest description

275: of a wave comes from solving the Helmholtz equation:

276: \[  ( - \Delta - \lambda^2 ) 

277: u = 0 \,, \ \ \Delta = \partial_x^2 + \partial_y^2 \,, 

278: \ \ ( x, y ) \in \Omega \,, \ \ u \rest_{\partial \Omega } = 0 \,. \]

279: Here we put our wave inside of a two dimensional region $ \Omega $.

280: In classical wave mechanics the limit $ \lambda

281:  \rightarrow \infty $ is described

282: using geometrical optics where the waves propagate along straight lines

283: reflecting in the boundary $ \partial \Omega $. Roughly speaking, we expect

284: something similar in the classical/quantum correspondence

285: with the Helmholtz equation replaced by  its quantum mechanical

286: version, the Schr\"odinger equation. For many fascinating

287: illustrations of this we refer to the web art gallery of Rick Heller:

288: {\small \tt http://www.ericjhellergallery.com}.

289: 

290: \begin{figure}[htb]

291: \includegraphics[width=9cm]{buni.ps}

292: \caption{An experimental image of bouncing ball modes

293: in a Bunimovich stadium cavity -- see \cite{ChHu96} and

294: {\small \tt http://www.bath.ac.uk/$\sim$pyscmd/acoustics}. With 

295: a certain amount imagination one can see our theorem in this

296: picture.}

297: \label{fig:bath}

298: \end{figure}

299: 

300: Many researchers on different aspects of 

301: semiclassical analysis have been interested in 

302: the the correspondence of solutions to the equation above and the 

303: classical geometry of balls bouncing from the walls of $ \Omega $:

304: B\"acker, Cvitanovi\'c, 

305: Eckhardt, Gaspard, Heller, Sridhar, in physics, and Colin de Verdi\`ere,

306: Melrose, Sj\"ostrand, Zelditch,  in mathematics, to mention some

307: (see \cite{BaScSt97},\cite{He01}

308:  for references to the physics literature, and \cite{La93},\cite{MeSj78-1},\cite{Ze03}

309: for mathematics).

310: %\nocite{He01,LaTe81, La93,KaHe98,FaNo03,BoLi03,Sa95, Cr93,  BaLeRa92, MeSj78-1, Do03,Sh74}

311: \renewcommand\thefootnote{\dag}% \small

312: 

313: Billiard tables for which the motion is chaotic are a particularly 

314: interesting model to study\footnote{Of course one would not 

315: want to play billiards on a table like that, and a completely 

316: integrable rectangular one can pose enough of a challenge. 

317: While discussing billiards and the classical/quantum correspondence we 

318: cannot resist mentioning that 

319: Pyotr Kapitsa (Nobel Prize in Physics '78) 

320: was fond of saying that trying

321: to detect the quantum nature of physical processes at room temperature 

322: was like trying to investigate the physical laws governing the collision 

323: of billiard

324: balls on a table aboard a ship going through rough seas.}. One of the

325: most famous is the Bunimovich billiard table shown in Fig.1. By adding two 

326: circular ``wings'' to a rectangular table the motion of a reflecting billiard

327: ball becomes chaotic, or more precisely, hyperbolic, in the sense that 

328: changes in initial conditions lead to exponentially large changes in motion as

329: time goes on. 

330: 

331: As a model for studying quantum phenomena in chaotic systems this billiard

332: table has become popular in experimental physics. A genuinely quantum example

333: is shown in Fig.2 -- it comes from the scanning tunnelling microscope work of

334: Eigler, Crommie, and others~\cite{Cr93}.

335: 

336: \begin{figure}[ht]

337: \includegraphics[width=10cm]{st.ps}

338: \caption{Quantum corral in the shape of the Bunimovich stadium. Courtesy of 

339: IBM{\tiny\textregistered}  Research.}\label{fig:qc}

340: \end{figure}

341: 

342: One question which is still mysterious to mathematicians and physicists alike

343: is if the states of this system (that is, solutions of the equation above)

344: can concentrate on the highly ustable closed orbits of the classical billiard.

345: {\em Quantum unique ergodicity} states that there is no such concentration

346: -- see \cite{Sa95},\cite{Sh74},\cite{Ze03} and references given there. In the arithmetic case, that

347: is for billiards given by arithmetic surfaces where the motion is given by 

348: the geodesic flow, spectacular advances have been recently achieved by Bourgain, 

349: Lindenstrauss \cite{BoLi03}, and Sarnak, while for the popular 

350: quantization of the {\em Arnold cat map} impressive results were

351: produced by Bonechi, De Bi\`evre, Faure, and Nonnenmacher \cite{FaNo03}, and also by 

352: Kurlberg and Rudnick \cite{KuRu01}.

353: 

354: Here we describe an elementary but 

355: striking result in the billiard case.

356: It follows from adapting the first author's

357: earlier work in control theory. Although motivated by the more general

358: aspects of \cite{BuZw03} we give a simple self contained proof.

359: 

360: For a state $ u ( x , y ) $ what counts is its {\em probability density},

361: $ | u ( x, y ) |^2 dx dy $ -- we assume here that $ |u|^2 

362:  $ is normalized to have

363: integral $ 1 $ over $ \Omega $. We say that it is bounded from below in a region if its

364: integral over that region is bounded from below by a positive constant. 

365: With this terminology we have, roughly speaking,

366: 

367: \medskip

368: 

369: \noindent

370: {\bf Theorem.} {\em For any normalized state of the Bunimovich billiard table,

371: the probability density in any neighbourhood of the

372:  wings of the table is bounded from below independently

373: of the energy $ \lambda^2 $.}

374: 

375: \medskip

376: In particular, the result says that single bouncing ball orbits (that

377: is orbits following an interval perpendicular to the horizontal straight

378: boundaries) cannot produce localized waves. Our result allows concentration 

379: on the full

380: invariant set of all vertical orbits over $ R $ -- that is 

381: consistent with the existing physical literature -- 

382: both numerical and experimental -- see \cite{BaScSt97} and \cite{ChHu96}.

383: In \cite{BuZw03} we show a stronger result, namely that the 

384: neighbourhood of the wings can be replaced by any neighbourhood of

385: the vertical intervals between the wings and the rectangular part.

386: The proof of that predictable (to experts) improvement is however no longer

387: elementary and is based on \cite{BaLeRa92}. 

388: 

389: 

390: The proof of theorem depends on the following unpublished result of the

391: first author (see \cite{BuZw03} for detailed references and background

392: material):

393: 

394: \medskip

395: \noindent

396: {\bf Proposition.} {\em 

397: Let $\Delta = \partial_x^2 + \partial_y^2 $, be 

398: the Laplace operator on the rectangle 

399: $R= [0, 1]_{x} \times [0,a]_{y}$. Then for any open $\omega\subset R$ 

400: of the form $ \omega= \omega_{x} \times [0,a]_{y}$ , there 

401: exists $C$ such that for any solutions of

402: \[

403: (-\Delta - \lambda^2) u = f + \partial_x g \ \text{ on $R$}\,, \ \ 

404:  u \rest_{\partial R}=0 \,, 

405: \] 

406: with an arbitrary $ \lambda \geq 0 $

407: we have 

408: \[

409: \int_R | u ( x , y ) |^2 dx dy \leq C \left( 

410: \int_R ( | f ( x , y )^2 | + | g ( x , y ) |^2 ) dx dy + 

411: \int_\omega | u ( x , y ) |^2 dx dy \right) \,. 

412: \]} 

413: \begin{proof}

414: We 

415: decompose $u$, and $ f + \partial_x g $ in terms of 

416: the basis of $L^2([0,a])$ formed by the Dirichlet eigenfunctions

417: $e_{k}(y)=  { \sqrt {{2}/a}}\sin(2k\pi y/a)$,

418: \begin{equation}

419: u(x,y)= \sum_{k}e_{k}(y) u_{k}(x), \qquad f(x,y) + 

420: \partial_x g ( x, y ) = \sum_{k}e_{k}(y) ( f_{k}(x) + \partial_x g_k ( x )) 

421: \end{equation}

422: we get for $u_{k}, f_{k}$ the equation

423: \[ 

424: \left(\partial_{x}^2 + z \right)

425: u_{k}= f_{k} 

426: + \partial_x g_k ,\qquad u_{k}(0)=u_k(1)=0 \,, \ \ 

427: z = \lambda^2 - \left({2k\pi}/{a}\right)^2 \,.

428: \] 

429: It is now easy to see that 

430: \begin{equation}

431: \label{eq:easy} 

432: \int_0^1 |u_{k}( x ) |^2 dx \leq C \left( \int_0^1 ( |f_k ( x ) |^2 + 

433: | g_k ( x ) |^2 ) dx + \int_{\omega_x } |u_k ( x ) 

434: |^2 dx \right) \,,   \end{equation}

435: where $ C $ is independent of $ \lambda_1 $. 

436: In fact, let us first assume that $ \omega_x = (0, \delta) $, $ \delta > 0 $,

437: and 

438: $ z = \lambda_1^2  $, with $ \Im \lambda_1 \leq C $. We then 

439: choose $ \chi \in \CIc ([ 0,1] ) $ identically zero near $ 0 $ and

440: identically one on $ [\delta/2,1] $. Then 

441: \[ \left(\partial_{x}^2 + \lambda_1^2 \right)

442: (\chi u_{k}) = F_{k} \,, \ \ F_k = \chi ( f_k + \partial_x g_k ) + 

443: 2 \partial_x \chi \partial_x u_k + \partial^2_x \chi u_k \,.\]

444: We can now use the explicit solution given by 

445: \[ \chi (x) u_k ( x ) = \frac{1}{\lambda_1} \int_0^x \sin( \lambda_1 ( x - y )) 

446: F_k ( y ) dy \,. \]

447: All the terms with $ \partial_x g _k $ and $ \partial_x u _k $ 

448: can be converted to $ g_k $ and $ u_k $ by integration by parts

449: (with boundary terms $ 0$ at both ends). Due to the $ \lambda^{-1}_1 $ 

450: factor that produces no loss and the estimate follows. The argument is 

451: symmetric under the $ x \mapsto -x $ change, so we can place our control

452: interval anywhere.

453: 

454: It remains to discuss the case $ z \leq - C < 0 $. Then the estimate

455: \eqref{eq:easy} follows from integration by parts (where now we do 

456: not need $ \omega_x $):

457: \[ \begin{split}

458: & \int_0^1 \left( f_{k} ( x ) \overline{u_{k} ( x)}  -

459:  g_{k}  ( x ) \overline{\partial_x u_{k} ( x)} \right) dx = 

460:  \int_0^1 ( f_{k} ( x ) + \partial_x g_{k} ( x )  ) \overline{u_{k}(x)} dx = \\

461: & \int_0^1 ( -\partial^2_x - z ) u_{k} ( x) \overline{u_{k} ( x )} dx 

462: = \int_0^1 \left( |\partial_x u_{k} ( x)|^2 + |z|

463: | u_{k} ( x)|^2 \right) dx \,. 

464: \end{split} \]

465: By the Cauchy-Schwartz inequality, the left hand side is bounded from above by 

466: \[  \left( \int_0^1 \left( |f_{k} ( x ) |^2 + | g_{k} (  x) |^2 \right)dx 

467:  \right)^{\frac12} 

468: \left( \int_0^1 \left( |u_{k} ( x ) |^2  + | \partial_x u_{k}  ( x ) |^2 \right) dx 

469:  \right)^{\frac12} \,.

470: \]

471: Since $ |z| > C > 0 $, \eqref{eq:easy} follows from elementary inequalities

472: (see \cite[Lemma 4.1]{BuZw03} for a general microlocal argument).

473:  We can now sum the estimate in $k$ to obtain the proposition.

474: \end{proof}

475: 

476: We can now present a more precise version of the theorem. For a yet

477: finer version we refer the reader to \cite[Theorem 3$'$]{BuZw03} and 

478: \cite[Fig.5]{BuZw03}.

479: 

480: \medskip

481: \noindent

482: {\bf Theorem${}'$.}{\em \ Consider $\Omega$ the Bunimovich stadium 

483: associated to a rectangle $R$. With the convention of Fig.1, let 

484: $ R_1 $ be any rectangle with the horizontal sides contained in the

485: sides of $ R $, strictly contained in $R $, and 

486: with $ R \setminus R_1 $ having two components.

487: 

488: There exists a constant $ C $ depending only on 

489:  $ \Omega $ and $ R_1 $ such that for any solution of the equation 

490: \[ (- \Delta - \lambda^2 )v =f\,, \ \  u \rest_{\partial \Omega}= 0 \,,  \ \ 

491: \lambda \geq 0 \,, \]

492: we have 

493: \[ \int_\Omega | v ( x , y ) |^2 dx dy \leq C\left( \int_\Omega|f ( x , y ) |^2 

494: d x dy + \int_{ \Omega \setminus R_1 }  |v ( x , y ) |^2 d x dy \right) \,.

495: \]

496: }

497: 

498: The ``wings'' of the billiard table in the original statement are given by 

499: $ \Omega \setminus R_1 $. We apply the second theorem with $ f = 0 $ 

500: to obtain the first one.

501: 

502: \begin{proof}

503: Let us take $x,y$ as 

504: the coordinates on the stadium, so that $x$ is the horizontal direction, 

505: $y$ the vertical direction, and the internal rectangle is 

506: $[0,1]_{x}\times [0,a]_{y}$. Let us then consider $u$ and $ f$  satisfying

507:  $(- \Delta- \lambda^2 )u =f$, $ u = 0 $  on the boundary of

508: the stadium, and $\chi(x)\in \CIc (0,1)$ equal to $1$ on $[\varepsilon, 1- \varepsilon]$. Then $\chi(x) u(x,y)$ is solution of

509: \[ (- \Delta- \lambda^2 )\chi u =\chi f + [\Delta, \chi] u \ \text{ in $R$}

510: \] 

511: with Dirichlet boundary conditions on $\partial R$.

512: Since $ [\Delta , \chi ] u = 2 \partial_x( \chi' u )  - \chi'' u  $ 

513: we can apply 

514: the proposition to obtain 

515: \[

516: \int_R |\chi ( x ) u ( x ,y ) |^2  dx dy  \leq 

517: C \left( \int_R | \chi( x )  f ( x, y ) |^2 dx dy  

518: +  \int_{ \omega_{\varepsilon} } | u ( x , y) |^2 dx dy \right)

519: \]

520: where $\omega_{\varepsilon}$ is a neighbourhood of the support of 

521: $ \partial_x\chi$. Since we can choose it to be contained in $ R 

522: \setminus R_1 $, the theorem follows.

523: \end{proof}

524: 

525: We conclude by remarking that the same argument holds in the

526: setting discussed recently in \cite{Do03} and \cite{Ze03}, since

527: in the argument above the rectangle can be replaced by a torus.

528: 

529: \medskip

530: \noindent

531: {\sc Acknowledgments.} The authors would like to thank the National 

532: Science Foundation for partial support under the grant  DMS-0200732.

533: They are also grateful to Arnd B\"acker, Stephan De Bi\`evre, Richard 

534: Melrose, 

535: and Steve Zelditch for helpful 

536: discussions and comments. They are also grateful to Paul Chinnery,

537: Victor Humphrey, Don Eigler, and The IBM Corporation for the figures.

538: 

539: \def\cprime{$'$}

540: \begin{thebibliography}{10}

541: 

542: \bibitem{ChHu96}

543: P.A. Chinnery and V.F. Humphrey.

544: \newblock Experimental visualization of acoustic resonances within a

545:   stadium-shaped cavity.

546: \newblock {\em Physical Review E}, 53:272--276, 1996.

547: 

548: \bibitem{BaScSt97}

549: A.~B\"acker, R.~Schubert, and R.~Stifter.

550: \newblock On the number of bouncing ball modes in billiards.

551: \newblock {\em J. Phys. A: Math. Gen}, 30:6783--6795, 1997.

552: 

553: \bibitem{He01}

554: E.J. Heller.

555: \newblock Quantum physics: Air juggling and other tricks.

556: \newblock {\em Nature}, 412(33), 2001.

557: \newblock News and Views.

558: 

559: \bibitem{La93}

560: V.F. Lazutkin.

561: \newblock {\em KAM theory and semiclassical approximation to eigenfunctions},

562:   volume~24 of {\em Ergebnisse der Mathematik und ihree Grenzgebiete}.

563: \newblock Springer Verlag, Berlin, 1993.

564: \newblock With an addendum by A.I. Shnirelman.

565: 

566: \bibitem{MeSj78-1}

567: R.B. Melrose and J.~Sj{\"o}strand.

568: \newblock Singularities of boundary value problems {I \& II}.

569: \newblock {\em Communications in Pure Applied Mathematics}, 31 \& 35:593-- 617

570:   \& 129--168, 1978 \& 1982.

571: 

572: \bibitem{Ze03}

573: S.~Zelditch.

574: \newblock Quantum unique ergodicity.

575: \newblock {\em {P}reprint}, 2003.

576: \newblock http://www.arXiv.org/abs/math-ph/0301035.

577: 

578: \bibitem{Cr93}

579: M.~F. Crommie, C.~P. Lutz, and D.~M Eigler.

580: \newblock Confinement of electrons to quantum corrals on a metal surface.

581: \newblock {\em Science}, 262(218), 1993.

582: 

583: \bibitem{Sa95}

584: P.~Sarnak.

585: \newblock Arithmetic quantum chaos.

586: \newblock In {\em The Schur lectures (1992) (Tel Aviv)}, volume~8 of {\em

587:   Israel Math. Conf. Proc.}, pages 183--236. Bar-Ilan Univ., Ramat Gan, 1995.

588: 

589: \bibitem{Sh74}

590: A.I. Shnirelman.

591: \newblock Ergodic properties of eigenfunctions.

592: \newblock {\em Uspekhi Mat. Nauk}, 29:181--182, 1974.

593: 

594: \bibitem{BoLi03}

595: J.~Bourgain and E.~Lindenstrauss.

596: \newblock Entropy of quantum limits.

597: \newblock {\em Comm. Math. Phys.}, 233(1):153--171, 2003.

598: 

599: \bibitem{FaNo03}

600: F.~Faure, S.~Nonnenmacher, and S.~De~Bievre.

601: \newblock Scarred eigenstates for quantum cat maps of minimal periods.

602: \newblock {\em {P}reprint}, 2003.

603: \newblock http://www.arXiv.org/nlin.CD/0207060.

604: 

605: \bibitem{KuRu01}

606: P.~Kurlberg and Z.~Rudnick.

607: \newblock Value distribution for eigenfunctions of desymmetrized quantum maps.

608: \newblock {\em Internat. Math. Res. Notices}, (18):985--1002, 2001.

609: 

610: \bibitem{BuZw03}

611: N.~Burq and M.~Zworski.

612: \newblock Control in the presence of a black box.

613: \newblock {\em {P}reprint}, 2003.

614: \newblock http://www.arXiv.org/abs/math.AP/0304184.

615: 

616: \bibitem{BaLeRa92}

617: C.~Bardos, G.~Lebeau, and J.~Rauch.

618: \newblock Sharp sufficient conditions for the observation, control and

619:   stabilization of waves from the boundary.

620: \newblock {\em {S.I.A.M.} Journal of Control and Optimization}, 305:1024--1065,

621:   1992.

622: 

623: \bibitem{Do03}

624: H.~Donnelly.

625: \newblock Quantum unique ergodicity.

626: \newblock {\em Proc. Amer. Math. Soc.}, 131:2945--2951, 2003.

627: 

628: \end{thebibliography}

629: 

630: \end{document}

631: 

632: