1: \documentclass[english]{article}

2: %%%\usepackage[dvips]{graphicx}

3: 

4: \usepackage{geometry} % see geometry.pdf on how to lay out the page. There's lots.

5: \geometry{a4paper} % or letter or a5paper or ... etc

6: % \geometry{landscape} % rotated page geometry

7: \usepackage{graphicx}

8: \usepackage{amssymb}

9: \usepackage{epstopdf}

10: 

11: \usepackage{babel}

12: \usepackage{inputenc}

13: \usepackage{epsfig}

14: 

15: 

16: \textwidth 16cm

17: \textheight 21cm

18: \oddsidemargin 0.0in

19: \topmargin 0.3in

20: 

21: \newcommand{\Qto}       {Q_{\mathrm {t_{out}}}}

22: \newcommand{\Qts}       {Q_{\mathrm {t_{side}}}}

23: \newcommand{\Ql}        {Q_{\mathrm \ell}}

24: \newcommand{\qz}        {q_0}

25: \newcommand{\qt}        {q_{\mathrm t}}

26: \newcommand{\ql}        {q_{\mathrm \ell}}

27: \newcommand{\kt}        {k_{\mathrm t}}

28: \newcommand{\Rto}       {R_{\mathrm {t_{out}}}}

29: \newcommand{\Rts}       {R_{\mathrm {t_{side}}}}

30: \newcommand{\Rlong}     {R_{\mathrm {long}}}

31: \newcommand{\Rlongto}   {R_{\mathrm {long,t_{out}}}}

32: \newcommand{\Rz}        {R_0}

33: \newcommand{\Rt}        {R_{\mathrm t}}

34: \newcommand{\Rl}        {R_{\mathrm \ell}}

35: \newcommand{\eto}       {\epsilon_{\mathrm {t_{out}}}}

36: \newcommand{\ets}       {\epsilon_{\mathrm {t_{side}}}}

37: \newcommand{\elong}     {\epsilon_{\mathrm {long}}}

38: \newcommand{\dz}        {\delta_0}

39: \newcommand{\dt}        {\delta_{\mathrm t}}

40: \newcommand{\dl}        {\delta_{\mathrm \ell}}

41: \newcommand{\YYK}       {Y_{\mathrm {YK}}}

42: 

43: 

44: 

45: \begin{document}

46: 

47: 

48: 

49: \begin{center}

50: {\Large {\bf Bose-Einstein Correlations in Multihadron Events at LEP}}

51: 

52: \vspace{0.8cm}

53: 

54: \normalsize{

55: C. Ciocca, M. Cuffiani, G. Giacomelli \\

56: %\vspace{0.5cm}

57: Dip. di Fisica of the University of Bologna and INFN Sezione di Bologna, 

58: I-40127 Bologna, Italy;\\ 

59: ciocca@bo.infn.it, cuffiani@bo.infn.it, giacomelli@bo.infn.it \\ } 

60: 

61: \par~\par

62: %\vspace{0.8cm}

63: 

64: {\bf Invited paper at the ``Ninth Workshop on Non Perturbative QCD''\par

65: Institut d'Astrophysique de Paris, Paris, France, 4-8 June 2007}

66: 

67: \par~\par

68: %\vspace{0.8cm}

69: 

70: \end{center}

71: 

72: %\vspace{1cm}

73: 

74: {\bf Abstract.} {\normalsize

75: Bose-Einstein correlations in pairs of identical particles were analyzed in 

76: $e^+ e^-$ multihadron annihilations at $\sim$91.2 GeV at LEP. The first studies involved 

77: identical charged pions and the emitting source size was determined. 

78: Then the study of charged kaons suggested that 

79: the radius depends on the mass of the emitted particles. Subsequenty the 

80: dependence of the source radius on the event 

81: multiplicity was analyzed. The study of the correlations in neutral pions and 

82: neutral kaons extended these concepts to neutral particles. The shape of the 

83: source was analyzed in 3 dimensions and was found not to be 

84: spherically symmetric. In recent studies at LEP the correlations 

85: were analyzed in intervals of the average pair transverse momentum and of the pair 

86: rapidity to study the correlations between the pion production points and their 

87: momenta (position-momentum correlations). The latest $e^+ e^-$ data are consistent with 

88: an expanding source.}

89: 

90: \section{Introduction}

91: Bose-Einstein Correlations (BECs) are a quantum machanical phenomenon which 

92: manifests in final multihadron states as an enhanced probability for identical 

93: bosons to be emitted with small relative four momentum Q, compared with non 

94: identical bosons under similar kinematic conditions \cite{brown, gold, cuffiani}. 

95: From the measured effect it is possible to determine the space 

96: time dimensions of the boson-emitting source. 

97: The BEC effect  arises from the ambiguity of path between 

98: sources and detectors and the requirement to symmetrise the wave function of two 

99: or more identical bosons. \par

100: 

101: In 1954 the radioastronomers R. Hanbury-Brown and R. Q. Twiss proposed a new 

102: interferometry technique to measure the angular dimension of a star. 

103: It required to measure the mixed intensities in two 

104: radiotelescopes; the dependence of the correlation on the distance between them 

105: yielded the angular diameter of the astronomical source 

106: \cite{brown}. 

107: G. Goldhaber et al applied the same principle in particle physics, 

108: in $\bar p p$ annihilations into two identical charged pions, obtaining the radius of the emitting source, $\sim$1 fm \cite{gold}. \par

109: 

110: The first LEP analyses  on BECs concerned identical charged pions, assuming a spherical emission source and yielded the size of the source (R$\sim$1 fm) and the 

111: chaoticity parameter \cite{cuffiani, acton}. Neutral $\pi^{\circ}$ were then 

112: considered \cite{abbiendi}.

113: Then the study was extended to neutral and charged kaons in order to determine if the source radius depends on the mass of the emitted particles \cite{kk, alexander}.

114: 

115: 

116: \begin{figure}

117:  \begin{minipage}[b]{8.5cm}

118:    \centering

119:    \includegraphics[width=8cm]{scheme.eps}

120: \begin{quote}   

121: \caption{\small Scheme of a measurement of BECs. a, b are two sources 

122: separated by a distance R; A, B are two detectors separated by a distance 

123: L. The 

124: emitted particles go from sources to detectors as a$\rightarrow$A, 

125: b$\rightarrow$B or as a$\rightarrow$B, b$\rightarrow$A. In astronomy L$<<$R, 

126: in particle physics L$>>$R.}

127: \label{scheme}

128: \end{quote} 

129:  \end{minipage}

130:  %\ \hspace{2mm} \hspace{3mm} \

131:  \begin{minipage}[b]{7.5cm}

132:   \centering

133:    \includegraphics[width=8.5cm]{signal.eps}

134: \begin{quote}   

135: \caption{\small Distribution of two-photon invariant mass, $M_{2 \gamma}$. The smooth curves are the total Monte Carlo expectation (solid line) and the background expectation (dashed line). The shaded region is the selected window for the $\pi^0$ signal. }

136: \label{signal}

137: \end{quote}

138:  \end{minipage}

139: \end{figure}

140: 

141: Further analyses were performed to establish if the emitting source radius depends on the particle multiplicity \cite{gg}. Other studies involved the search for BE correlations in multipions \cite{ackerstaff}.\par

142: 

143: BECs were studied in two and three dimensions and one discovered that the emitting source is not spherical \cite{multipions}. Many studies were made for WW correlations \cite{GA} and also in $\nu$ interactions \cite{allasia}. 

144: Finally they involved the study of expanding sources and trials to determine the emission time \cite{momentum}.\par

145: 

146: We shall make a brief survey of these studies, concentrating finally on the very recent works which prove that even in $e^+ e^-$ collisions one has expanding sources.

147: 

148: 

149: \section{Experimental Procedure}

150: A detailed description of the OPAL experiment may be found  in ref \cite{ahmet}. The most important subdetector for BEC studies is the Central Tracking 

151: Detector, the Jet Chamber. For $\pi^0$ studies we needed also the barrel electromagnetic calorimeter. 

152: A sample of 4.3 million multihadronic events from $Z^0$ decays were used. 

153: A set of quality cuts was applied and one used cuts specific for BEC studies.\par

154: 

155: First, the event thrust axis was computed, using tracks with a

156: minimum of 20 hits in the jet chamber, a minimum transverse momentum of 

157: 150 MeV and a maximum momentum of 65 GeV.

158: Clusters in the electromagnetic calorimeter were used for energies

159: exceeding 100 MeV in the barrel or 200 MeV in the endcaps.

160: Only events well contained in the detector were accepted,

161: requiring $|{\rm cos}\theta_{\mathrm {thrust}}|<0.9$,

162: where $\theta_{\mathrm {thrust}}$ is

163: the polar angle of the thrust axis with respect to the beam axis. Tracks were required to have a maximum momentum of 40 GeV and to

164: originate from the interaction vertex.

165: Electron-positron pairs from photon conversions were rejected.

166: The selected events contained a minimum of five 

167: tracks and were reasonably balanced in charge, i.e. 

168: $|n_{\mathrm {ch}}^{+}-n_{\mathrm {ch}}^{-}|/(n_{\mathrm {ch}}^{+}+

169: n_{\mathrm {ch}}^{-}) \leq 0.4$,

170: where $n_{\mathrm {ch}}^{+}$ and $n_{\mathrm {ch}}^{-}$ are the number

171: of positive and negative charge tracks, respectively.

172: About 3.7 million events were left after all cuts.

173: All charged particle tracks that passed the selections were used,

174: the pion purity being approximately 90\%. \par

175: 

176: Since in multihadron events more than 90\% of the measured tracks are charged pions, the study of BEC for like-sign charged pion pairs was usually performed without proper particle identification and without purity correction. This choice 

177: introduces a small error in the chaoticity parameter $\lambda$ and in the 

178: radius R of the emitting region. However some analyses were performed with 

179: properly identified pions. That required some effective cuts on the fraction of the global solid angle acceptance.\par

180: 

181: For $\pi^0 \pi^0$, $K^{\pm}K^{\pm}$ and $K^0 K^0$ correlations, 

182: particle identification was necessary \cite{abbiendi, kk}. 

183: Fig. 2 shows the distribution of the two-photon invariant mass and the $\pi^0$ event 

184: selection. 

185: 

186: 

187: \section{BECs from a static source}

188: {\bf BECs in one dimension.} The measured BEC function is defined as the ratio 

189: $~C(Q)=\rho(Q)/\rho_0(Q)~$, were $Q$ is a Lorenz-invariant variable expressed in 

190: terms of the two pion four momenta $p_1$ and $p_2$ as 

191: $Q^2$=-($p_1$ - $p_2$)$^2$, $\rho(Q) = (1/N) dN/dQ$ is the measured $Q$ 

192: distribution of the two pions and $\rho_0 (Q)$ 

193: is a reference distribution which should contain all the correlations included 

194: in $\rho(Q)$, except BECs. For the determination of $\rho_0(Q)$, 

195: different methods were used: for identical $\pi^+ \pi^+$ and $\pi^- \pi^-$ one used 

196: the $\pi^+ \pi^-$ sample, but also the event mixing reference sample, 

197: where pion pairs are formed from pions belonging to different events; 

198: also a Monte Carlo (MC) reference sample without BECs was used. \par

199: 

200: The correlation distribution $C(Q)$ was parametrised using the Fourier 

201: transform of the expression for a static sphere of emitters with a Gaussian density:

202: \begin{equation}

203: C(Q) =N(1+ \lambda exp(-R^2 Q^2))(1 + \delta Q + \epsilon Q^2).

204: \end{equation}

205: 

206: \noindent $\lambda$  is the chaoticity parameter, $R$ is the radius of the source, 

207: and $N$ a normalization factor. The empirical term $(1 + \delta Q + \epsilon Q^2)$ 

208: accounts for the behaviour of the correlation function at high $Q$ due to 

209: any remaining long-range correlation.

210: The largest difference among results from different experiments lies in the choice 

211: of the reference sample: the statistical errors on $R$ is small, but the systematic 

212: uncertainty is large. \par

213: 

214: Fig 3 shows a typical distribution of $C(Q)$ versus $Q$; it is relative to 

215: a three dimensional analysis, but the observed features are typical of all BECs: notice 

216: the BEC peak at low $Q$ and the tail at large $Q$; the solid line is a fit 

217: to eq. 1, excluding the $Q$-intervals indicated in the figure, which contain 

218: effects from known hadron resonances.\par

219: 

220: 

221: \begin{figure}

222:  \begin{minipage}[b]{8cm}

223:    \centering

224:    \includegraphics[width=6cm]{figura3.eps}

225: \begin{quote}

226: \caption{\small The BEC distribution $C'(Q)$ for charged pions vs $\Qts$. 

227: The smooth solid curve is the fitted correlation function (the excluded 

228: regions contain effects from known hadron resonances). }

229: \label{fig3}

230: \end{quote}   

231:  \end{minipage}

232: % \ \hspace{2mm} \hspace{3mm} \

233:  \begin{minipage}[b]{8.5cm}

234:   \centering

235:    \includegraphics[width=8.5cm]{figura4.eps}

236:    \begin{quote}

237: \caption{\small Radius of the emitting region for BECs of 2 identical bosons and 

238: for FDCs of 2 identical baryons produced in $e^+ e^-$ collisions at LEP. }

239: \label{fig4}

240: \end{quote}

241:  \end{minipage}

242: \end{figure}

243: 

244: The same analysis was repeated for $KK$ BECs. A similar analysis was performed on 

245: Fermi Dirac Correlations (FDCs) for identical fermions: in this case there is no 

246: peak at small values of $Q$, but a dip. The analysis gives the radius of the 

247: emitting regions as shown in Fig. 4. Note that there probably is a decrease of 

248: $R$ with increasing mass of the emitted identical particles.\par

249: 

250: Fig. 5a shows the variation of the emitting radius with the charged multiplicity of the event \cite{gg}: there is an increase of about 10\% of the radius when the 

251: multiplicity increases from 10 to 40 charged hadrons in 

252: the final state. This may be related to the number of hadron jets: one has 

253: $R_{4jets} > R_{3jets} > R_{2jets}$. Notice that there is a corresponding 

254: decrease of the chaoticity parameter, Fig. 5b.\par

255: 

256: 

257: 

258: \begin{figure}

259:  \begin{minipage}[b]{8.5cm}

260:    \centering

261:    \includegraphics[width=10.5cm]{figura5.eps}

262: \begin{quote}

263: \caption{\small a) Increase of the emitting radius with increasing event 

264: multiplicity and b) decrease of $\lambda$. }

265: \label{fig5}

266: \end{quote}   

267:  \end{minipage}

268: %\ \hspace{2mm} \hspace{3mm} \

269:  \begin{minipage}[b]{8.5cm}

270:  \centering

271:    \includegraphics[width=9cm]{figura7.eps}

272:    \begin{quote}

273: \caption{\small The distributions in $\Qto$ and $\Qts$ are broader than in $\Ql$: 

274: thus $\Rl > \Rts \sim \Rto$.  }

275: \label{fig7}

276: \end{quote}

277:  \end{minipage}

278: \end{figure}

279: 

280: 

281: In ref. \cite{multipions} it was found that there are 3$\pi$ BECs, that is 

282: after removing the effect of 2$\pi$ correlations on the 3$\pi$ sample. 

283: The present situation is consistent with the relation 

284: 

285: \begin{equation}

286: R_{3\pi} = R_{2\pi} / \sqrt{2}

287: \end{equation}

288: 

289: In ref. \cite{ackerstaff} it was found that there are true multiparticle correlations up to 5$\pi$.\par

290: 

291: 

292: \begin{figure}

293:  \begin{minipage}[b]{8.5cm}

294:    \centering

295:    \includegraphics[width=9cm]{figura6.eps}

296: \begin{quote}   

297: \caption{\small Comparison of the BE Correlation functions for PbPb and 

298: $e^+ e^-$ collisions. }

299: \label{fig6}

300: \end{quote}

301:  \end{minipage}

302:  %\ \hspace{2mm} \hspace{3mm} \

303:  \begin{minipage}[b]{8.5cm}

304:   \centering

305:    \includegraphics[width=6.5cm]{pr423_08.eps}

306:    \begin{quote}

307: \caption{\small $\YYK$ vs pion pair rapidity $Y$. Vertical bars 

308: include statistical and systematic errors. $\YYK$=$Y$ corresponds 

309: to a source expanding boost-invariantly.}

310: \label{fig:yy}

311: \end{quote}

312:  \end{minipage}

313: \end{figure}

314: 

315: 

316: \noindent {\bf BECs in two and three dimensions.} 

317: Multidimensional static analyses were performed in 3 dimensions using the 

318: Longitudinal Center of Mass System (LCMS): the sum of the impulses of the emitted 

319: $q \bar q$ pair lies in the plane perpendicular to the event axis, defined by the $q \bar q$ direction. 

320: The components of the  3-dimensional distribution in the longitudinal, 

321: out and side projections indicate that the last ones are larger, see Fig. 6. Thus the longitudinal radius is about 20\% larger than the transverse radius: the 

322: emitting source is ellissoidical, elongated in the $q \bar q$ direction.\par

323: 

324: \noindent {\bf Comparison of BECs in $e^+ e^-$ and Nucleus-Nucleus collisions.}

325: Fig. 7 shows the BEC functions in $e^+ e^- \rightarrow$ hadrons  and 

326:  Pb Pb $\rightarrow$ hadrons: note how much narrower is the 

327: distribution in Pb Pb collisions: the distribution yields a radius 

328: $R \simeq 6-7$ fm for the emissions of pion pairs.

329: 

330: 

331: \section{Expanding sources}

332: BECs have been analyzed in Nucleus-Nucleus collisions in order to find evidence for expanding sources due to the formation of a quark-gluon deconfined plasma \cite{aggraval, adams}. 

333: Expanding sources may arise in $e^+ e^-$ collisions because of string fragmentation \cite{geiger}. In order to study BECs in non static, expanding sources we analyze the correlation functions

334: 

335: 

336: \begin{equation}

337: C'=\frac{C^{\rm DATA}}{C^{\rm MC}}=

338: \frac{N^{\rm DATA}_{\rm like}/N^{\rm DATA}_{\rm unlike}}

339: {N^{\rm MC}_{\rm like}/N^{\rm MC}_{\rm unlike}},

340: \end{equation}

341: 

342: in bins of the average pair four-momentum with respect to the event thrust direction

343: 

344: \begin{equation}

345: k_t =(\vec{p}_{{\mathrm t},1} + \vec{p}_{{\mathrm t},2}) 

346: \end{equation}

347: 

348: and of the pair rapidity:

349: 

350: \begin{equation}

351: |Y| = \frac{1}{2} \ln \left[

352: {\frac{(E_1+E_2)+(p_{{\mathrm \ell},1}+p_{{\mathrm \ell},2})}

353: {(E_1+E_2)-(p_{{\mathrm \ell},1}+p_{{\mathrm \ell},2})}}\right]

354: \end{equation}

355: 

356: The experimental distributions in $dN / d|Y|$ and $dN /dk_t$ 

357: are in good agreement with the distributions from the Jetset Monte Carlo. 

358: The dependences of $C$ and $C'$ on $K$ were studied in three bins of $|Y|$

359: ($0.0\leq|Y|<0.8$, $0.8\leq|Y|<1.6$, $1.6\leq|Y|<2.4$) and five bins of 

360: $\kt$ ($0.1\leq\kt<0.2$ GeV, $0.2\leq\kt<0.3$, $0.3\leq\kt<0.4$, 

361: $0.4\leq\kt<0.5$ and $0.5\leq\kt<0.6$ GeV).\par

362: 

363: Two-dimensional projections of the correlation function 

364: $C'(\Ql,\Qts,\Qto)$ for a single bin of $|Y|$ and $\kt$ are shown

365: in Fig.~\ref{proBP_2}a, b. BEC peaks are visible at low $\Ql,\Qts,\Qto$. 

366: 

367: \begin{figure}[!h]

368: \centerline{\includegraphics[width=10cm]{pr423_03.eps}}

369: \begin{quote}

370: \caption{\small Two-dimensional (a), (b) and one-dimensional (c), (d) and (e) 

371: projections of the correlation function $C'(\Ql,\Qts,\Qto)$ 

372: for 0.8 $\leq |Y| <$ 1.6 and 0.3 $\leq \kt <$ 0.4 GeV. $\Qto <$ 0.2 GeV in 

373: (a), $\Ql <$ 0.2 GeV in (b).  In (c), (d), (e) the projections at low values 

374: ($<$0.2 GeV) of the other variables.}

375: \label{proBP_2}

376: \end{quote}

377: \end{figure}

378: \par

379: 

380: To extract the spatial and temporal extensions of the pion source from

381: the experimental correlation functions,

382: the Bertsch-Pratt (BP)

383: \vspace{-0.2cm}

384: 

385: %\begin{flushleft}

386: %$C'(\Ql,\Qts,\Qto)$

387: %\end{flushleft}

388: %\begin{equation}

389: \begin{equation}

390: C'(\Ql,\Qts,\Qto) = N (1 + \lambda {\rm e}^{-(\Ql^2 \Rlong^2 +

391: \Qts^2 \Rts^2 +

392: \Qto^2 \Rto^2 +

393: 2 \Ql \Qto \Rlongto^2

394: )})

395: F(\Ql,\Qts,\Qto)

396: \end{equation}

397: and the Yano-Koonin (YK)

398: \begin{equation}

399: C'(\qt,\ql,\qz) =

400: N (1 + \lambda {\rm e}^{-(\qt^2 \Rt^2 +

401: \gamma^2(\ql - v \qz)^2 \Rl^2 +

402: \gamma^2(\qz - v \ql)^2 \Rz^2

403: )})

404: F(\qt,\ql,\qz)

405: \end{equation}

406: 

407: \noindent parameterizations were fitted to the measured correlation

408: functions in intervals of $\kt$ and $|Y|$. 

409: In both parameterizations, $N$ is a normalization factor, 

410: $\lambda$ is the degree of incoherence of the pion sources, related to the fraction of pairs that interfere.

411: The parameters $N$ and $\lambda$, whose product determines the size of the

412: BEC peak, are significantly (anti)correlated. The two functions 

413: $F(\Ql,\Qts,\Qto) =

414: (1 + \elong\Ql + \ets\Qts + \eto\Qto)$

415: and $F(\qt,\ql,\qz) = (1 + \dt\qt + \dl\ql + \dz\qz)$,

416: where $\epsilon_{\mathrm i}$ and $\delta_{\mathrm i}$ are free

417: parameters, were introduced in Eq.~(6) and (7) to take into

418: account residual long-range two-particle correlations due to energy and

419: charge conservation. The interpretation of the other parameters in Eq~(6), is:\\

420: - $\Rts$ and $\Rlong$ are the transverse and longitudinal 

421: radii in the longitudinal rest frame of the pair;\\

422: - $\Rto$ and the cross-term $\Rlongto$ are a

423: combination of both spatial and temporal extentions of the source.

424: The difference ($\Rto^2 - \Rts^2$) is proportional to the duration of

425: the particle emission process, and $\Rlongto$ to the source

426: velocity with respect to the pair rest frame.\par

427: 

428: In the YK frame Eq.~(7), where $\gamma=1/\sqrt{1-v^2}$, the free

429: parameters are interpreted as follows:\\

430: - $v$ is the longitudinal velocity, in units of $c$,

431: of the source element in the CMS frame;\\

432: - $\Rz$ measures the time interval, times $c$, during which

433: particles are emitted, in the rest frame of the emitter (source element).

434: The limited phase-space available limits the analysis for $\Rz^2$ ;\\

435: - $\Rt$ and $\Rl$ are the transverse and longitudinal radii, in the rest frame of the

436: emitter.\par

437: 

438: The parameters $\Rz$, $\Rt$ and $\Rl$ are

439: evaluated in the rest frame of the source element. 

440: The two parameterizations are not independent, so that a

441: comparison between the BP and YK fits is an important test.

442: In the YK picture the source velocity $v$ of each element does not depend on 

443: $k_t$, while it is correlated with the pair rapidity Y. $Y_{YK}$ measures the 

444: rapidity of the source element with respect to the c.m frame 

445: $Y_{YK} = \frac{1}{2} \ln \left[ (1+v) / (1-v) \right]$. 

446: A non expanding source corresponds to $Y_{YK} \simeq 0$ for any Y. For a 

447: longitudinally boost invariant source (for which the velocity of each element 

448: is $v = z/t$, where $t$ is the time elapsed since the collision and $z$ is the 

449: longitudinal coordinate of the element) the correlation $Y_{YK} =Y$ 

450: is expected as in Fig. 8. \par

451: 

452: The following relations hold between the BP and YK parameters:

453: \begin{eqnarray}

454: \Rts^2 & = & \Rt^2 \\

455: \Rlong^2 & = &

456: \gamma_{\mathrm {LCMS}}^2(\Rl^2+\beta_{\mathrm {LCMS}}^2 \Rz^2) \\

457: (\Rto^2-\Rts^2) & = & \beta_{\mathrm t}^2

458: \gamma_{\mathrm {LCMS}}^2(\Rz^2+\beta_{\mathrm {LCMS}}^2\Rl^2).

459: \end{eqnarray}

460: $\beta_{\mathrm {LCMS}}$ is the velocity of the

461: source element

462: in the LCMS, i.e. with respect to the pair longitudinal rest frame;

463: $\gamma_{\mathrm {LCMS}}=1/\sqrt{1-\beta_{\mathrm {LCMS}}^2}$. 

464: For a boost-invariant source, $\beta_{\mathrm {LCMS}}=0$: (9)

465: and (10) become:

466: \begin{eqnarray}

467: \Rlong^2 & \simeq & \Rl^2 \\

468: (\Rto^2-\Rts^2) & \simeq & \beta_{\mathrm t}^2 \Rz^2.

469: \end{eqnarray}

470: \par

471: In Fig. 10 the best-fit BP and YK parameters are compared: \\

472: - The longitudinal parameter $\Rlong^2$ is larger than

473: $\Rl^2$ in all rapidity intervals, Fig. 9(a),(d) and (g). $\Rlong^2 > \Rl^2$

474: corresponds to $\beta_{\mathrm {LCMS}}$ $>$0, i. e. to a pion source 

475: whose expansion is not exactly boost-invariant.\\

476: - The equality of the transverse parameters, $\Rts^2$ $\Rt^2$, is 

477: confirmed; there may be deviations at low $\kt$. \\

478: - $R^2_0$ and $(\Rto^2-\Rts^2)$ are essentially equal to zero, suggesting that the present technique does not allow to mesure the duration of the emission process.

479: 

480: 

481: \begin{figure}[!h]

482: \centerline{\includegraphics[width=14cm]{pr423_10.eps}}

483: \begin{quote}

484: \caption{\small BP and YK fits. (a)(d)(g) The best-fit longitudinal 

485: radius $\Rlong^2$ in the BP frame (open dots) compared 

486: with the 

487: YK $\Rl^2$ (full dots). (b)(e)(h) The BP transverse correlation length 

488: $\Rts^2$ (open dots) compared with the YK $\Rt^2$ (full dots). (c)(f)(i) The difference of the BP transverse radii $(\Rto^2-\Rts^2)$ (open dots) compared with the YK time parameter $\Rz^2$ times $\beta_{\mathrm t}^2$ (full dots). }

489: \label{all9par_bp}

490: \end{quote}

491: \end{figure}

492: \par

493: 

494: \section{Conclusions}

495: We have first summarized the results obtained on BECs in 

496: $e^+ e^-$ collisions at the $Z^0$ peak assuming a static source. Then we presented 

497: an analysis in bins of the average 4-momentum of the pair. Based on this, 

498: the dynamic features of the pion emitting source were investigated in the YK 

499: and BP formalisms.\par

500:  

501: The transverse and longitudinal radii of the pion sources

502: decrease for increasing

503: $\kt$, indicating the presence of correlations between the particle

504: production points and their momenta.

505: The YK rapidity scales with the pair rapidity,

506: in agreement with a nearly boost-invariant expansion of the pion source.

507: Phase space limitation did not allow the measurement of the

508: duration of the particle emission process.

509: \par

510: Similar results have been observed in more complex systems, such as the

511: pion sources created in pp and heavy-ion collisions, which are now

512: complemented with measurements in the simpler hadronic system formed

513: in $e^+ e^-$ annihilations. The unexplained similarities between BECs in 

514: different reactions might indicate a present limitation of our understanding of these correlations \cite{csorgo}.\\

515: 

516: {\bf Acknowledgements}. We thank all the members of the OPAL Collaboration, in particular the Bologna members. We acknowledge the contribution of Ms. Anastasia Casoni.

517: 

518: 

519: \begin{thebibliography}{99}

520: \footnotesize{

521: \bibitem{brown}

522: Hanbury-Brown et al., Phil. Mag. 54 (1954) 633; Nature 178 (1956) 1046.

523: \vspace{-0.3cm}

524: \bibitem{gold}

525: G. Goldhaber et al., Phys. Rev. 120 (1960) 130.

526: \vspace{-0.3cm}

527: \bibitem{cuffiani}

528: M. Cuffiani and G. Giacomelli, Il Nuovo Saggiatore 18 N1-2 (2002) 46.

529: \vspace{-0.3cm}

530: \bibitem{acton}

531: P. D. Acton et al., Phys. Lett. B267 (1991) 143.

532: \vspace{-0.3cm}

533: \bibitem{abbiendi}

534: G. Abbiendi et al., Phys. Lett. B559 (2003) 131. M. Boutemeur et al., 

535: hep-ex/0510027.

536: \vspace{-0.3cm}

537: \bibitem{kk}

538: P. D. Acton et al., Phys. Lett. 298 (1993) 456. 

539: G. Abbiendi et al., Eur. Phys. J. C21 (2001) 23. 

540: R. Akers et al., Z. Phys. C67 (1995) 389. 

541: G. Abbiendi et al., Eur. Phys. J. C8 (1999) 559.

542: \vspace{-0.3cm}

543: \bibitem{alexander}

544: G. Alexander et al., Phys. Lett. B452 (1999) 159.

545: \vspace{-0.3cm}

546: \bibitem{gg}

547: G. Alexander et al., Z. Phys. C72 (1996) 389. 

548: G. Giacomelli, Nucl. Phys. Proc. Suppl. 25B (1991) 30.

549: A. Breakstone et al., Phys. Lett. B162 (1985) 400.

550: \vspace{-0.3cm}

551: \bibitem{ackerstaff}

552: K. Ackerstaff et al., Eur. Phys. J. C5 (1998) 239. G. Abbiendi et al., Phys. Lett. B523 (2001) 35.

553: \vspace{-0.3cm}

554: \bibitem{multipions}

555: G. Abbiendi et al., Eur. Phys. J. C16 (2000) 423.

556: M. Acciarri et al., Phys. Lett. B458 (1999) 517. P. Abreu et al., Phys. Lett. 

557: B471 (2000) 460. D. Abbaneo et al. Eur. Phys. J. C36 (2004) 147.

558: \vspace{-0.3cm}

559: \bibitem{GA}

560: G. Abbiendi et al., Eur. Phys. J. C8 (1999) 59;  Eur. Phys. J. C36 (2004) 297.

561: \vspace{-0.3cm}

562: \bibitem{allasia}

563: D. Allasia et al., Z. Phys. C37 (1988) 527.

564: \vspace{-0.3cm}

565: \bibitem{momentum}

566: G. Abbiendi et al., arXiv:0708.1122 [hep-ex], Eur. Phys. J. C52 (2007) 787.

567: \vspace{-0.3cm}

568: \bibitem{ahmet}

569: K. Ahmet et al., Nucl. Instrum. Meth. A305 (1991) 275.

570: \vspace{-0.3cm}

571: \bibitem{aggraval}

572: M. M. Aggraval et al.[WA98 Coll] Phys. Rev. Lett. 93 (2004) 022301. 

573: S. Kniege et al. [NA49 Coll] J. Phys. G30 (2004) S 1073.

574: \vspace{-0.3cm}

575: \bibitem{adams}

576: J. Adams et al. [Star Coll] arXiv: nucl-ex/0411036. 

577: P. Chung et al. [Phenix Coll] Nucl. Phys. A749 (2005) 275.

578: \vspace{-0.3cm}

579: \bibitem{geiger}

580: K. Geiger et al., Phys. Rev. D61 (2000) 054002.

581: \vspace{-0.3cm}

582: \bibitem{csorgo}

583: T. Cs$\ddot{o}$rg$\ddot{o}$, J. of Phys. Conf. Series 50 (2006) 259.

584: 

585: }

586: 

587: 

588: 

589: 

590: \end{thebibliography}

591: 

592: \end{document}

593: 

594: 

595: 

596: 

597: 

598: 

599: 

600: 

601: 

602: 

603: