New particles from Belle Stephen L. Olsen Department of Physics & Astronomy,University of Hawai’i at Manoa, Honolulu, HI 96822, 5 USA 0 0 E-mail: [email protected] 2 n Abstract. I report recent results on hidden charm spectroscopy from Belle. These include: a observation of a near-threshold enhancement in the ωJ/ψ invariant mass distribution for J exclusive B → KωJ/ψ decays; evidence for the decay X(3872) → π+π−π0J/ψ, where the 7 π+π−π0invariantmassdistributionhasastrongpeakbetween750MeVandthekinematiclimit of 775 MeV, suggesting that the process is dominated by the sub-threshold decay X →ωJ/ψ; 2 andtheobservation ofapeaknear3940MeVintheJ/ψ recoilmassspectrumfortheinclusive v continuumprocesse+e− →J/ψX. Theresultsarebasedonastudyofa287fb−1 samplee+e− 8 annihilationdatacollectedatcenter-of-massenergiesaroundtheΥ(4S)intheBelledetectorat 6 the KEKBcollider. 0 2 1 4 . 0 / x e 1. Introduction - The recent surge in activity in hadron spectroscopy and, I suppose, the main motivation for p e the formation of the Topical Group on Hadron Physics, is the result of renewed interest in a h rather old question: are there hadronic states with a more complex structure than the simple qq¯ : v mesons and qqq baryons of the original quark model? This revival of interest has been driven i X by experimental reports of pentaquarks [1], the narrow DsJ states [2, 3], and the X(3872) [4]. In spite of considerable theoretical and experimental effort, the existence of non-qq¯ mesons r a and/ornon-qqq baryonsremainsanopenquestion. Whiletheidentification ofastrangeness=+1 (or charm=-1) baryon would be definitive evidence for a non-qqq baryon, the experimental situationregardingtheexistenceofsuchstatesremainsunsettled(andamajortopicofdiscussion at this meeting [5]). On the other hand, while the D and and X(3872) are experimentally sJ well established, the theoretical interpretation is not so clear. The D states could be standard sJ P-wave cs¯states and their narrowness is only surprisingbecause the relativistic potential model calculations that predicted them to be heavier (and above DK threshold) are wrong [6]. Some theorists, including our opening speaker [7], remain hopeful that a cc¯ charmonium assignment can be found for the X(3872). To sort this all out, I think that the so-called hidden charm mesons can and will play a decisive role for reasons that include: • the theory for these systems is well founded(and recently blessed by this year’s Nobel Prize Committee) and has fewest ambiguities; • the experimental signatures tend to be clean; • cc¯meson states below open-charm threshold are narrow and do not overlap; and • lotsofnon-cc¯-typemesonshavebeenconjectured,includingDD¯∗ molecules [8]andcc¯-gluon hybrids [9]. Although the Belle detector [10] is specialized to studies of CP violation in B meson decays, it has proven to be a useful device for studying particles containing cc¯ pairs. Belle detects cc¯ systems produced via weak decays of b quarks —the b → cc¯s process is a dominant b-quark decay mode— and the continuum production process e+e− → cc¯cc¯, which has been found to be surprisingly large. The KEKB asymmetric energy e+e− collider [11] operates at a center- of-mass (cms) energy corresponding to the Υ(4S) resonance and routinely delivers luminosities thatareinexcessof1034cm−2s−1,therebyprovidingBellewithahugedatasamplethatcontains about300 million BB¯ meson pairevents andover onebillion e+e− → qq¯continuum annihilation events. Belle results in the hidden charm meson sector include first observations of: ′ ′ ′ • the η via the sequence B → Kη , η → K Kπ [12]; c c c S • anomalously large cross sections for the exclusive process e+e− → J/ψη and the inclusive c process e+e− → J/ψ(cc¯) [13]; • the X(3872) meson [4]; • a near-threshold ωJ/ψ mass enhancement in exclusive B → KωJ/ψ decays [14]; and • a peak at 3940 MeV in the J/ψ recoil mass spectrum in the inclusive e+e− → J/ψX process [15]. In this talk I will discuss the last two items as well as recent results on properties of the X(3872). I will not have time to cover any of the many other Belle results on hadron ∗∗ spectroscopy, suchasourmanyinterestingresultsoncharmedbaryonspectroscopy[16], D [17] and D mesons [3] and two-photon physics [18]. In addition, I will not have time to report on sJ Belle’s lack of observation of pentaquarks [19] or the D (2632) [20]. All unpublished numbers sJ reported here are preliminary. 2. A near-threshold ωJ/ψ mass enhancement in B → KωJ/ψ decays At the Υ(4S), BB¯ meson pairs are produced with no accompanying particles. As a result, each B meson has a total cms energy that is equal to E , the cms beam energy. We beam identify B mesons using the beam-constrained B-meson mass M = E2 −p2 and the bc q beam B energy difference ∆E = E −E , where p is the vector sum of the cms momenta of the B beam B B meson decay products and E is their cms energy sum. For the final states discussed here, the B experimental resolutions for M and ∆E are approximately 3 MeV and 13 MeV, respectively. bc We select B → Kπ+π−π0J/ψ candidate events (J/ψ → ℓ+ℓ−) track combinations with M bc and ∆E values that are within 2.5σ of their nominal values. Figure 1 shows a scatterplot of M(π+π−π0J/ψ) (vertical) versus M(π+π−π0) for selected events in the ∆E-M signal bc region. Here a distinct vertical band corresponding to ω → π+π−π0 decays is evident near M(π+π−π0)= 0.782 GeV. We identify three-pion combinations with M(π+π−π0) within 25 MeV of m as ω candidates ω and form the Dalitz plot of M2(ωJ/ψ) (vertical) versus M2(ωK) (horizontal) shown in Fig. 2. The clustering of events near the left side of the plot corresponds to B → K J/ψ; K → Kω X X ∗ events, whereK denotes strange meson resonances such as K (1270), K (1400), and K (1430) X 1 1 2 that are known to decay to Kω. There is also a clustering of events with low ωJ/ψ invariant masses near the bottom of the Dalitz plot. To study these, we suppress K → Kω events by X only looking at events in the region M(Kω)> 1.6 GeV, to the right of the dashed line in Fig. 2. The M and ∆E distributions of the selected events indicate that about half of the entries bc in the M(ωK) > 1.6 GeV Dalitz plot region are due to background. To perform a background 4.8 24 4.55 21.5 yJ/) (GeV) 4.3 2yJ/) (GeV) 19 p3 M( 2wM( 4.05 16.5 3.8 0.50 0.75 1.00 1.25 1.50 14 M(3p ) (GeV) 1 2 3 4 5 M2(w K) (GeV2) Figure 1. A scatterplot of M(π+π−π0ℓ+ℓ−) (vertical) versus M(π+π−π0) for events in Figure 2. the Dalitz-plot distribution for B → KωJ/ψ candidate events. the ∆E-M signal region. The vertical band bc indicated by the arrows corresponds to ω → π+π−π0 decays. 12 n bi 8 nts/ Eve 4 30 0 n bi 8 nts/ Eve 4 20 0 n bi 8 nts/ Eve 4 10 0 n bi 8 nts/ Eve 4 0 50.200 M 5 .(2G50eV) 5.200 M 5 .(2G50eV) 5.200 M 5 .(2G50eV) 5.300 3880 M4(w08J0/y ) (MeV) 4280 bc bc bc Figure 4. B → KωJ/ψ signal yields vs Figure 3. M distributions for events in bc M(ωJ/ψ). The curve in (a) indicates the the ∆E signal region for 40 MeV-wide bins in result of a fit that uses an S-wave Breit- M(ωJ/ψ). Wignerresonancetermandaphase-space-like threshold function for the background. subtraction and determine the level of B → KωJ/ψ signal events, we separate the data into 40 MeV-wide bins of M(ωJ/ψ) and measure the B meson signal levels in the M and ∆E bc distributions. The histograms in Fig. 3 show the M distributions for the twelve lowest bc M(ωJ/ψ) mass bins, where strong peaks at M = m are evident at low ωJ/ψ masses, bc B especiallyforthemassregionscoveredbyFigs.3(b)and(c). Thecorresponding∆E distributions (notshown)showsimilarstructure. WeestablishtheB → KωJ/ψsignallevelforeachM(ωJ/ψ) mass bin by performing binned fits simultaneously to the M and ∆E distributions with bc Gaussian functions for the signal and smooth background functions. The smooth curves in Fig. 3 indicate the fitted M curves for each ωJ/ψ mass bin. bc The bin-by-bin signal yields are plotted vs M(ωJ/ψ) in Fig. 4. An enhancement is evident around M(ωJ/ψ) = 3940 MeV. The curve in Fig. 4 is the result of a fit with a S-wave ∗ ∗ Breit Wigner function threshold function of the form f(M) = A q (M), where q (M) is the 0 momentum of the daughter particles in the ωJ/ψ restframe. This functional form accurately reproduces the threshold behavior of Monte Carlo simulated B → KωJ/ψ events that are generated uniformly distributed over phase-space. The fit gives a Breit-Wigner signal yield of 58±11 events with a peak peak mass and total width of M = 3943±11(stat)±13(syst) MeV Γ = 87±22(stat)±26(syst) MeV, where the systematic errors are determined from variations in the values when different bin sizes, fitting shapes and selection criteria are used. The event yield translates into a product branching fraction (here we denote the enhancement as Y(3940)): B(B → KY(3940))B(Y(3940) → ωJ/ψ) = (7.1±1.3(stat)±3.1(syst))×10−5, The statistical significance of the signal, determined from −2ln(L /L ), where L and 0 max max p L are the likelihood values for the best-fit and for zero-signal-yield, respectively, is 8.1σ. 0 A cc¯charmonium meson a mass of 3943 MeV would dominantly decay to DD¯ and/or DD¯∗; hadroniccharmoniumtransitionsshouldhaveminusculebranchingfractions. Ontheotherhand, decays of cc¯-gluon hybridcharmoniumto D(∗)D¯(∗) meson pairsareforbiddenor suppressed,and ∗∗ the relevant “open charm” threshold is mD +mD∗∗ ≃ 4285 MeV [21, 22], where D refers to the JP = (0,1,2)+ charmed mesons. Thus, a hybrid state with a mass equal to that of the peak ′ we observe would have large branching fractions for decays to J/ψ or ψ plus light hadrons [23]. Moreover, lattice QCD calculations have indicated that partial widths for such decays can be comparable to the width that we measure [24]. However, these calculations predict masses for these states that are between 4300 and 4500 MeV [25], substantially higher than our measured value. 3. The X(3872) with 253 fb−1 The X(3872) was discovered by Belle as a narrow π+π−J/ψ mass peak in exclusive B− → K−π+π−J/ψ decays [4, 26]. Figure 5 shows the X(3872) signal from a 253 fb−1 data sample containing 275 million BB¯ pairs. The observed mass and the narrow width are not compatible with expectations for any of the as-yet unobserved charmonium states [27]. Moreover, the π+π− invariant mass distribution, shown in Fig. 6, peaks near the upper kinematic limit of M(π+π−) = 775 MeV, and has a shape that is consistent with ρ→ π+π− decays. Charmonium decays to ρJ/ψ final states violate isospin and are expected to be suppressed. The X(3872) and its above-listed properties were confirmed by the BaBar [28], CDF [29] and D0 [30] experiments. The X(3872) mass (3871.9 ± 0.5 MeV [31]) is within errors of the D0D¯0∗ threshold (3871.3±1.0 MeV [32]); the difference is 0.6±1.1 MeV. This has led to speculation that the X 12 10 30 8 V e M 0 2eV/c20 nts/1 6 M e nts/5 ev 4 e v E 10 2 0 0 0.40 0.50 0.60 0.70 0.80 M(p +p -) (GeV) 3820 3840 3860 3880 3900 M(p +p -J/y ) (MeV/c2) Figure 6. M(π+π−) for events in the Figure 5. The X(3872) → π+π−J/ψ signal X(3872) signal peak. The shaded histogram from the 253 fb−1 data sample. icsurtvheeissidthebearnedsu-dltetoefrmaifintedwbitahckagrρo→undπ;+tπh−e lineshape. mightbeaD0D¯0∗ boundstate[33,34,8]. Accordingtoref.[33],thepreferredquantumnumbers for such a bound state would be either JPC = 0−+ or 1++. The decay of an C = +1 state to π+π−J/ψ would proceed via an I = 1 ρ0J/ψ intermediate state and produce the π+π− mass spectrum like that we see. In this meson-meson bound state interpretation, the close proximity of the X mass to D0D¯0∗ threshold compared to the D+D−∗-D0D¯0∗ mass splitting of 8.1 MeV produces a strong isospin violation. Swanson made a dynamical model for the X(3872) as a D0D¯0∗ hadronic resonance [34]. In this model, JPC = 1++ is strongly favored and the wave function has, in addition to D0D¯0∗, an appreciableadmixtureofωJ/ψ plusasmallρJ/ψ component. Thelatterproducestheπ+π−J/ψ decays that have been observed; the former gives rise to π+π−π0J/ψ decays via a virtual ω that are enhanced because of the large ωJ/ψ component to the wavefunction. Swanson’s model predicts that X(3872) → π+π−π0J/ψ decays should occur at about half the rate for π+π−J/ψ and with a π+π−π0 invariant mass spectrum that peaks near the upper kinematic boundary of 775 MeV (7.5 MeV below the ω peak). X(3872) → π+π−π0J/ψ decays would populate the horizontal band indicated by the horizontallinesinthescatterplotofFig.1. Thiscorrespondstothe±3σband|M(π+π−π0J/ψ)− m |< 16.5 MeV. X(3872) Figure 7 shows the M distributions for events that are in the ∆E and X → π+π−π0J/ψ bc signal regions for 25 MeV-wide π+π−π0 invariant mass bins; Fig. 8 shows the corresponding∆E distributions for events in the M and X signal regions. There are distinct B meson signals in bc both the M and ∆E distributions for the M(π+π−π0)> 750 MeV bin and no evident signals bc for any of the other 3π mass bins. The curves in the figures are the results of binned likelihood fits that are applied simultaneously to the M and ∆E distributions. bc Figure 9 shows the fitted B-meson signal yields vs M(π+π−π0). All of the fitted yields are consistent with zero except for the M(π+π−π0) > 750 MeV bin, where the fit gives 5 6 bin bin 4 s/2.5 s/ Evt Evt 2 0 0 bin bin 4 s/2.5 s/ Evt Evt 2 0 0 bin bin 4 s/2.5 s/ Evt Evt 2 0 0 bin bin 4 s/2.5 s/ Evt Evt 2 50.200 5.250 5.200 5.250 5.200 5.250 5.300 -00.20 0.00 -0.20 0.00 -0.20 0.00 0.20 M (GeV) M (GeV) M (GeV) D E (GeV) D E (GeV) D E (GeV) bc bc bc Figure 7. M distributions for 25 MeV- Figure8. ∆Edistributionsfor25MeV-wide bc wide π+π−π0 invariant mass bins. π+π−π0 invariant mass bins. 16 3 V e M n vents/25 8 Events/bi 12 E 0 0 1.60 1.70 1.80 1.90 2.00 M(Kp +p -p 0) (GeV) 480 605 730 M(p +p -p 0) (MeV) Figure 10. The M(Kπ+π−π0) distribution Figure 9. The B-meson signal yields from for events in the M -M(3π) signal region. X thefitstotheM -∆E signals vs3π invariant bc mass. 12.4 ± 4.1 events. The statistical significance of the signal in this one bin, determined from −2ln(L /L ), where L and L are the likelihood values for the best-fit and for zero- 0 max max 0 p signal-yield, respectively, is 6.6σ. Figure 10 shows the M(Kπ+π−π0) distribution for for the X → π+π−π0J/ψ signal events. The distribution is spread across the limited allowed kinematic region and there is no evident structure that might be producing the high mass peak in Fig. 9 by some sort of a kinematic reflection. A possible background to the observed signal would be feed-across from the near-threshold ωJ/ψ enhancement in B → KωJ/ψ decays described above. Since the ω → π+π−π0 resonance peak is at m = 782.5 MeV, which is 7.5 MeV above the maximum possible 3π invariant mass ω value for X → π+π−π0J/ψ decays, there is no overlap between the centroids of the ωJ/ψ and X → π+π−π0J/ψ signal bands in Fig. 1. However, there is some overlap in the tails of the kinematically allowed regions for the two processes that might result in some events from one signal feeding into the other. We determine the level of signal cross-talk to be 0.75 ± 0.14 events from the integral of the fitted function in Fig. 4 over the region of overlap with the X(3872) signal band. As an independent check, we refitted for the X(3872) → π+π−π0J/ψ signal yield with a tighter restriction on M(π+π−π0J/ψ), namely m − 3σ < M(π+π−π0J/ψ) < m + 1σ, that has X X no overlap with the ω band. The X → π+π−π0J/ψ signal yield in the truncated region is 10.6±3.6 events. For a Gaussian signal distribution with no feed-across background, we expect the truncation of the signal region to reduce the signal by 2.1 events (16%); the observed reduction of 1.8 events is consistent with a feed-across level that is less than one event. AnotherpossiblesourceofbackgroundtotheX(3872) → π+π−π0J/ψsignalarenon-resonant B− → K−π+π−π0J/ψ decays. To determinethelevel of these, welooked for B-mesonsignals in theM -∆E distributionsforeventsinM(π+π−π0J/ψ)sidebandsaboveandbelowtheX(3872) bc mass region. There is no evidence for significant signal yields in the M -∆E distributions of bc either sideband. Fits gives an estimate of thenon-resonant background in theX → π+π−π0J/ψ signal bin of 1.3±1.0 events. Todeterminethebranchingfraction, weattribute all of thesignal events with M(π+π−π0)> 750 MeV to X → π+π−π0J/ψ decay. We compute the ratio of π+π−π0J/ψ and π+π−J/ψ branchingfractionsbycomparingthistothenumberofX → π+π−J/ψ inthesamedatasample, corrected by MC-determined relative detection efficiencies. The ratio of branching fractions is B(X → π+π−π0J/ψ) Nev(π+π−π0J/ψ)επ+π−J/ψ = = 1.1±0.4(stat)±0.3(syst), (1) B(X → π+π−J/ψ) Nev(π+π−J/ψ)επ+π−π0J/ψ wherethesystematicerrorreflectstheuncertaintyintherelativeacceptance, thelevelofpossible feed-across and nonresonant background, and possible event loss due to the M(3π) > 750 MeV requirement, all added in quadrature. If we allow for cross-talk and non-resonant contributions attheirmaximum(+1σ)values,thestatistical significanceoftheX(3872) → π+π−π0J/ψ signal is reduced to ≃ 4σ. 4. A new charmonium state in inclusive e+e− → J/ψX annihilations. SomeofthebiggestsurprisesfromBellehavenothingtodowithB-mesonphysicsatallandhave come, instead, from the inclusive e+e− → J/ψX annihilation process. This is demonstrated in Fig. 11, which shows the distribution of masses for systems with more than two charged tracks thatrecoil against J/ψ mesonsproducedinthee+e− continuumator neartheΥ(4S)resonance. Inthis figure, which is based on a 280 fb−1 data sample, the histogram indicates thebackground level derived from the J/ψ → ℓ+ℓ− mass sidebands. The prominent peak at M ≃ 2.98 GeV in Fig. 11 corresponds to the η . From the yield recoil c of events we determine a cross-section branching-fraction product [15] σ (e+e− → J/ψη )B(η →> 2tracks) = 25.6±4.4 fb. Born c c This is more than an order of magnitude higher than non-relativistic QCD (NRQCD) calculations of ∼ 2 fb−1[35]. There is no evident signal for any recoils with mass below M , ηc which is also the cc¯mass threshold. Also contrary to NRQCD expectations, the four-charmed- quark process e+e− → cc¯cc¯ dominates inclusive J/ψ production. From the total number of charmonium states and charmed particles found in the recoil system, we determine the cross section ratio [36] σ(e+e− → J/ψ(cc¯) = 0.82±0.21; σ(e+e− → J/ψX) NRQCD predicts this ratio to be ∼ 0.1 [37]. ′ The second and third prominent peaks in Fig. 11 are at the masses of the χ and η , c0 c respectively. The fourth peak is well fitted by a Gaussian function with a peak mass of 3940 ± 12 MeV and a signal significance of 4.5σ. The width of this state is consistent with 2c 2MeV/c 80 DD thresholdX(3938) N/20 MeV/ 23 0 60 unresolved puzzle of J/y X 2 N/ 1 40 glueball or 0 3.3s fluctuation 3 20 2 0 2 2.5 3 3.5 4 4.5 1 Recoil Mass(J/y ) GeV/c2 0 3.5 4 4.5 Recoil Mass(J/y ) GeV/c2 Figure 11. The distribution of masses recoiling from the J/ψ in e+e− → J/ψX annihilations. Figure 12. The DD¯∗ (top) and DD¯ (bottom) invariant mass distributions for e+e− → J/ψDD¯(∗). experimental M resolution. Since this is rather poor, we can only derive an upper limit on recoil the total width of Γ< 96 MeV (90% CL). We investigated this peak further by studying events where a D meson is identified in the J/ψ recoil system, e.g. in events of the type e+e− → J/ψDX. Figure 12 shows the distribution of masses recoiling against the J/ψ for e+e− → J/ψDX events where MX = mD∗ (top) and M = m (bottom). There is an evident 9.9±3.3 event signal for the 3940 MeV state in the X D DD¯∗ mass spectrum, with a statistical significance of 4.5σ. The signal level is the DD¯ mass spectrum is 4.1±2.2 events with a significance of only 2.1σ. This peak cannot be identified with any known charmonium state. An obvious guess is that it is either the χ′ or the η′′. However, χ′ → DD¯∗ is forbidden and, thus, ruled out. Likewise c0 c c0 η′′ → DD¯ decays are also forbidden, but, since the DD¯ “signal” is ambiguous, we can’t use this c ′′ to rule out this assignment. On the other hand, an η assignment to the observed peak would c imply a m −m mass splitting of ∼ 100 MeV, about twice as large as the measured ψ(3S) ηc(3s) splitting for the 2S states. This seems unlikely. The mass of this fourth peak is very similar to that of the ωJ/ψ peak seen in B → KωJ/ψ and described above, and a search for it in the ωJ/ψ decay channel is in progress. In addition, we are examining B → KDD¯∗ decays for a DD¯∗ component of the ωJ/ψ enhancement. 5. Summary We observe peaks near 3940 MeV in the ωJ/ψ mass distribution from B → KωJ/ψ decays and in the recoil mass spectrum in the inclusive annihilation process e+e− → J/ψX. The latter peak is also seen in the exclusive process e+e− → J/ψDD¯∗ and, thus, cannot be assigned to the χ charmonium state. At this stage, we cannot tell whether or not the state seen in B decays c0 and the one seen in inclusive J/ψ production are one and the same. Further investigation is in progress. We observe a ∼ 4σ signal for X(3872) → π+π−π0J/ψ. This is the first measurement of an X decay mode other than π+π−J/ψ. The π+π−π0 invariant masses are strongly clustered above 750 MeV, near the upper kinematic boundary; this is suggestive of a sub-threshold decay via a virtual ωJ/ψ intermediate state. Such a decay, at near the measured branching fraction, was predicted by Swanson based on a model where the X(3872) is considered to be primarily a D0D¯0∗ hadronic resonance [34]. Thepresence of the X(3872) → ωJ/ψ decay process would establish the Charge-Conjugation quantum number of the X(3872) as C = +1. This, in turn, would mean that the π+π− system in X → π+π−J/ψ decay comes from the decay of a ρ meson. The large isospin violation implied by the near equality of the ρJ/ψ and ωJ/ψ decay widths is difficult to accomodate in a cc¯ charmonium interpretation of the X, but a natural consequence of the meson-meson bound state model point of view. 6. Acknowledgements I thank the organizers of the APS Topical Group on Hadron Physics for inviting me to give this talk. My Belle colleagues and I thank the KEKB group for the excellent operation of the accelerator, the KEK cryogenics group for the efficient operation of the solenoid, and the KEK computer group and the NII for valuable computing and Super-SINET network support. We acknowledge support from MEXT and JSPS (Japan); ARC and DEST (Australia); NSFC (contract No. 10175071, China); DST (India); the BK21 program of MOEHRD and the CHEP SRC program of KOSEF (Korea); KBN (contract No. 2P03B 01324, Poland); MIST (Russia); MESS (Slovenia); Swiss NSF; NSC and MOE (Taiwan); and DOE (USA). References [1] T. Nakano et al. (LEPS Collaboration), Phys. Rev. Lett. 91, 012002 (2003). See also K. Hicks, these proceedings. [2] B. Aubert et al. (BaBar Collaboration), Phys. Rev. Lett. 90, 242001 (2003) and D. Besson et al. (CLEO Collaboration), Phys. Rev.D 68, 032002 (2003). [3] Y. Mikami et al. (Belle Collaboration), Phys. Rev. Lett. 92, 012002 (2003) and P. Krokovny et al. (Belle Collaboration), Phys. Rev. Lett. 91, 262002 (2003). For recent Belle results see A. Drutskoy, these proceedings. [4] S.K.Choi et al. (Belle Collaboration), Phys. Rev.Lett. 91, 262001 (2003). [5] A.Dzierba, these proceedings; see also A.R. Dzierba et al. Phys.Rev.D 69, 051901 (20034). [6] E. Eichten, these proceedings; see also W. Bardeen, E.J. Eichten and C. Hill, Phys. Rev. D 68, 054024 (2003). [7] C. Quigg, these proceedings; see also T. Barnes and S. Godfrey, Phys. Rev. D 69, 054008 (2004) and E.J. Eichten,K. Lane and C. Quigg, Phys.Rev. D 69, 094019 (2004). [8] See, for example, M.B. Voloshin and L.B. Okun, JETP Lett. 23, 333 (1976); M. Bander, G.L. Shaw and P.Thomas, Phys.Rev.Lett.36,695 (1977); A.DeRujula,H.Georgi andS.L.Glashow, Phys.Rev.Lett. 38, 317 (1977); N.A. T¨ornqvist, Z. Phys. C 61, 525 (1994); and A.V. Manohar and M.B. Wise, Nucl. Phys.B 339, 17 (1993). [9] D.Horn and J. Mandula, Phys.Rev.D 17, 898 (1978). [10] A. Abashian et al. (Belle Collab.), Nucl. Instr. and Meth. A 479, 117 (2002) and Y. Ushiroda (Belle SVD2 Group),Nucl. Instr. and Meth.A 511, 6 (2003). [11] S. Kurokawa and E. Kikutani, Nucl. Instr. and. Meth. A499, 1 (2003), and other papers included in this volume. [12] S.K.Choi et al. (Belle Collaboration), Phys. Rev.Lett. 89, 102001 (2002). [13] K.Abeet al.(Belle Collaboration), Phys.Rev. Lett. 89, 142001 (2002). [14] S.K.Choi et al. (Belle Collaboration), hep-ex/0408126. [15] P.Pakhlov, hep-ex/0412041; see also T. Ziegler, these proceedings. [16] T.Lesiaketal.(BelleCollaboration),hep-ex/0409065,R.Misuketal.(BelleCollaboration),hep-ex/0412069, and K. Abeet al. (Belle Collaboration), Phys. Lett. B 524, 33 (2002). [17] K. Abe et al. (Belle Collaboration), hep-ex/0410091 and K. Abe et al. (Belle Collaboration), Phys. Rev. D 69, 112002 (2004). [18] H. Nakazawa et al. (Belle Collaboration), hep-ex/0412058; K. Abe et al. (Belle Collaboration), Eur. Phys. J. C 32, 323 (2003); and K.Abeet al. (Belle Collaboration), Phys. Lett.B 540, 33 (2002). [19] R.Mizuk, talk at PENTA04, Kobe Japan, July 2004, hep-ex/0411005. [20] K.Abe, talk at PENTA04, KobeJapan, July 2004. [21] N.Isgur, R. Kokoskiand J. Paton Phys.Rev. Lett. 54, 869 (1985). [22] F.E.Close andP.R. Page, Nucl.Phys B443,233(1995); F.E.Close andP.R. Page, Phys.Lett. B366,323 (1996). [23] F.E. Close, Phys.Lett. B 342, 369 (1995). [24] C. McNeile, C. Michael and P.Pennanen (UKQCD Collaboration), Phys. Rev.D 65, 094505 (2002). [25] C. Banner et al., Phys. Rev. D 56, 7039 (1997); Z.-H. Mei and X.-Q. Luo, Int. J. Mod. Phys. A 18, 5713 (2003); and X. Liao and T. Manke, hep-lat/0210030. [26] Theinclusion of charge-conjugate modes is always implied. [27] S.L.Olsen, hep-ex/0407033 and K. Abe et al. (Belle Collaboration), hep-ex/0408116. [28] B. Aubert et al. (BaBar Collaboration), hep-ex/0406022. [29] D.Acosta et al. (CDF-IICollaboration), Phys. Rev.Lett.93, 072001 (2004). [30] V.M. Abazov et al.,(D0 Collaboration), hep-ex/0405004. [31] This is theweighted average of all of the reported X(3872) mass measurements. [32] S.Eidelman et al. (Particle Data Group), Phys. Lett. B 592, 1 (2004). [33] N.A. T¨ornqvist, Phys. Lett. B 590, 209 (2004); F.E. Close and P.R. Page, Phys. Lett. B 578, 119 (2003); S. Pakvasa and M. Suzuki, Phys. Lett. B 579, 67 (2004); C.-Y. Wong, Phys. Rev. C 69, 055202 (2004); and E. Braaten and M. Kusunoki,Phys. Rev.D 69, 114012 (2004). [34] E.S. Swanson, Phys. Lett.B 588, 189 (2004); [35] G.T. Bodwin, J. Lee and E. Braaten, hep-ph/0212181. [36] K.Abeet al.(Belle Collaboration), BELLE-CONF-0331 (2003). [37] P. Cho and A.K. Leibovich, Phys. Rev. D 54, 6690 (1996); E. Braaten and Yu-Qi Chen, Phys. Rev. Lett. 76, 730 (1996); V.V.Kiselev, A.K.Likhoded and M.V. Shevlyagin Phys. Lett. B 332, 411 (1994).