In a recent article we presented an argument which, we believe, shows to be incorrect an estimate, by Boglione and Pennington, of corrections to the valence (quenched) approximation predictions for properties of the lightest scalar glueball. Boglione and Pennington's reply to our article, it appears to us, fails to address the specific technical issues we raised.

The fragmentation of gluons to gluonium states is analyzed qualitatively in the non-perturbative region. The convolution of this mechanism with perturbative gluon radiation leaves us with a hard component in the fragmentation of gluon to gluonium.

The constituent nature of candidate gluonic mesons can be studied by comparing their production in quark and gluon jets. The production rate for such mesons depends on the colour confinement processes at the end of the perturbative evolution. Whereas we expect enhanced production of hybrids in the fragmentation region of a gluon jet, the rate for glueballs depends on the relative importance of colour triplet and colour octet neutralization. These neutralization processes can be studied independently in events with large rapidity gaps. If octet processes turn out important the recently suggested lightest J^{PC}=0^{++} glueball with mass around 1000 MeV should become visible already in the spectrum of the leading charged particle pairs.

We explain the phi and t dependences of mesons with JPC = 0pm +,1^++,2pm +$ produced in the central region of pp collisions. For the 0++ and 2++ sector this reveals a systematic behaviour in the data that appears to distinguish between qqbar and non-qqbar or glueball candidates.

Based on a $3\times3$ mass matrix describing the mixing of the scalar states $f_0(1370)$, $f_0(1500)$ and $f_0(1710)$, the hadronic decays of the three states are investigated. Taking into account the two possible assumptions concerning the mass level order of the bare states N, S and G in the scalar sector, $M_G > M_S > M_N$ and $M_G > M_N > M_S$, the glueball-quarkonia content of the three states is obtained by solving the unlinear equations. Some predictions about the hadronic and two-photon decays of the three states in two cases are presented. It is pointed out that the predictions about the two-photon decay width ratio for the three states can provide a stringent consistency check of the two assumptions.

We examine the possibility that the xi(2230) meson is a member of the Pomeron trajectory. A method of connecting the xi --> p pbar decay width and the pp cross sections through the Pomeron residue function is presented. We have used a relativistic, singularity-free form factor to make the analytic continuation of the residue function between crossed channels. We predict that if the xi(2230) meson is a Pomeron, then it should have a xi --> p pbar decay width of about 2 MeV.

We study the mixing angle \theta_{O\psi} and mixing amplitude f_{O\psi} of J/psi and vector glueball O, in the framework of potential models of heavy quarks and constituent gluons. While the state vectors of J/psi and O are constructed from the wave functions of few-body Schroedinger equations, the mixing dynamics is governed by perturbative QCD. We obtain a value of the mixing angle of |\tan \theta_{O\psi}| \approx 0.015 and the mixing amplitude of |f_{O\psi}(m_{O\psi}^2)| \approx 0.008 GeV^2, which is compatible with phenomenological analysis.

Lee and Weingarten have recently criticized our calculation of quarkonium and glueball scalars as being "incomplete" and "incorrect". Here we explain the relation of our calculations to full QCD.

We evaluate the continuum limit of the valence (quenched) approximation to the mass of the lightest scalar quarkonium state for a range of different quark masses and to the mixing energy between these states and the lightest scalar glueball. Our results support the interpretation of $f_0(1710)$ as composed mainly of the lightest scalar glueball.

The present discussion focusses on the dominant production of the 0++ glueball in central collisions in the suppressed yet observable channel 0++ --> pi+ pi- l+ l-.

We use Ward identities of broken scale invariance to infer the amount of scalar glueball--$\bar{q}q$ meson mixing from the ratio of quark and gluon condensates in the QCD vacuum. Assuming dominance by a single scalar state, as suggested by a phase-shift analysis, we find a mixing angle $\gamma \sim 36^{\circ}$, corresponding to near-maximal mixing of the glueball and $\bar{s}s$ components.

Recent results concerning the evidence and classification of the J^{PC}=0^{++} states, obtained with P. Minkowski, are presented: The isoscalars f_0(980) and f_0(1500) are classified as members of the 0^{++} nonet, while the broad state called f_0(400-1200) and the state f_0(1370) are considered as different components of a single broad resonance, the lowest-lying 0^{++} glueball. Furthermore, we propose the investigation of glueball production in the fragmentation region of gluon jets.

A study of central meson production as a function of the difference in transverse momentum (dPT) of the exchanged particles shows that undisputed qqbar mesons are suppressed at small dPT whereas the glueball candidates are enhanced and that the production cross section for different resonances depends strongly on the azimuthal angle between the two outgoing protons.

We elaborate on the hypothesis that in high energy hadron hadron and nucleus nucleus collisions the lowest mass glueballs are copiously produced from the gluon rich environment especially at high energy density. We discuss the particular glueball decay modes: $0^{++}, 2^{++} \to K \bar{K}$ and $0^{++} \to \pi^{+} \pi^{-} \ell^{+} \ell^{-}$.

The method of equivalent quanta is applied both to photon-photon and, by analogy, to double pomeron exchange in heavy-ion collisions. This Weizs\"acker-Williams approach is used to calculate production cross sections for the glueball candidate $f_J(1710)$ meson via photon-photon and pomeron-pomeron fusion in peripheral heavy-ion collisions at both RHIC and LHC energies. The impact-parameter dependence for total and elastic cross sections are presented, and are compared to results for proton-proton collisions.

The evidence for the low mass J^{PC}=0^{++} states is reconsidered. We suggest classifying the isoscalars f_0(980) and f_0(1500) as members of the 0^{++} nonet, with a mixing rather similar to that of the pseudoscalars \eta' and \eta. The broad state called f_0(400-1200) or `sigma'' and the state f_0(1370) are considered as different signals from a single broad resonance, which we take to be the lowest-lying 0^{++} glueball. The main arguments in favor of these hypotheses are presented and compared with theoretical expectations.

An unambiguous identification of glueballs, quarkless hadrons dramatically predicted by non-perturbative quantum chromodynamics, will be of great significance. This has so far been problematic owing to the scarcity of glue-rich production channels as well as possible mixing between glueballs and flavour singlet quark-antiquark mesons. The proposal, advanced here, to experimentally search for glueballs in the invariant mass of a leading K_S-pair fragmenting from an energetic jet out of high-statistics three-jet events in hadronic decays of the weak neutral Z boson, is relatively free from this problem. A hard gluon can perturbatively go into a quark-antiquark pair, which then hadronizes, but it will not fragment directly into a leading q q(bar)meson. Yet, a sufficiently hard gluon jet will have a substantial fragmentation rate into a leading glueball. For a mixed glueball-meson hadron, this rate will only be reduced by the cosine squared of the mixing angle. Production and detection rates of a spin-zero glueball are estimated here within reasonable fragmentation models using realistic cuts and efficiencies. These numbers make it very likely that a search, along the lines suggested here by any of the four groups at the Large Electron Positron collider at CERN, will prove fruitful.

We demonstrate that the azimuthal dependence of central meson production in hadronic collisions, when suitably binned, provides unambiguous tests of whether the Pomeron couples like a conserved vector-current to protons. We discuss the possibility of discriminating between q-qbar and glueball production in such processes. Our predictions apply also to meson production in tagged two-photon events at electron--positron colliders and to vector-meson production in ep collisions at HERA.

Based on the general analysis of branching ratio of two pseudoscalar meson channels, discriminants between mesons and glueballs for $I=0, J^{PC}=even^{++}$ unflavored hadrons with mass between 1.2 GeV and 2.9 GeV are suggested. Known $I=0, J^{PC}=2^{++}, f_2(1525)$ particle is discriminated as a typical meson. The way to discriminate new $I=0, J^{PC}=even^{++}$ unflavored hadrons is discussed.

It is shown that instantons provide a natural mechanism to explain an unusual azimuthal angular dependence of the production of the even-parity glueball candidates in central pp collision. A different azimuthal angular dependence for instanton-induced production of the odd-parity glueballs is predicted.

We discuss the theoretical expectations and phenomenological evidence for the lightest glueballs and the members of the meson nonet with quantum numbers J^PC=0^{++}. We reconsider the recent evidence for candidate states with masses below ~1700 MeV, but include also the results from earlier phase-shift analyses. Arguments are presented to classify the scalars f_0(980) and f_0(1500) as members of the 0^{++} nonet, with a mixing rather similar to that of the pseudoscalars eta' and eta. The S-wave states called f_0(400-1200) and f_0(1370) are considered as different signals from a single broad resonance, which we take to be the lowest-lying 0^{++} glueball. This state together with eta(1440) and f_J(1710) with spin J=2 form the basic triplet of binary gluonic bound states. We argue that these hypotheses are consistent with what can be expected theoretically.

The current status of the accepted glueball candidates is reviewed. The difference between bare and dressed hadrons is emphasised. What two photon processes, both production and decay, have taught us about these is discussed.

Within the meson-exchange model we calculate $f_0$-meson production cross section in $\pi N$ and $NN$ reactions and investigate the possibility for $f_0$ observation via the $K\bar{K}$ decay mode in $pp$ collisions. Our studies indicate that an extraction of the $f_0$ signal is unlikely due to the large background from nonresonant reaction channels.

We investigate the two-body decay modes $\rho\rho, \pi\pi^{\ast}(1300)$ and $\sigma \sigma$ of the $f_0(1500)$, all leading to the $4\pi$ decay channel, in a three-state mixing scheme, where the $f_0(1500)$ is a mixture of the lowest-lying scalar glueball with the nearby isoscalar states of the $0^{++} Q\bar{Q}$ nonet. In the leading order of this scheme, the decay mechanism of the $f_0(1500)$ proceeds dominantly via its quarkonia components, which can be described in the framework of the $^3P_0$ $Q\bar Q$ pair creation model. We predict the hierarchy of decay branching ratios B with $B(\rho \rho) ^{>}_{\sim} B(\pi\pi) ^{>}_{\sim} B(\sigma \sigma) > B(\pi\pi^{\ast})$,providing a key signature for the proposed mixing scheme in this leading order approach.

We provide the first explanation of the counter-intuitive scalar glueball couplings to pseudoscalar mesons found in lattice QCD and predict hitherto uncalculated decay modes. Significant a_1 pi and (pi pi)_S (pi pi)_S couplings are found. We demonstrate the equivalence of linear and quadratic mass matrices for glueball-quarkonium mixing. The equivalence of formalisms which deal with a glueball-quarkonium basis and only a quarkonium basis is demonstrated. We show that the f_0(1500) is not the heaviest state arising from glueball-quarkonium mixing for a glueball mass consistent with lattice QCD. The masses and couplings of scalar mesons, as well as their valence content, are calculated.

We examine the mixing between the sigma meson $\sigma$ and the "pure" glueball field $H$ and study the decays of the scalar glueball candidate $f_0(1500)$ (a linear combination of $\sigma$ and $H$) into two and four pions in effective linear sigma model. From recent experimental data on $f_0(1500)$ decay to pions we point out that the mixing angle is of order of 0.1.

We show that linear Regge trajectories for mesons and glueballs, and the cubic mass spectrum associated with them, determine a relation between the masses of the \rho meson and the scalar glueball, M(0^{++})=3/\sqrt{2} M(\rho ), which implies M(0^{++})=1620\pm 10 MeV. We also discuss relations between the masses of the scalar and tensor and 3^{--} glueballs, M(2^{++})=\sqrt{2} M(0^{++}), M(3^{--})=2M(0^{++}), which imply M(2^{++})=2290\pm 15 MeV, M(3^{++})=3240\pm 20 MeV.

Using the ``glueball dominance'' picture of the mixing between q\bar{q} mesons of different hidden flavors, we establish new glueball-meson mass relations which serve as a basis for glueball spectral systematics. For the tensor glueball mass 2.3\pm 0.1 GeV used as an input parameter, these relations predict the following glueball masses: M(0^{++})\simeq 1.65\pm 0.05 GeV, M(1^{--})\simeq 3.2\pm 0.2 GeV, M(2^{-+})\simeq 2.95\pm 0.15 GeV, M(3^{--})\simeq 2.8\pm 0.15 GeV. We briefly discuss the failure of such relations for the pseudoscalar sector. Our results are consistent with (quasi)-linear Regge trajectories for glueballs with slope \simeq 0.3\pm 0.1 GeV^{-2}.

Recently performed investigations of meson spectra allow us to determine the resonance structure for the waves $IJ^{PC}=00^{++}$, $10^{++}$, $02^{++}$, $12^{++}$ = $IJ^P= 1/2 0^+$ in the mass region up to 1900 MeV, thus establishing the meson multiplets $1^3P_0q\bar q$ and $2^3P_0q\bar q$. Experimental data demonstrate that there are five scalar/isoscalar states in this mass region. Four of them are $q\bar q$ states, that is, members of the $1^3P_0q\bar q$ and $2^3P_0q\bar q$ nonets, while the fifth state is an extra one not accomodated by $q \bar q$ systematics; it has the properties of the lightest scalar glueball. Analysis of the $00^{++}$-wave performed within the framework of the dispersion relation technique allows us to reconstruct the mixing of a pure gluonium with neighbouring scalar $q \bar q$ states belonging to $1^3P_0q\bar q$ and $2^3P_0q\bar q$ nonets: three scalar mesons share the gluonium state between each other -- those are two comparatively narrow resonances $f_0(1300)$ and $f_0(1500)$ and a broad resonance $f_0(1530^{+90}_{-250})$. The broad state is a descendant of the gluonium, keeping about 40-50% of its component.

The unexpectedly large branching ratios for $B\to \eta' K (\eta' X_s)$ decays could be of gluonic origin. We study the implications for $B\to \eta K (\eta X_s)$ and $P K (PX_s)$, where $P$ is the pseudoscalar glueball. In the mechanism proposed by Fritzsch, large branching ratios are predicted for these modes. The $B\to \eta K$ rate is barely within the experimental limit, and $B\to P K$, $PX_s$ could be at the 0.1% and 1% level, respectively. Smaller but less definite results are found for the mechanism of $g^* \to \eta' g$ via the gluon anomaly.

I review what lattice QCD simulations have to tell us about the glueball spectrum. Talk given at HEP97, Jerusalem, August 1997.

The hadronic dressing of the ten lightest scalar mesons, the $q{\bar q}$ nonet and a glueball, is discussed.

The production of $J^{PC} = 1^{++},0^{-+}$ and $2^{-+}$ mesons in double tagged $e^+e^- \to e^+e^-R$ is calculated and found to have the same polarisation and dynamical characteristics as observed in $pp \to ppR$. Implications for the spin structure of the Pomeron are considered. Production of $0^{++},2^{++}$ mesons in these two processes may enable the dynamical nature of these mesons to be determined.

We argue that the $1^{--}$ glueball $O$ is much cleaner than other glueball candidates. Anomalously large $J/\psi \to \rho\pi, K^*\bar K$ decays and the scalar glueball mass scale suggest that $m_{O} \simeq m_{J/\psi}$.

Motivated by the recent experimental data about candidates for glueball from different processes, we discuss in this paper the production of 0^{++} glueball from double diffractive scattering at momentum transfer |t|< 1GeV^2 in high energy p+p(\bar p) collision. We apply the phenomenology of Pomeron (IP) of Donnachie-Landshoff, the field theory model of IP of Landshoff-Nachtmann and the relevant calculating approachs. We assume while IP coupling with glueball, the 0^{++} glueball can be considered as a bound state of two non-perturbative massive gluons. We evaluate the dependence of cross section for 0^{++} glueball production on system energy $\sqrt{s}$ and show that it could be tested experimentally.

The application of our glueball-qqbar filter to the centrally produced KKbarpi system shows that the f1(1285) and f1(1420) have the same behaviour; namely consistent with the f1(1420) being the partner to the f1(1285) in the 3P1 nonet of axial mesons. We determine a flavour singlet-octet mixing angle of approximately 50 degrees for this nonet and highlight that the existence of the supposed f1(1510) needs confirmation.

Computations in the quenched approximation on the lattice predict the lightest glueball to be a scalar in the 1.5-1.8 GeV region. Here we calculate the dynamical effect the coupling to two pseudoscalars has on the mass, width and decay pattern of such a scalar glueball. These hadronic interactions allow mixing with the $q \overline q$ scalar nonet, which is largely fixed by the well-established K_0^*(1430). This non-perturbative mixing means that, if the pure gluestate has a width to two pseudoscalar channels of ~100 MeV as predicted on the lattice, the resulting hadron has a width to these channels of only ~30 MeV with a large eta-eta component. Experimental results need to be reanalyzed in the light of these predictions to decide if either the f_0(1500) or an f_0(1710) coincides with this dressed glueball.

The background due to the direct diffractive dissociation of the photon into the $\pi^+\pi^-$-pair to the "elastic" diffractive $\rho^0$-meson production in electron-proton collisions is calculated. The amplitude for the background process $\gamma p \to \pi^+ \pi^- p$ is proportional to the $\pi$-meson - proton cross section. Therefore, describing the HERA data, we can estimate $\sigma (\pi p)$ at energy $s_{\pi p}\sim (2 - 3) 10^{3} GeV^2$ that is considerably higher the existing data. At large $Q^2$ the interference between resonant and non-resonant $\pi^+ \pi^-$ production leads also to the more slow increase of the $\sigma^L/\sigma^T$ ratio with the mass of the $2\pi$ (i.e. $\rho^0$-meson) state.

Basing on the previously performed K-matrix analysis of experimental data, we investigate, in the framework of the propagator matrix (D-matrix) technique, the 1100-1900 MeV mass region, where overlapping resonances $f_0(1300)$, $f_0(1500)$, $f_0(1530^{+90}_{-250})$, and $f_0(1780)$ are located. Necessary elements of the D-matrix technique are developed. The D-matrix analysis confirms previous K-matrix results: in the region 900-1900 MeV five scalar/isoscalar states are located. Four of them are members of the two $q\bar q$-nonets, while one state is an extra for the $q\bar q$ systematics, being a good candidate for the lightest scalar glueball. The D-matrix analysis shows that this extra state, a candidate for the lightest scalar glueball, is dispersed, due to a mixing with $q\bar q$- states, over three resonances: $f_0(1300)$, $f_0(1500)$, and $f_0(1530^{+90}_{-250})$. The broad resonance $f_0(1530^+{90}_{-250})$ is a descendant of the lightest glueball carrying about 50% of the gluonium component, the rest of the gluonium is shared between $f_0(1300)$ and $f_0(1500)$.

The $f_0(1500)$ is identified in three glueball favoured production mechanisms: $p\bar{p}$ annihilation, $\psi \to \gamma f_0$ and Central Production. The production rate for glueballs in $\psi \to \gamma G$ has been quantified and the $f_0(1500)$ is found to be consistent with a glueball - $q\bar{q}$ mixture. We illustrate a remarkable property of central production where kinematic cuts appear to make a systematic separation between glueballs and $q\bar{q}$ of the same $J^{PC}$. When the cut favouring glueballs is applied, the $f_0(1500)$ and other enigmatic states appear prominently while confirmed $q\bar{q}$ states are empirically suppressed.

We have stumbled upon a remarkable empirical feature of central meson production which separates established $q\bar{q}$ mesons from glueball candidates. This does not appear to have been noted previously and we have no simple explanation for it. We suggest that glueballs and $q\bar{q}$ of the same $J^{PC}$ are distinguishable due to their boson versus fermion internal structure and that this leads to a different topology for central production of glueballs and $q\bar{q}$. Upon application of this test to data from the WA102 experiment we find that the $f_0(1500)$ and the $f_{2}(1900)$ show behaviour consistent with glueballs and opposite to that exhibited by established $q\bar{q}$ states.

Basing on the results of the K-matrix fit of $(IJ^{PC}=00^{++})$ wave, we analyze the analytic structure of the amplitude and $q \bar q$/glueball content of resonances in the mass region 1200-1900 MeV, where an extra state for $q \bar q$-systematics exists being a good candidate for the lightest scalar glueball. Our analysis shows that the pure glueball state dispersed over three resonances: $f_0(1300)$, $f_0(1500)$ and $f_0(1530^{+90}_{-250})$, while the glueball admixture in $f_0(1750)$ is small. The broad resonance $f_0(1530^{+90}_{-250})$ is the descendant of the lightest pure glueball. The mass of pure glueball is $1630\pm {70\atop 30}$ MeV, in agreement with Lattice calculation results.

In this paper we continue previous efforts in the literature to determine phenomenological values for the gluon mass by confronting theoretical results obtained in a theory of massive gluons with experimental values or results directly referring to the nontrivial structure of the Yang-Mills vacuum, e.g. to the presence of the gluon condensate. The decays of heavy quarkonia into 3 gluons and 2 gluons + photon are considered in detail as well as the correlators of the electromagnetic current and the scalar glueball current. Based on the analysis for the latter quantities a value for the gluon mass in the range of 500-600 MeV is estimated from the standard SVZ-value of the gluon condensate.

We summarize the results of the investigation of the $IJ^{PC}=00^{++}$ wave which is devoted to the search for the lightest scalar glueball. In the framework of the K-matrix formalism, the $q\bar q$-nonet classification of the scalar/isoscalar states is performed in the mass region below 1900 MeV basing on the following data: GAMS data for $\pi^-p\to\pi^0\pi^0n$, $\eta\eta n$, $\eta\eta'n$; CERN-M\"unich data for $\pi^-p\to\pi^+\pi^-n$; Crystal Barrel data for $p\bar p\to\pi^0\pi^0\pi^0$, $\pi^0\pi^0\eta$, $\pi^0\eta\eta$; BNL data for $\pi\pi\to K\overline K$. It is shown the existence of an extra state for $q\bar q$ classification: this state is a good candidate for the lightest scalar glueball. Analysis of the glueball/$q\bar q$ mixing shows that the glueball state has mainly dispersed over three real resonances: $f_0(1300)$, $f_0(1500)$ and $f_0(1530^{+90}_{-250})$. The broad resonance with complex mass $m=1530^{+90}_{-250}-i(560\pm 140)$ MeV is descendant of the pure glueball. The restored mass of the pure glueball is equal to 1695 MeV. This value agrees well with the results of lattice gluodynamics.

A simple mixing scheme describing the $f_0(1500)$ and the $f_0(1700)$ as mixed states of a $\bar s s$ meson and a digluonium is reconsidered at the light of new experimental data.

I briefly review lattice QCD, the valence (quenched) approximation, and the application of both to the determination of the mass and two-body decay couplings of the lightest scalar glueball. Results in agreement with the observed properties of $f_J(1710)$ strongly suggest this resonance is largely a scalar glueball.

The ratios of partial widths for the decay of a glueball into two pseudoscalar mesons are calculated under the assumption that the production of light quark pairs ($u\bar u,d\bar d,s\bar s$) in soft gluon-\-induced reactions goes on within universal symmetry breaking. Parameter of the violation of flavour symmetry is fixed by the central hadron production data in high energy hadron collisions and/or by the ratios of radiative decay amplitudes $J/\Psi \to \gamma\eta/\gamma\eta'$ and $J/\Psi \to \gamma \phi \phi/\gamma \omega \omega$. The ratios of coupling constants $glueball\to\pi\pi,K\bar K, \eta\eta,\eta\eta'$ which are calculated with this parameter coincide reasonably with those of $f_0(1500)$, supporting an idea about glueball nature of $f_0(1500)$.

On the basis of the results of $\xi(2230)\rightarrow\pi^{+}\pi^{-}, p\bar{p}$ and $K\bar{K}$, measured by the BES Collaboration in radiative J/$\psi$ decays, combined with the upper limit of Br($\xi\rightarrow p\bar{p}$)Br($\xi\rightarrow K\bar{K}$), measured by PS185 experiment, we argue that the distinctive properties of $\xi$(2230), the flavor-symmetric decays and the narrow partial decay widths to $\pi\pi$ and $K\bar{K}$ as well as its copious production in radiative J/$\psi$ decay, would strongly favor the glueball interpretation of $\xi$(2230).

Following the discovery of two new scalar mesons, $f_0(1370)$ and $f_0(1500)$ at the Low Energy Antiproton Ring at CERN, we argue that the observed properties of this pair are incompatible with them both being $\qqbar$ mesons. We show instead that $f_0(1500)$ is compatible with the ground state glueball expected around 1500 MeV mixed with the nearby states of the $0^{++}$ $Q\bar{Q}$ nonet. Tests of this hypothesis include the prediction of a further scalar state, $f_0'(1500 - 1800)$ which couples strongly to $K\bar{K}$, $\eta\eta$ and $\eta\eta'$. Signatures for a possible tensor glueball at $\sim $2 GeV are also considered.

We show that the newly discovered scalar meson $f_0(1500)$ at LEAR has properties compatible with the lightest scalar glueball predicted by lattice QCD and incompatible with a $\qqbar$ state. We suggest that decays of glueballs are into pairs of glueballs (including $\eta, \eta'$ or $(\pi\pi)_S$) or by mixing with nearby $Q\bar{Q}$ states. The partial widths of $f_0(1500)$ are in accord with this hypothesis, tests of which include characteristic radiative decays to $\gamma\phi,\gamma\omega, \gamma\rho$ and the prediction of a further scalar state, $f_0'(1500 - 1800)$ which couples strongly to $K\bar{K}$, $\eta\eta$ and $\eta\eta'$.

On the basis of the recent results of $\xi(2230)\rightarrow \pi^+\pi^-, p\bar p$ and $\xi(2230)\rightarrow K^+K^-,~K_SK_S$, measured by the BES Collaboration in $J/\psi$ radiative decays, combined with the PS185 experiment of $p\bar p\rightarrow \xi(2230)\rightarrow K\bar K$, we argue that because of its very narrow partial decay widths to $\pi\pi$ and $K\bar K$ ( less than $1~MeV$ ), its large production rate in $J/\psi$ radiative decays, and its flavor- symmetric couplings to $\pi\pi$ and $K\bar K$, the $\xi(2230)$ is very likely to be a $J^{PC}=(even)^{++}$ glueball.

We discuss the influence of glueball coupling to nucleons on the weak axial-vector coupling constants including singlet channel. We consider a possibility of introduction of constituent gluon contribution to the proton spin. The estimated value for this quantity seems to be rather small.