List of points to clarify received by Sept 15
        =============================================


*** low mass dileptons

Are the S-U and Pb-Pb data of NA38/50 compatible with the rho mass
shift interpretation of the Ceres low mass dilepton enhancement?  Can
we see a quantitative comparison of the dilepton spectra from CERES
and NA50 in the mass region where the two acceptances overlap?  If
CERES doesn't see the rho anymore (no peak left), how can NA50 measure
a phi/(rho+omega) ratio?

Is there any drop in the A-dependence of the rho+omega production
cross section, measured by NA38/50, from p-A to S-U and Pb-Pb
collisions?  Can NA38/50 show that the enhancement of the ratio
phi/(rho+omega) is *not* a consequence of the drop in number of
"bare-mass" rhos?

Is the NA49 phi yield in central Pb-Pb collisions compatible with the
NA38/50 data?

Is there any experimental evidence that rules out the hypothesis that
the "excess signal" observed by Ceres is actually unsubtracted
background?  To probe this, make the assumption that the "excess
signal" *is* indeed unsubtracted background and see if that hypothesis
leads to unreasonable mass and pt signal distributions, or their
charged multiplicity dependence, or any other relevant plot.

How well does one know the expected physics contributions to the
dilepton and photon spectra?  What is the contribution of the eta to
the inclusive photon and dilepton spectra?

What do we really learn about chiral symmetry restoration from the
low-mass dilepton data?  Not in the sence of the 'dropping masses'
*conjecture* but on the basis of vector and axial-vector mixing, a
model independent prediction of QCD.


*** direct photons

How sensitive are the results of WA98 to the assumption of mt-scaling
in the calculation of the background from eta and omega decays?  What
would be the extracted level of direct photon production in Pb-Pb
after taking into account the growth of T with the particle mass?
What about the sensitivity to the shape of the pion spectrum at large
pt? (pi0 vs charged pions and sensitivity to extrapolation of pion and
eta spectra beyond the measured pt range).


*** psi suppression

Show the NA50 data for J/psi and DY separately as a function of ET,
not only as the ratio psi/DY.

Present a critical and quantitative discussion of the high ET
"fluctuations".

Evaluate the robustness of the results against rebinning by showing
plots of J/psi over DY as a function of ET for different ET binnings
(sizes and positions).

Review all the available information on how well we understand the DY
continuum measured by NA38/50.

Show the pt or (even better) the mt spectra of J/psi (and DY) instead
of just giving mean pt; compare to available models with just comover
interactions and with Debye screening.  There is a lot more
information in the shape of the spectrum than in the average pt.
Also, the spectrum gives a much better impression on the quality of
the data.

Why is the number of participants not the pertinent variable in the
study of J/psi production and suppression?

Which model-independent conclusions (in particular concerning the chi
behaviour) can we derive from a comparison of the psi and psiprime
suppression patterns?  Consider 3 "centrality" ranges : pA; S-U and
peripheral Pb-Pb; central Pb-Pb.

If the psiprime starts to be suppressed much earlier than the chi and
the chi has a bigger radius than the psiprime, can we rule out the
model of Debye screening, where the size of the quarkonia is expected
to be the pertinent variable?

What is the meaning of the "sudden onset" seen by NA50 in the psi over
DY ratio at ET around 40 GeV?  If the psi physics depends on the
impact parameter of the Pb-Pb collision, the significant fluctuations
in particle production between different collisions of identical
centrality would smear out any "sharp jump", if measured versus ET.
Must we understand the data as evidence that ET is *the* variable upon
which depends the psi production and suppression?

Is it by coincidence that the A-dependence of J/psi is so similar to
that of charged particle production?

[Added after Sept 15]
Show the NA50 data for J/psi over DY as a function of EZDC, not only
as a function of ET.

Is the ET dependence of the ratio psi/DY compatible with the ET
dependence of the average pt**2 of the psi, within (for instance) the
deconfinement scenario?

In order to make the "drop", in the psi/DY versus L pattern, shift to
a higher or a lower value of L (or of energy density or of some other
variable), what is the most suited collision system to use?  What are
the pros and cons of doing, for instance, Ag-Ag collisions instead of
Pb-Ag collisions?


*** intermediate mass lepton pairs

The excess in IMLP (1.5 - 2.5 GeV) has been shown to vary more or less
smoothly with the number of participants and its mass and pt spectrum
seems to be consistent with the expected distributions from
charm.  Concerning the data, we would like to see the following more
quantitative analysis:

pA data: All nuclei have been fitted with a constant charm/DY ratio,
and the data seems to be consistent with this. However, one should be
able to quantify the possible A dependence of this ratio by fitting
all nuclei independently (with a free ratio DDbar/DY) and give limits
on the A and participant number dependence, eg in the form
A**(alpha_DDbar-alpha_DY).  Can one make a statement that there is a
sigificant difference (and quantify the significance level) between
the Npart dependence in pA and AA?

AA data: We would like to see the mass, pt, y and costheta
distributions of the excess (i.e. after subtracting DY and other
contributions in the relevant mass window). What is the experimental
limit on any change in these distributions? (This can be shown eg by
dividing some representative 'small excess' and 'large excess'
centrality samples)

Interpretation: Current possible interpretations include increased
charm production, change in the charm kinematic distributions (eg via
initial/final state rescattering), thermal pairs:

Change in charm kinematic distributions: Given the limit (or extend)
on changes of kinematic distributions (m, pt, y, costheta, see above)
allowed by the experimental data (in a restricted acceptance), can one
make any model-independent statements about this explanation? One
particular model (Xin-Nian Wang) has been published, can it be
compared with (excluded by) the data?

Thermal pairs: are there any thermal calculations one could compare
the data (mass, pt, mt) distributions with (obviously in the NA50
acceptance)?

Question to Theory: if the open charm cross section is enhanced (by a
factor 3), would one not expect that also the J/Psi production cross
section changes? Or vice versa, which QCD processes could one imagine
to increase charm but not influence J/Psi (at the production level)?

Is there a connection between the low and the intermediate mass
enhancements?

[Added after Sept 15]
Is there any experimental evidence that rules out the hypothesis that
the "excess signal" observed by NA38/NA50 is actually unsubtracted
background?  To probe this, make the assumption that the "excess
signal" *is* indeed unsubtracted background and see if that hypothesis
leads to unreasonable mass and pt signal distributions, or their
charged multiplicity dependence, or any other relevant plot.


*** strangeness

Is the Omega inverse mt slope for peripheral Pb-Pb events different
from the value for central collisions?  Consider the border to be
where the J/psi data has a "break", about 160 participant nucleons.

We need a thorough and detailed discussion of phase space
extrapolations, corrections due to not (or incompletely) identified
particles, etc., influencing the results on global strangeness
production in S-S and Pb-Pb collisions.  Could we have a general rule
for extrapolating a measured spectra to full phase space?

How is strangeness distributed and balanced in y and pt in S-S, Pb-Pb
and other collision systems?  Which pieces are still missing (and how
can we get them)?

How sensitive to the assumed hyperon rapidity distributions are the
extrapolations to the domain -0.5<y<0.5, in the case of the (very
asymmetric) p-Pb system?

How do the multistrange hyperon yields change from p-Be (a much more
symmetric system) to p-Pb?  If the measured yields of multi-strange
hyperons per event and per participant nucleon *decrease* from p-Be to
p-Pb, the "enhancements" become a problem of phase space coverage and
rapidity shifts.  Specify where in rapidity multistrange hadrons are
measured for different systems and compare with p and pbar
distributions.

Is the strangeness enhancement a smooth curve from pp to the most
central Pb-Pb collisions, going through p-A, S-A and peripheral Pb-Pb
data, or is there a discontinuity for a certain collision size?  When
will we be able to answer this essential question?

Is the possible disagreement between NA49 and WA97 lambda yields in
Pb-Pb collisions due to feed down from cascades?

Do the strange particle yields in p-Nucleus agree among the several
experiments (NA35, NA36 and WA97)?

What can we learn from lambda polarisation (if measured)?

What are the precise predictions on omega and xi yields at 40 A GeV/c?

Microscopical models tests on p-Pb for hyperons.  VENUS model
overestimates (one order of magnitude for Omega) the WA97 data on
hyperon multiplicities in p-Pb collisions.  It seems to be necessary
to compare all microscopic models to these p-Pb data in order to check
the relabiality of the mechanisms introduced to explain the
strangeness enhancement in nuclear collisions.  Can we organise a
common effort to compare different models with p-Pb data?

Can we learn something by comparing the Omega with the phi?  Both
particles are composed of strange quarks only.

Is the Omega yield in central Pb-Pb collisions compatible with the
other hadronic species in terms of the chemical equilibrium picture?

What is the meaning of the relatively low inverse slope of the Omega
mt spectrum (relative to the NA44/NA49 trend)?  Is there a separate
trend for the multi-strange hyperons?

What can we conclude from NA49's finding of essentially no event by
event K to pi ratio fluctuations in central Pb-Pb collisions?

Is the relative Kaon production really enhanced compared to pp?


*** HBT

A recent NA49 paper presents the first 'physics' result from HBT at
the SPS extracting the freeze-out temperature and expansion velocity.
Can we trust the quantitative parameters given?  The application of
equation (6) seems to have a problem for the values of r and beta_perp
(is beta getting larger than one or is r cutted off?).  Comment on the
application of equations (8) and (9) to the pion spectra.

Compare NA44 results to NA49 published systematics; Discussion of
difference between the pi+ and pi- 'radii' in NA44 (opposite to what
is expected from Coulomb).

What are the final NA44 results on Kaon interferometry in central
Pb+Pb collisions?

What do we learn from the mt dependence of two particle correlations?


*** exotic

How do the state-of-the-art theoretical estimates for exotic light
nuclei production compare with the NA52 sensitivity?


*** thermal/chemical equilibration scenarios

Are the neutral pion spectra of WA98 compatible with the NA49 pion HBT
in what concerns freeze-out scenarios?

Can we have a deuteron mt spectra from NA44?  Is the corresponding
inverse slope compatible with the NA49 value?

Since the slopes of the mt spectra are continuously changing, as has
been shown by several experiments, the results of exponential fits are
very sensitive to the range of the fit.  Can we trust the relevance of
the "inverse slope systematics"?  How is this taken into account in
the extraction of transverse flow velocities?

It would be very interesting to have the ratios of the pt spectra
between heavy-ion data and pp or p-Be data.  What's the place of the
Cronin effect and of random-walk models in the understanding of pt
distributions?  Even better, we would like to see the different
spectra plotted together or, if too many details get lost in a log
plot, show the local slopes of the different spectra together.

How good is the evidence for thermalization and/or chemical
equilibration?  What more is needed to settle these questions?

Are the thermal models of Becattini et al. consistent with the one of
Braun-Munzinger et al.?  Do we observe strangeness equilibration,
enhancement or suppression?

How naive are our pictures of thermal and chemical freeze-out?  Does
the shape of the freeze-out hyper-surface influence the thermal model
interpretation?


*** collective behaviour, flow, ...

Concerning the directed (and elliptic) flow, can we rule out
non-hydrodynamic effects, like shadowing effects of the pions or
cluster breakup of baryons, as the mechanisms generating the flow?
Flow is interesting if it indicates hydrodynamic behaviour.  Compare
the pt dependence in the models with data.

What is the existing evidence for collective behaviour?  We need a
compilation and critical discussion of all available pieces of the
puzzle: radial, elliptic, directed flow; comparisons between different
experiments.


*** general

For every measured observable whose evolution pattern is expected to
provide information on the properties of heavy-ion collisions, a
complete "review" of reference data should be provided.  The reference
"baseline" should include pp and p-A data, peripheral B-A collisions,
data at lower (and higher) energies.

This has been done for the J/psi and low mass dileptons in more detail
than for the other observables.  There are HBT radii as a function of
particle multiplicity in pp collisions, p_T broadening in p-A, etc.

We need quantitative comparisons of yields, spectra and two-particle
correlations between the different experiments, in the regions of
overlapping acceptance.

Which systematic, discontinuous or continuous, changes of observables
exist (a) from pp to p-nucleus to nucleus-nucleus collisions, (b) at
the SPS or at other available energies, (c) in nuclear collisions
versus centrality.

Is the observation of *no* J/psi suppression in SS and SU collisions
understandable in the same context as the observation of strangeness
enhancement and low mass dilepton enhancement between p-A and S-A?

Can different observations like apparent chemical equilibrium and
thermalization, flow, strangeness enhancement, J/psi suppression and
dilepton enhancement be related with each other?

Can we agree on a method to estimate the primordial energy density?

Are the time duration parameters from two-pion interferometry in
central Pb-Pb (mean decoupling time ~8 fm/c, mean duration of emission
~4 fm/c) compatible with our expansion models and with the quantity of
finally observed "flow" radial velocity?

What exactly is a "thermal hadronic system"?  What should be the
proper hadronic (or, better, non-partonic) state right after the phase
change?

How can detailed balance be delt with in cascade type models at
energies where multiparticle production dominates?

What observable is sensitive to a large latent heat jump?

Where, when and how does the radial collective velocity field build up
in the hydro models which incorporate a parton-hadron phase change?
How can it be understood that at AGS the "flow" magnitude seems to be
the same as at SPS?

Can we have as a general rule to show both the subtracted and
unsubtracted spectra whenever the background is not negligible?  This
is particularly relevant when differential distributions like m_T
spectra are presented (Lambda, Xi, phi, J/psi, etc).

Is the effect of different beam energies negligible in the comparison
of S-S and Pb-Pb particle yields?

Should the m_T range of all particles used in the famous mass versus T
plot be the same?

What is the dependence of the particle yields on the number of
participants (a) in thermal models and (b) in microscopical models ?

For which physics observables the number of participants is the good
quantity to compare different systems, e.g. Pb-Pb and S-S?

Different solutions of hydro (many of them approximate) give different
velocities for the same temperature describing the same spectra.
People in the community pick at their liking whatever formula suits
them and come up with different parameters.  Somebody should clarify
(maybe Uli Heinz, since he is generating and using many of the
different formulae).


*** questions to Marek : [Added after Sept 15]

To derive his recent claim that the observed strangeness enhancement
is an indication of QGP formation M. Gazdzicki assumes that the number
of s and sbar is not altered during hadronization.  We would like to
ask him the following questions: what do we know about the
fragmentation of gluon jets? what do we know about the changes between
u, d and s quarks during hadronization?


To argue for the formation of a QGP at the SPS in Pb-Pb and S-S
collisions you have proposed a model for the QGP phase transition
which assumes :

1) applicability of a grand canonical partition function 

2) creation of a globally equilibrated system at very early times in
the collision

3) a first-order phase transition 

4) a mixed-phase state at the critical temperature

5) a step increase in entropy density from s-hadronic to s-qgp across
the critical temperature

6) an entropy-dominated free energy in the neighborhood of the
critical point

7) A deterministic correlation between equilibrated early-stage energy
density and the energy variable F.

8) Early-stage entropy production conserved throughout the collision
(with a small transfer from pions to baryons)

9) early-stage strangeness production conserved throughout the
collision
 
This model is compared to data in two forms.

1) entropy/baryon ratio vs F. Prediction: slope increases stepwise
with increasing F across the critical region due to transition from
s-hadron to s-qgp.

The slope of the prediction is adjusted to agree with data below the
estimated critical region. The slope above is taken from the model
alone.

2) Strangeness/entropy ratio vs F. Prediction: ratio decreases
stepwise with increasing F across the critical temperature due to
transition from s-hadron to s-qgp.

Data from N-N are compared with A-A and the prediction. The N-N data
are scaled up by a factor 3.6 to agree with A-A at AGS energies.

These comparisons seem to indicate evidence for QGP formation at SPS
in the form of possible discontinuities in these ratios, consistent
with the model.  However, other models not involving QGP formation or
differing strongly in basic assumptions may do equally well.  The
available data are consistent with these predictions but don't require
them.  Applying the usual standard, is the proposed model the simplest
one consistent with the available data?

Questions:

1) Independent of data, can a grand canonical approach be defended ab
initio for a very finite many-body system (< 10k `particles' total at
the SPS).

2) Can one reconcile the assumption of early-stage global
thermalization with a parton cascade picture, a string picture or
flow?

3) Items 3) - 6) above imply a negligible `width' to the phase
transition, whereas lattice calculations indicate a finite width of at
least 20 MeV in temperature, ignoring finite-size effects which may
further broaden the transition region. Why should one start with such
an extreme assumption.

4) This model assumes that the system is created in a particular
thermodynamic macrostate determined by F whose relevant properties
(entropy and strangeness abundance) are then preserved throughout the
expansion period. In this picture the collision system in some
respects does not transit the phase-transition region as a coupled
many-body system in either direction. Is this a reasonable ab initio
assumption?

5) What steps have been or could be taken to explain these ratio
trends in terms of a hadronic scenario?

6) If a grand canonical approach can be assumed for S-S and Pb-Pb at a
given F why doesn't it work for N-N? What is the dividing line? Why
aren't we as interested in a careful study of the A dependence of
these ratios for light A as in the F or root-S dependence?

7) In a small system is it relevant to talk about a mixed phase? What
would be the typical space-time structure of such a mixed phase?

8) Why is it reasonable to rescale the strangeness/entropy ratio for
N-N by 3.6, other than to make the trend agree at AGS energies? What
is the physics justification for this scaling?

9) Why assume an entropy-dominated free energy near the critical point
(most probable state determined solely from entropy arguments)?

10) Can we obtain separately an entropy-density estimate and a
strangeness abundance estimate vs A and F or root-S, rather than
simply the ratios?