c'est plutot quelque chose qui doit aller dans la discussion de l'erreur systematique. J'effacerais ces 4 mots.

j'ai un petit mal avec cette phrase, je trouve mieux:

"EXCALIBUR was used to generate events with colour reconnection following the ansatz of ....."

je propose:

"[...] between data and Monte Carlo."

Phrase suivante: pas de "final"

je trouve que c'est trop tot, donc je propose "delete" ou de le mettre au debut de 2.2.3Results

difficile a digerer, c'est mieux comme ca:

"[...] the energy of the most energetic quark jet [...]"

pour eviter des commentaires de Wasserbaech/JFG/Patrick

"[...] and by the uncertainty introduced by the cut on energy detected close to the beam axis in the tvqq analyses."

"[...] obtained assuming lepton universality only is:"

j'aime mieux:

"[...] obtained assuming only lepton universality is:"

"[...] adding together the data samples collected at [...]"

plutot

"[...] combining the data samples collected at [...]"

Dernier commentaire: je trouve que la discussion de la deviation de la mesure par rapport a ce qu'on a mesure et les trucs DPA IBA doivent aller dans la section 3 au moment on donne des resultats et pas dans les conclusions.

-par 3: Rather than the "measurement of cross-sections ..." I would say the "separation of WW events in all possible W decay channels allows the determination of W br...". The measurement of the total ee-->WW reaction is another test of the SM, separate from the BR measurements (if you want to say it so, before the comment on the br mesurements).

-par 7 (line3): "an excellent agreement" can we replace "excellent" with a quantitative xsection or efficiency difference. (or just say it is taken into account in the syst)

-last line: "which, for sqrt(s)=....[18], gives sigma_WW="

-par 2: "Beam related...the beam" Sounds unclear. What about "Beam related background and detector noise that is not simulated in Monte Carlo events, affect the efficiencies of the cut which has the purpose of removing gamma-gamma events depositing energy ...the beam"

-par 4: "Events are classified into six di-lepton channels making use of electron and muon identification criteria. A jet or a track is classified as a tau either if no lepton identification is fulfilled or if the identified lepton has an energy lower than 25 GeV."

-par 1:"Similary to the lower energy measurements three selection procedures are applied.One selection is optimised for WW events with an electon or a muon, and requires an identified (remove energetic!) electron or muon track. The other two are designed to select other tvqq events, based on ... events. As the ... selections select efficiently tvqq events with leptonic tau decays, and the tau selections select also e/munuqq events, a reclassification of selected events is finally performed."

--Move the beginning of par 2.2.3 "The probability for ... (Fig1)." to the end of par 2.2.1 .

Change "from a signal process..." to " from each of the three evqq,mvqq or tvqq signal processes...".

I don't know exacly how the selection has been finally applied but there should be a general cut on the total lvqq probability like Prob(lvqq)>0.40 {Prob(lvqq)=prob(evqq)+Prob(mvqq)+Prob(tvqq)} to select all events. After the 3 three e-mu-tau prob components should be used only for reclassification purposes.If this is what is done we should try to state it . The global Prob(lvqq) should be stated in the 2.2.1 par and the reclassification business with partial Probs in par 2.2.3 .

-page4 line3:"This classification improves .." --> "This classification reduces the electon-tau and muon-tau correlations (section 4)."

-sect 3. last par: Bleah. sigma_WW as a simple sum is not nice !

Ok we had a lot of discussions about this ... but my personal taste is still the same. Let me say again that such a summed total xsection is built violating the consistency of W BR within the 3 samples. It is not bad because it does not have SM BR but because it just does not have a common hadronic/leptonic BR for the produced W-pairs at all.

And if we put together these 3 topologies we do need some common BR if we assume that all 3 of them come from W-pair decays. No?

-Remove ALEPH 99-?... add CERN-EP

INTRODUCTION- "...determination of the W branching ratios and, indirectly, the coupling of the W to cs pairs."

If I look at table 2 the main systematic errors are

MC stat, bkg normalization and beam bkg. I don't

understand what do you mean by signal eff.

A general comment on the paper... you have chosen to mention

only what are the main systematic errors in each section.

The problem is that by looking at table 2 it is very

difficult in some cases to understand what is the meaning

of the error just looking at the tag in this table. Even

worse, nowhere is explained how in fact this errors have

been calculated, what actually has been done. This problem

is particularly relevant in section 2.3 for the 4q channel.

- As you have decided to compare the result in section 2.3

with the SM predictions, why you don't do the same in sections

2.1 and 2.2? it would be nice to write a sentence after the

result with the SM prediction to be compared for the sake

of consistency.

SECTION 2.2- "... and by the uncertainty coming from the cut on energy around the beam for tnqq events." Is this lepton isolation in table 2??? ... again the same problem: difficult to understand the meaning of the

errors in table 2 and what has been done to determine them.

- Same comment than in section 2.1 for the SM predictions.

- "... gives an identical result, but slighly less precise."

- why you don't simply quote the results mentioning that they

are compatible, in place of quoting 10 fb lower and 10% higuer

error? if you quote 6.74 +- 0.25, the reader does not need to

use the calculator...

• same comment "... 40 fb lower with a 18%..."

- "Small but significant disagreements appear for jet energy

distributions in particular. These differences can be propagated

at LEP2 by energy rescaling and all alter the shapes of the

background distributions in such a way that the resulting

W pair cross-section is lowered. The cross-section is

corrected upwards by 100 fb accordingly:"

I see that you have decided a radical change of philosophy

here, which is not what it was agreed in the W meeting. We have

agreed in the meeting that this 100 fb should be considered

as an aditional source of systematic error, and not a shift.

To conclude that this is a shift on the central value you

have to show that the error in this 100 fb is negligible. You

have to explain what are the variables not reproduced by the

MC, and if we decided our MC is not reproducing data we

have to be consistent with this in ALL our papers ( W mass

for instance). Are you sugesting that our MC does not reproduce

Jet Energies even after applying the Jet Energy corrections?

If, finally, you have decided this is a shift, then it is

not more a systematic error. This is in contradiction with

the next paragraph:

"The uncertainty on the variables used for the event

preselection cuts, and for the NN input is assessed by

changing jet directions and energies according to possible

residual differences between LEP1 Z data and MC..."

My personal opinion is: do not take this as a shift

but as a systematic error.

• "... as a function of the c.o.m. energy."

- Table 2 cannot be understood unless somewhere in the text is

explained what is the meaning of the tags, and what has been

done to compute the systematic errors. For instance, I'm afraid

that what is called FSI only corresponds to BE, and so on…

better in section 3, and to have in section 5 only a summary

of the results.

'Combining all final states and assuming lepton universality, the total

W-pair cross section is measured ....'

24fb are not clearly explained and referred to in Table 2. MC stats does not appear to account entirely for the 'background normalisation' error (this is 7fb in the text). Is 'MC stats' the error in the signal

efficiencies part of which comes from the 0.5% uncertainty in the random triggers? The 11fb 'beam bkgd' is not explained.

Remove the sentence ' For this purpose the topology....'

Is it true that the e/mu ID criteria are applied

AFTER the candidate lepton is chosen - seems quite inefficient?

event....' and 'If an event is...' should be transferred to the end of

section 2.2.1.

However, the second sentence can be re-worded: 'If the

probability is greater than 0.40 the event is kept in its corresponding channel otherwise it is considered by the tvqq selection.

Then, it would be better to transfer the rest of the 1st paragraph

in section 2.2.3 to the end of sec 2.2.2

Remove 'combined' from 2nd sentence.

If 'MC stat' is really the error in the subtraction of bkgds it should be relabelled 'Bkgd subtraction'. The origin of the 'Beam bkgd' could be stated as well as the bkgd normalisation etc.

These differences are quantified using the 14 NN variables taking correlations and energy rescaling to 189 GeV into account. For example, Fig~\ref{Ejetmax} shows the effect of these adjustments in the case of the maximum jet energy distribution from 4-jet events at 189 GeV where the agreement with data is improved where the QCD background dominates. Assuming that these adjustments are independent of\sqrt(s), the MC events in the background NN distribution are reweighted accordingly. The corrected fit result is:

\sigma(WW>4q) = 6.89 +- 0.23 (stat) +- 0.13(syst) pb

If only events with a NN output > 0.3 are kept a value of ?+-? is found showing no significant bias due to the QCD background. The measured cross section is 2.5 standard deviations lower than the Standard Model expectation of 7.59 pb [21].

The sources contributing to the systematic error are summarised in Table 2 ..........(Here we should relabelled 'qqgg jet' as 'background shape' changed to 90(?) MeV); relabel 'jet correction' as 'jet energy calibration';

point out that the bkgd normalisation assumes a 5%(?) variation in its

xsection; say what beam background is and separate 'MC stats' into

'efficiency' and '4f-cc03 correction'. So, the text should refer to the list of specific tags used in Table 2 where some clarification is needed.)

Uncertainties with the WW generator have been evaluated...... coupled to the parton level of KORALW signal events (all in the same paragraph).

Several cross checks have been performed on the fit result to search

for possible biases arising in the selection and full simulation of the WW events. The previous version of the NN [1],..... from 6.65 to 6.77pb (stop at this point).

We should then give the fit result obtained assuming SM branching ratios for comparison.

1. No hyphen in "cross section".

2. Use an italic "s" in \sqrt{s}.

3. For consistency, make the particle symbols "Z", "q", and "e" roman,

not italic.

4. Be consistent: use $l$ or $\ell$ for leptons (including in subscripts,tables, etc.).

5. Because there is only one appendix, it doesn't need a letter ("A").

E.g., try "\section*{Appendix: Hadronic...}".

6. Move the figures and tables earlier so they appear near the places

they are mentioned.

7. Do we really want to have italic table and figure captions? I don't see any advantage over roman; I guess it is a matter of taste. But it does have the disadvantage that half of the symbols and numerical stuff come out roman(!) and it would take some additional effort to make everything italic (including material that is now made roman by your macros). See table 1 and fig 3 for examples.

8. Remove "PRELIMINARY" from all figures!

9. Concerning colors in figures: I have no evidence that they will cause problems, but do we know for sure what the color figures will look like when they are rendered in black and white in the journal? I think it would be prudent to prepare truly black and white versions for the journal, and use the color ones for the preprint, etc.

abstract, 3rd sentence: it is peculiar to say we measure the WW cross

section while assuming Standard Model WW production. I think we have

to be more specific about what we assume if we want this sentence to

make sense.

"assumptions for W-pair" (delete "the").

p 1.3:

No comma between "recorded" and "at".

next sentence: delete "the" between "in" and "t-channel".

"exchange, referred to as the CC03 diagrams~[7]."

next sentence: don't start a new paragraph.

"The measured cross sections are corrected for the difference, denoted

the "4f-CC03 correction", in the accepted cross sections for CC03

processes and all Standard Model four-fermion final states consistent

with W-pair decays."

two sentences later: "Comparison samples generated with".

Hyphen in "four-fermion".

"diagrams show excellent agreement."

p 1.7:

How about "$\gamma\gamma$" instead of "gamma-gamma"?

Insert "and" between "production," and "BHWIDE".

next sentence: "which gives sigma_WW = blech for sqrt{s} = blech

and m_W = blech."

Use one pair of $$ around each of these equations in order to get the

spacing correct.

two sentences later: "of the cut that removes".

How about "$\gamma\gamma$" instead of "gamma-gamma"?

"depositing energy within 12degrees of the beam axis."

p 2.2:

I don't think the "resp." is used in English. Until I joined ALEPH I had never read such a thing. [If you insist on using it, please write

"resp.\ " to get the spacing correct.]

p 2.3:

"assigned to a tau" is a bit awkward.

next sentence: "A maximum likelihood fit is applied to determine the

cross section for each individual fully leptonic...". [Each channel

has one cross section.]

p 2.5:

How about this? "The results of the fit are" [no colon after "are" or

"yield".]

Period at the end of the last equation.

p 2.8:

"The three lnuqq selection procedures".

two sentences later: "selections have sizeable efficiencies for taunuqq events, as does the tau selection for e/munuqq events, a reclassification is performed."

p 3.1:

Remove "energy" between "sqrt{s}" and "the".

next sentence: "In this analysis, the selection of the lepton track

therefore relies...".

p 3.2:

"theta_lh the angle of the track to".

two sentences later: "so that taunuqq events where...will also be".

two sentences later: no comma between "direction" and "that".

[If you want to keep that comma, you need to add one after "cone".

But I recommend no comma.]

Space between "5" and "GeV".

two sentences later: "were described in detail in previous papers~[1],".

"energy-based cut values".

Get rid of the spurious space after sqrt{s}.

next sentence: no space between "0.5" and "%".

p 4.7:

"network" instead of "net".

p 4.8:

"where a four-jet event becomes a three-jet event".

next sentence: "contain" instead of "contains". [subjunctive!]

next sentence: hyphen in "Four-fermion".

p 5.0:

"are described in the appendix and are" [no comma].

next sentence: "simplified with respect to that used".

"with fewer input variables".

p 5.2:

Don't start a new paragraph at the sentence "The fit result is...".

p 5.3:

"gives" instead of "has given".

"with an 18%".

two sentences later: "non" is not a word (in English!), so it must be

attached to other stuff, at least by a hyphen: "non-four-quark".

"output" is misspelled.

"are kept, yielding a value".

next sentence: space between "100" and "fb".

next sentence: don't start a new paragraph.

two sentences later: "The uncertainties related to the variables".

No comma after "cuts".

"the NN input are assessed".

Spell out "Monte Carlo".

No hyphen in "miscalibration".

next sentence: "Uncertainties in the WW generator" [or "associated with"

or...].

"are evaluated by comparing".

Spell out "Monte Carlo".

Semicolon instead of comma after "samples".

"EXCALIBUR is also used to".

next sentence: "non-b".

two sentences later: no space between "0.3" and the period.

The column headings (especially "qqqq") aren't centered above the numbers.

Units: don't write "in"; write "(%)" and "(fb)".

Don't let the \hlines hit the words in the first column (bottom few rows).

[I can provide TeXnical help if you wish.]

p 6.5:

"previous sections and uses the". [no comma]

next sentence: no colon after "is".

Use one pair of $$ around this equation.

next sentence: "of the centre-of-mass energy."

next sentence: "performed adding" is bad grammar.

In the table:

Don't capitalize "state interactions".

Spell out "Monte Carlo" and "identification".

p 7.5:

"taunu versus enu or munu," is evidently a spurious sentence fragment.

p 7.8:

Period at the end of eq 1.

Do we want to mention "to first order in alpha_s"?

p 7.9:

Make the "cs" roman in the displayed equation.

The vertical axis values are too close to the axis.

Use a symbol font for the axis values, in order to get minus signs

instead of hyphens.

There are too many tick marks and values on the horizontal axes.

(Don't need two ticks per bin or one value every 2.5 bins.)

The HIGZ fonts (in the keys) are UGLY!

The text in the keys is too small, as are the labels "a)", "b)", "c)".

I don't think Phys Lett will appreciate all the empty space in this

figure; if you stack the three parts with no space in between you can

save a lot of room. The aspect ratio of the plots could also be

changed so the stacked version doesn't come out too narrow when it

is placed in one column of the journal.

Caption: "for the (a) enuqq, (b) munuqq, and (c) taunuqq selections."

two sentences later: combine sentences: "background contribution, the

dark shaded histogram..."

The vertical axis values are too close to the axis.

Space instead of hyphen in the horizontal axis label.

There are too many tick marks and values on the horizontal axis.

Caption: hyphen in "four-quark".

other figures?

Space between "sigma_WW" and "(pb)".

Space between "E_cm" and "(GeV)". [How about sqrt{s} instead of E_cm?]

We use lower case "m" for mass elsewhere...

While you are at it, we could use a space before and after the "="

in the m_W equation on the plot.

Caption: remove the spurious spaces after "MeV/c^2" and "m_W".

p 11.0:

"Using" is bad because the subject of the sentence ("cross section")

isn't using.

Hyphen in "W-pair" for consistency.

p 11.1:

"New calculations with the double pole approximation (DPA)~[24] have

recently appeared, and two Monte Carlo programs, YFSWW~[25] and

RacoonWW~[26], are being developed."

two sentences later: no space between "0.5" and "%".

next sentence: "It is currently thought [believed] that the...".

No space between "2" and "%".

next sentence: "After adding" is bad. How about this?

"After inclusion of the data collected at centre-of-mass energies

of 161, 172, and 183 GeV, the hadronic decay branching ratio is

found to be blech%, [comma] which is used...".

Roman "cs" in that last equation.

p 11.7:

Need more vertical space and no indentation before "Global event

properties".

"Confidence level" instead of "probability".

p 12.0:

Space between "WW" and "kinematics".

Do you really want a roman "WW"?!

Last two \items: "smallest" instead of "minimum".

general comments:

1. It is nonstandard to give the titles of published papers. (It

is certainly not part of the journal's specifications.) And when

you put them in italics, you get the same problems I mentioned at

the beginning. For example, in [1] we have "e+e-" with one italic

"e" and one roman.

2. Use a backslash to prevent a big inter-sentence spacing after

abbreviated words: Phys.\ Lett.\ B453.

3. Be consistent: bold or not bold volume numbers?

specific comments:

[1] Delete the 2nd and 3rd occurrences of "ALEPH Collaboration"--it

isn't necessary to repeat the author(s) if they are the same for

consecutive papers within one \bibitem.

Semicolon instead of period after "435".

[5] "...(1992) 305; S.~Jadach..."

Period at the end.

[17] Nucl.\ Phys.\ (Proc.\ Suppl.)

[18] Capitalize "Group".

Unlike a preprint, the web page will not necessarily still say

"80.39 GeV" in one month, one year, or 100 years, and the reader

will not be able to find out how the number was obtained (e.g.,

which experiments). I think we should refer to a real document.

[20] Remove the extra space after "e+e-". This paper is unpublished,

so it is OK to leave in the title, in my opinion. But we need to

have two italic e's in "e+e-"!

Comma after the title, and don't necessarily capitalize the next

word.

Refer to a contributed paper, not to an abstract. And say which

conference!

[21] et al.,

[22] G.~Marchesini

[23] et al.\ (Particle

[25] Writing out four author names is pushing it. Five is going too

far! Use et al.

[26] Here we have the title in roman, with quotation marks. Be

consistent.

[27] Bohm instead of Bhm.

"...(1992) 29; B391 (1993) 483 (erratum)."