Dear all,

Thank you for sending your comments. I added the issues raised in the meeting. If you have any remaining comments, please send them to me before tomorrow, Tuesday 9 July.

Cheers, Veerle

Comments on "Observation of the decay Bs -> Dphi"

=== Title/Abstract: ===

As a byproduct of the analysis of the normalisation sample, you also give an measurement of the Bs->DbK*b/B->DbK*b ratio of branching fractions, but this is not reflected in the title. That update might therefore go by unnoticed, which would be a pity. Why not: Observation of the decay Bs->Dbphi and updated of the Bs->DbK*b branching fraction.

=== Introduction: ===

The current description of the physics case is misleading. As written, it is suggested that by simply collecting a larger sample of Bs->Dbphi one will be able to measure the CKM angle gamma unambiguously, which is not the case. In fact, the mode Bs->Dbphi itself is not sensitive to gamma. One needs the combined information from Bs->Dbphi and Bs->Dphi to extract gamma and in addition also Bs->D(CP)phi to get an unambiguous solution. We think the presentation of the physics case would benefit: - when showing Feynman diagrams - from a discussion of the modes involved and their differences - from a clear explanation of which modes you are and are not measuring, and why. [Are you distinguishing between the D and Dbar? Why is Bs->Dphi not measured?] As this is only a branching ratio measurement, some of the details regarding the determination of gamma [like time-integrated vs time-dependent] could be left out and/or replaced by references only.

L.29: eigenstates L.33: "sufficient number of different D final states" could you say how many?

=== Event Selection: === L. 69: "high sum of the pt of the tracks", could you specify the cut more precisely, or reference to the trigger document [12] which has the required information on the topo lines? L. 89: For the cut on the flight distance of B->Dbar K*, you probably corrected for the acceptance. How large was this correction? L.122: larger than 0.2 is irrelevant to the reader as no distributions are given. Please remove. L.124: which assumptions on the BF? Why did you not use the Punzi FoM that avoids the problem? Line 123: Why did you chose S/Sqrt(S+B) as the optimisation criteria. The Punzi FoM is better suited for optimising for first observations. Line 124: Please give the "assumptions on the decay branching fraction". Line 126: Please specify what you define as the signal region. Is it the whole mass range 5150 to 5600 MeV?

=== Signal Yield: === Figure 2: - Why are there no red data points between -0.4 and 0.3? If the cut at 0.4 is already applied, you would not expect to see an event at 0.3. - If bins with zero events where suppressed, why is there a partial red line in the -0.3 to -0.2 bin? - the difference between the red/blue is very difficult to distinguish in a black and white printout. Can the colors be changed?

L.133: You decide not to use the information of whether you have a DbarK* or a DbarK*bar final state and hence are loosing information. With that info you would have 2 mass plots with the same B, resp Bs signal but half the background. Why don't you do that?

Figure 1: - Bs->DbarK*bar (bar missing) - The fit for the partially reconstructed background in pink does not look very convinding. Maybe the RooKeysPdf tuning is too fine?

L.173: is -> are L.201 and later : ^{bkg} -> ^{\rm bkg}

Line 164: Change "Figure" to "Fig." Line 203: Does this result on fs/fd already contain the latest improvements from CLEO-c?

=== Systematic uncertainties: === L.219: contribution->'contributions' or 'the contribution' L.226: could you explain why you have a fit bias? L.237-242: you account for DCS decays but not for double mis-ID which give you a broader D mass. How are they treated? Line 237: It would be helpful to write (part of) the information in this paragraph already in the introduction.

=== Results and Conclusions: === - Line 246+9: Change reference [20] to [8] as the LHCb result is the only entry in the PDG.

- f_s/f_d should be in math mode. Are you using the latest error with the new CLEO measurement?

- "... compared to a previous measurement [20]" : If it's one measurement cite it and not the PDG.

=== References: === Line 266: The arxiv link is broken. Remove "hep-ph/"