{"id":"75120e8d-707f-4a66-9bce-9514ef5236fe","arxiv_id":"2610.41395","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"An absolute proper-motion map of about 400,000 bulge stars from 27 HST fields shows dispersion increasing toward the Galactic center and asymmetric across longitude, consistent with a barred bulge.","lead":"Using archival Hubble Space Telescope images of 27 fields, the authors measured how about 400,000 stars in the southern Galactic bulge move across the sky. The resulting map shows motions becoming more disordered toward the center and differing between the two sides, pointing to the Milky Way's bar-shaped inner structure.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek/deepseek-v4-flash","headline":"Disc contamination is not removed in the fits that establish the l<0/l>0 dispersion and covariance asymmetry; the ~4 mas/yr disc component seen in Section 4.4 can create both the steeper l<0 gradient and C_lb ~ 0 on that side, so the bar-asymmetry claim is not yet separated from it.","rationale":"The reader identified the same load-bearing assumption: the observed l-asymmetry in dispersions and covariance must be intrinsic to the bulge, not produced by spatially varying foreground disc contamination. My reading of the paper supports that this is the decisive unresolved step. The measurement pipeline itself has independent support: the absolute calibration comparison with Gaia DR3 and OGLE (Fig. 1), the mu_b agreement with Sgr A*, and the l>0 covariance being consistent with Kozlowski et al. (2006) and Clarke et al. (2019) all suggest the catalogue is not fabricated and the basic trends are plausible. However, the headline asymmetry is computed from an all-star sample, while the paper's own non-Gaussianity analysis shows that this sample contains a disc component with a large mu_l offset. Because the l<0 and l>0 field sets are small and not matched in b or extinction, a side-dependent disc fraction can mimic both the steeper l<0 dispersion slope and the near-zero l<0 covariance. The bootstrap-only error bars make the quoted significance look stronger than the systematic uncertainty justifies. The reader's CONDITIONAL verdict already captures this, so I recommend no verdict change. The concrete contamination-corrected refit described above is the one check that would settle whether the asymmetry is dynamical or a selection artifact.","tokens_in":24202,"tokens_out":12000,"duration_ms":83746,"concrete_test":"Refit Eqs. (8)-(12) after correcting for foreground disc contamination: apply the Section 4.5 CMD-based bulge selection (removing the blue upper-left disc branch) in the 16 fields with F606W, and for all 27 fields use a Besancon or TRILEGAL synthetic disc model to predict and subtract the disc contribution to sigma_l, sigma_b and C_lb. If the l<0/l>0 slope ratios move toward 1 or the C_lb difference disappears, the asymmetry is a contamination artifact; if the slopes and covariance difference survive the correction, the concern is retired.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—a steeper dispersion gradient for l<0 and opposite C_lb signs on the two sides—rests on Sections 4.2 and 4.3, where all stars with 18.5 <= m_F814W <= 23 are used with no CMD-based disc removal. The paper itself shows in Section 4.4 (Fig. 7) that foreground disc stars form a second component shifted by about 4 mas/yr in mu_l, and the abstract attributes the non-Gaussianity mainly to this disc contamination. A second component with a mean offset near 4.8 mas/yr contributes w(1-w)(Delta_mu)^2 to sigma_l^2 and, if its intrinsic C_lb is small, dilutes the bulge covariance toward zero. The l<0 sample has only six fields and is not matched to the l>0 sample in b or extinction, so the disc fraction is free to differ between the two sides; a higher disc fraction on l<0 would produce both a steeper sigma_l gradient and C_lb close to zero without any intrinsic bar asymmetry. The bootstrap errors quoted in Eqs. (8)-(12) are internal sampling errors only and do not include this contamination systematic. In addition, the field with the largest sigma_l is excluded from the sigma_l fits (Fig. 5 caption) with no stated reason, and if that field is on one side it changes the slope ratios. Section 4.5 applies CMD cleaning only to the 16 fields with F606W and fields 26 and 27 are not extinction-corrected; that cleaning is not used in the fits that establish the headline asymmetry.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek/deepseek-v4-flash","summary":"The paper compiles archival HST/WFC3-UVIS data in 27 southern bulge fields (l in [-6, 4] deg, b in [-5, -1] deg) with 3--13 yr baselines, calibrates absolute proper motions against Gaia DR3 (with an OGLE/Uranus cross-check), and delivers ~400,000 stars at ~0.1 mas/yr precision for the brightest sources. The analysis computes mean proper motions, dispersions, anisotropy, and the covariance C_lb. The mean mu_b agrees with the VLBA Sgr A* value, and mu_l agrees with Sgr A* after extrapolating the b-trend to the plane. The headline results are a dispersion gradient that steepens toward the centre, with a steeper slope on the l<0 side (slope ratios r_sigma_l = 0.70 +/- 0.09, r_sigma_b = 0.75 +/- 0.07), anisotropy sigma_l/sigma_b = 1.131 +/- 0.005, and C_lb ~ -0.085 on the l>0 side versus ~0.004 on the l<0 side. The authors attribute the asymmetry to the non-axisymmetric bar. The central statistics use the all-star sample 18.5 <= m_F814W <= 23; CMD-based disc removal and extinction correction are applied only in Section 4.5, and only for the 16 fields with F606W data.","tokens_in":24581,"tokens_out":5656,"duration_ms":36880,"significance":"If the asymmetry is intrinsic, this is a valuable dataset and result: it extends bulge proper-motion studies from a handful of fields to 27 sightlines, at precision adequate to resolve bulge internal kinematics, and the claim of an l<0/l>0 dispersion asymmetry would be a direct probe of the bar geometry. The catalogue construction is careful in several respects worth crediting: the absolute frame is anchored externally (Gaia DR3) and cross-checked independently (OGLE/Uranus); the Sgr A* comparison uses an independent VLBA measurement; the RC reference values (m_RC = 16, C_RC = 1) are conventions rather than quantities fitted to the kinematics; and the bootstrap convergence is explicitly tested. The central physical claim, however, is not yet separated from the foreground-disc contamination that the paper itself documents in Section 4.4, and the quoted errors are internal sampling errors only. The result is therefore conditional on a contamination-controlled re-analysis rather than ready as stated.","major_comments":[{"comment":"The l-asymmetry that the abstract advertises is established on the all-star sample (18.5 <= m_F814W <= 23, Section 3.1) with no CMD-based disc removal; Section 4.5 applies disc removal and extinction correction only to the 16 fields with F606W data, and neither is used in the fits behind Eqs. (8)-(12). The l<0 side contains just six fields (05, 06, 07, 09, 10, 22) that are not matched to the l>0 side in b or extinction. Section 4.4/Fig. 7 shows a second component offset by roughly 4.8 mas/yr in mu_l from the bulge peak; such a component contributes w(1-w)(Delta_mu)^2 to sigma_l^2 (of order 15 per cent for a 10 per cent disc fraction at sigma_l ~ 3 mas/yr) and, having small intrinsic C_lb, dilutes the observed |C_lb| toward zero. A side-dependent disc fraction would then produce both the steeper l<0 sigma_l gradient and the C_lb ~ 0 value without any intrinsic bar asymmetry, and the bootstrap errors in Eqs. (8)-(12) do not include this systematic. Please recompute the asymmetry statistics on the CMD-cleaned 16-field subset, or quantify the disc fraction per side and propagate a contamination model into sigma_l and C_lb.","section":"Sections 4.2, 4.3; Eqs. (8a)-(9b), (12); Fig. 5","authors_acknowledge":"This discrepancy is likely because our mean proper motions are averaged over all stars with18.5< 𝑚F814W <23in the 27 fields, so spatial offsets and contamination by disc stars can affect our results.","authors_acknowledge_p":0.65},{"comment":"The field with the largest sigma_l is excluded from the sigma_l fits with no stated criterion. Since the slope ratio r_sigma_l = 0.70 +/- 0.09 is one of the two numbers carrying the bar-asymmetry conclusion, the fit must be reported both with and without that field, and the exclusion criterion stated explicitly; if the excluded field falls on one side, its removal directly changes the slope ratio.","section":"Fig. 5 caption; Eqs. (8a)-(9b)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The covariance asymmetry (C_lb = -0.085 +/- 0.003 for l>0 versus 0.004 +/- 0.011 for l<0) is presented as part of the bar-asymmetry evidence, but the l<0 value is exactly where the disc-dilution effect lands: a mixture of bulge stars (C_lb ~ -0.1) with a modest disc fraction of near-zero covariance reproduces both numbers. The authors do flag the limited l<0 data in the paragraph following Eq. (12), but the abstract and conclusions do not carry that condition. Quantify the disc fraction in the l<0 fields (e.g., from the double-Gaussian fits of Section 4.4) and show the contamination-corrected C_lb rather than the raw value.","section":"Section 4.3, Eq. (12); Fig. 6","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The OGLE/Uranus catalogue is cited as 'in preparation' and the statement that OGLE gives smaller calibration errors in large fields but larger in small fields rests on that unpublished source; since Gaia DR3 is adopted this is not blocking, but the claim should be marked as preliminary.","section":"Section 2.3, Fig. 1","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The caption places the purple solid line and shaded region description only in the last sentence; indicate in each panel which symbol/line corresponds to which statistic.","section":"Fig. 4","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The catalogue underlying the main results is offered 'on reasonable request' rather than archived; given the paper's primary deliverable is this catalogue, a public archive (e.g., MAST or Zenodo) would better support reproducibility.","section":"Data Availability","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The astrometric error model assumes gamma = 0.6 and the text says tests 'describe the distribution well'; state what was tested and whether a single gamma is adopted for all SNR rather than fitted.","section":"Section 2.4, Eq. (1)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Minor text issues: 'Koz owski' appears in the Fig. 5 and Fig. 6 captions, and the 3-sigma clipping of Section 3.1 interacts with the outlier-field exclusion in Fig. 5 in a way worth a sentence of clarification.","section":"Section 3.1, Figs. 5-6 captions","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"major_revision","confidential_remarks":"The catalogue itself is likely sound and publishable as a data product; the issue is that the paper's central physical conclusion (l<0/l>0 asymmetry from the bar) is drawn from the same all-star sample whose contamination the paper itself characterises in Section 4.4. That is fixable within the manuscript's scope: recompute Eqs. (8)-(12) on the 16-field CMD-cleaned subsample of Section 4.5, or provide an explicit contamination model with side-dependent disc fractions. I would also note that the OGLE/Uranus dependence on an unpublished catalogue is acceptable only because Gaia DR3 is the adopted reference."},"author_rebuttal":null,"desk_editor":{"model":"deepseek/deepseek-v4-flash","letter":"Two things to know. First, this is a real data paper: absolute proper motions for ~400,000 stars in 27 WFC3/UVIS fields across the southern bulge, anchored to Gaia DR3, reaching sub-0.1 mas/yr for brighter stars. The reduction uses the mature hst1pass pipeline with CTE-corrected _flc data, and the absolute calibration is cross-checked against both Gaia DR3 and OGLE/Uranus. The mean-motion checks pass: mu_b matches Sgr A*, and mu_l extrapolates to the Reid & Brunthaler value at b=0. If the catalogue becomes public, it will be a useful resource, especially at magnitudes where Gaia cannot resolve individual stars.\n\nSecond, the interpretive claims outrun the analysis. The headline results — a steeper dispersion gradient on the l<0 side (slope ratio 0.70 ± 0.09) and C_lb near zero on l<0 versus -0.085 on l>0 — are computed from all stars with 18.5 <= m_F814W <= 23, with no CMD-based disc removal. Only 16 of 27 fields have F606W data; Section 4.5 does the cleaning but only for that subset, and none of it feeds the fits that establish the asymmetry. This is not pedantry. The paper itself shows (Fig. 7) a foreground disc component shifted by ~4-5 mas/yr in mu_l. A second component with that offset adds w(1-w)(Delta_mu)^2 to sigma_l^2, of order 1-2 (mas/yr)^2 — 10-20% of the observed variance — and a disc population with small intrinsic C_lb dilutes the covariance toward zero. If the disc fraction is higher on the l<0 side, which the paper never checks, both halves of the asymmetry follow without any intrinsic bar signal. The l<0 sample is at most seven fields, the field with the largest sigma_l is dropped from the fits with no stated reason, and the sides are not matched in b or extinction. The bootstrap errors in Eqs. (8)-(12) cover sampling only, not contamination or the ~0.1 mas/yr absolute calibration term. And only three of the l<0 fields have colour data, so the CMD-clean check is weakest exactly where the new claim lives.\n\nThe paper is not careless — it flags disc contamination for the mean motions and the non-Gaussianity, and the authors plainly know how to remove disc stars — but that control never reaches the asymmetry fits. Minor: catalogue on request, no pipeline code, and Fields 26 and 27 lack extinction correction (stated).\n\nWho it is for: bulge-kinematics people who will use the catalogue and should treat the l-asymmetry as a hypothesis, not an established measurement. It deserves a serious referee; the main request should be the asymmetry fits on the CMD-cleaned subset with per-field disc accounting, plus a justification for the outlier cut.","headline":"A genuinely useful new HST proper-motion catalogue whose headline bar-asymmetry claims still need the disc-contamination control the authors have in hand but did not apply to the main fits.","tokens_in":25262,"tokens_out":9566,"would_cite":true,"duration_ms":62496,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek/deepseek-v4-flash","headline":"HST proper motions of about 400,000 stars in 27 southern bulge fields show velocity dispersions rising toward the Galactic centre with a steeper gradient on the $l<0^\\circ$ side, a kinematic signature of the bar's non-axisymmetric shape.","keywords":["Galactic bulge","proper motions","Milky Way bar","HST WFC3 astrometry","Gaia DR3 reference frame","velocity dispersion anisotropy","red clump stars","disc contamination"],"falsifier":"Recompute the longitude gradients of $\\sigma_l$ and $\\sigma_b$ using only the 16 CMD-cleaned fields, or after removing the disc subpopulation in every field via the double-Gaussian decomposition the paper itself uses; if the slope ratio between the $l>0^\\circ$ and $l<0^\\circ$ sides returns to within error of 1, the bar-asymmetry claim loses its support. A complementary test is to repeat the dispersion fits with additional negative-longitude fields reduced through the same pipeline; the claim stands only if the steeper $l<0^\\circ$ gradient survives the added sightlines.","tokens_in":25601,"feed_emoji":"🌌","tokens_out":14130,"duration_ms":105331,"temperature":0.7,"pith_summary":"The paper tries to settle what the transverse motions of bulge stars look like across the whole southern bulge, not just in a few windows. From archival HST WFC3/UVIS images with 3–13 year baselines, anchored to Gaia DR3 for absolute calibration, it measures proper motions for about 400,000 stars in 27 fields at roughly 0.1 mas/yr precision for the brighter stars. The central finding is that proper-motion dispersions rise toward the Galactic centre on both sides, but the gradient is significantly steeper on the $l<0^\\circ$ side, with slope ratios between the two sides of $0.70\\pm0.09$ in $\\sigma_l$ and $0.75\\pm0.07$ in $\\sigma_b$, which the authors read as the kinematic imprint of the bar's non-axisymmetric shape. Supporting evidence includes a mean anisotropy $\\sigma_l/\\sigma_b = 1.131\\pm0.005$ and a normalised covariance that is negative on the $l>0^\\circ$ side but essentially zero on the other. If the interpretation holds, the bulge is a bar-driven pseudo-bulge whose velocity-dispersion field is measurably lopsided, giving dynamical models a new, bulge-wide constraint.","feed_headline":"Lopsided star motions reveal the Milky Way's bar","feed_subtitle":"HST proper motions of 400,000 stars show random velocities rising faster on one side of the bulge: the bar's fingerprint.","key_machinery":"The machinery is a set of absolute proper motions built from two epochs of HST WFC3/UVIS F814W images per field, separated by 3 to 13 years and reduced with the hst1pass software, which performs PSF photometry and astrometry, corrects geometric distortion and the telescope's 'breathing' focus changes, and uses CTE-corrected _flc data. Gaia DR3 stars are cross-matched to place the relative HST motions on an absolute frame, with the OGLE/Uranus catalogue used as a cross-check on calibration errors. From these measurements the paper computes, per field, the mean motions $\\langle\\mu_l\\rangle$ and $\\langle\\mu_b\\rangle$, the intrinsic dispersions $\\sigma_l$ and $\\sigma_b$ with measurement errors subtracted in quadrature following the Jones (1970) formula, the anisotropy ratio $\\sigma_l/\\sigma_b$, and the normalised covariance $C_{lb}$, with bootstrap resampling (3,000 steps) for the errors. In the 16 fields with F606W data, colour–magnitude diagrams separate bulge red-clump stars (core-helium-burning giants used as standard candles) and red giants from foreground disc main-sequence stars; fits to the red-clump magnitude and colour provide extinction and reddening corrections, and double-Gaussian fits to the proper-motion distributions quantify the disc subpopulation that causes the observed non-Gaussianity.","core_discovery":"On its own terms, the paper claims that the bulge's proper-motion dispersion field is asymmetric about the longitude of the Galactic centre. Linear fits of $\\sigma_l$ and $\\sigma_b$ against longitude, done separately for the two sides, give slopes whose ratios between $l>0^\\circ$ and $l<0^\\circ$ are $0.70\\pm0.09$ and $0.75\\pm0.07$: the dispersions climb more steeply on the negative-longitude side, a pattern the paper attributes to non-axisymmetric structure such as the bar. Two further statistics reinforce the claim: the velocity distribution is anisotropic, with $\\sigma_l/\\sigma_b = 1.131 \\pm 0.005$ and all fields above 1.0, and the normalised covariance between the two proper-motion components is negative for $l>0^\\circ$ ($-0.085\\pm0.003$) while essentially zero for $l<0^\\circ$ ($0.004\\pm0.011$). The paper also establishes that the motion of the bulge centre itself is reproduced: the mean proper motion in latitude matches that of the central black hole Sgr A*, and the longitude mean, extrapolated from its observed latitude trend to the plane, gives $-6.383\\pm0.065$ mas/yr against Sgr A*'s $-6.379\\pm0.026$ mas/yr. A caveat the paper itself states is that the covariance trend with latitude rests on limited $l<0^\\circ$ data and needs further verification.","pith_inferences":["The steeper dispersion gradient on the $l<0^\\circ$ side is plausibly a near-side/far-side projection effect of a rotating bar: stars on the side closer to the Sun sample the bar's gravitational asymmetry differently, so the slope-ratio amplitude could in principle be converted into a constraint on the bar's viewing angle and pattern speed.","A testable extension the paper does not run is to push the same two-epoch pipeline through the existing negative-longitude HST fields from the literature, so the two-sided comparison rests on more than the seven fields at $l<0^\\circ$ used here.","The red-clump calibration assumes the same intrinsic red-clump magnitude and colour in every field; if the fields differ systematically in metallicity along the bar, the extinction-corrected magnitude and colour bins of Section 4.5 could inherit a small bias that a metallicity-aware calibration would remove."],"forward_implications":["If the asymmetry is real, the bulge's velocity-dispersion field is measurably lopsided, and the slope ratio between the two longitude sides becomes a quantitative target that bar models must reproduce rather than a free parameter.","The agreement between the extrapolated $\\langle\\mu_l\\rangle$ and Sgr A*'s proper motion strengthens the practice of treating Sgr A* as the dynamical anchor of the bulge centre in kinematic studies.","The quantified double-Gaussian disc component, offset by about 4 mas/yr in $\\mu_l$ and concentrated in the brighter stars, gives future surveys a template for subtracting foreground contamination even where colour information is missing.","Matching the roughly 400,000-star absolute proper-motion catalogue against radial-velocity surveys would deliver a three-dimensional velocity-ellipsoid map of the bulge, extending the two-dimensional kinematics presented here.","The sign flip of $C_{lb}$ between the two longitude sides, together with its opposite latitude trends, implies the velocity ellipsoid is tilted and oriented differently on the two sides of the bar, a property future dynamical models can test directly."],"supporting_citations":[{"why":"Supplies the Gaia DR3 absolute proper-motion reference frame that places the relative HST motions on an absolute system.","marker":"Gaia Collaboration et al. (2016, 2023)"},{"why":"The hst1pass software performs the source detection, PSF photometry, and astrometry for every field, including distortion and breathing corrections.","marker":"Anderson (2022)"},{"why":"Provides the empirical CTE-loss correction method; its corrected products, the _flc data, are the basis of the astrometric measurements.","marker":"Anderson & Bedin (2010)"},{"why":"Adapts the CTE correction to the WFC3/UVIS camera, where CTE loss is severe at the low background levels of these images.","marker":"Anderson et al. (2021)"},{"why":"The VLBA proper motion of Sgr A* is the reference value against which the mean motions and the plane extrapolation are checked.","marker":"Reid & Brunthaler (2004)"},{"why":"The prior HST bulge proper-motion survey whose anisotropy ratio, covariance near -0.1, and astrometric error formula this study extends and compares with.","marker":"Kozłowski et al. (2006)"},{"why":"Supplies the formula used to subtract measurement errors in quadrature and obtain intrinsic dispersions.","marker":"Jones (1970)"},{"why":"The red-clump fitting method used to derive extinction and reddening corrections in the fields with colour data.","marker":"Nataf et al. (2013)"},{"why":"The SWEEPS double-Gaussian decomposition of disc and bulge proper-motion populations is the template for the non-Gaussianity analysis.","marker":"Clarkson et al. (2008)"},{"why":"VIRAC proper-motion maps provide the independent anisotropy and quadrupolar covariance patterns with which the measured values are compared.","marker":"Clarke et al. (2019)"}],"fun_headline_variants":["Bulge star motions show the bar's lopsided pull","400,000 stars trace the Milky Way's barred bulge","HST proper motions expose the bulge's non-axisymmetric bar","Star velocity maps reveal the bar's asymmetric footprint","Bulge dynamics from HST proper motions of 400,000 stars"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The load-bearing assumption is that the asymmetry in dispersions and covariance between the two longitude sides is intrinsic to the bulge, not an artifact of foreground disc contamination that differs between the sides: the headline statistics are computed from all stars with $18.5 \\le m_{\\rm F814W} \\le 23$ without colour-based disc removal, and only 16 of the 27 fields have the colour information needed to test that assumption directly.","fun_headline_variants_meta":{"raw":{"variants":["Bulge star motions show the bar's lopsided pull","400,000 stars trace the Milky Way's barred bulge","HST proper motions expose the bulge's non-axisymmetric bar","Star velocity maps reveal the bar's asymmetric footprint","Bulge dynamics from HST proper motions of 400,000 stars"]},"model":"deepseek/deepseek-v4-flash","effort":"low","cost_usd":8.7e-05,"raw_usage":{"cost":8.73782e-05,"is_byok":false,"cost_details":{"upstream_inference_cost":8.73782e-05,"upstream_inference_prompt_cost":3.84202e-05,"upstream_inference_completions_cost":4.8958e-05},"total_tokens":1556,"prompt_tokens":1287,"completion_tokens":269,"prompt_tokens_details":{"audio_tokens":0,"video_tokens":0,"cached_tokens":1081,"cache_write_tokens":0},"completion_tokens_details":{"audio_tokens":0,"image_tokens":0,"reasoning_tokens":198}},"tokens_in":1287,"tokens_out":269,"cost_source":"openrouter_usage","duration_ms":8898,"image_pages":[],"temperature":1.0,"billed_attempts":1,"reasoning_tokens":198,"transport_provider":"openrouter","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-11-02T15:16:39.256533+00:00","model_set":{"reader":"deepseek/deepseek-v4-flash"},"falsifier":"Recompute the longitude gradients of $\\sigma_l$ and $\\sigma_b$ using only the 16 CMD-cleaned fields, or after removing the disc subpopulation in every field via the double-Gaussian decomposition the paper itself uses; if the slope ratio between the $l>0^\\circ$ and $l<0^\\circ$ sides returns to within error of 1, the bar-asymmetry claim loses its support. A complementary test is to repeat the dispersion fits with additional negative-longitude fields reduced through the same pipeline; the claim stands only if the steeper $l<0^\\circ$ gradient survives the added sightlines.","supporting_citations":[],"review_version":1}