{"id":"c7c6ed9a-8674-4728-895b-16182c1edfb5","arxiv_id":"gr-qc/0005050","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Generalizing the Schwarzschild photon sphere, the paper defines photon surfaces in arbitrary spacetimes and proves that black holes in static spherical spacetimes must be surrounded by one.","lead":"This paper defines a photon surface in an arbitrary spacetime and proves that, under an energy condition, any black hole in a static spherically symmetric spacetime must be surrounded by one. It gives relativists a general geometric language for light-trapping surfaces, extending the Schwarzschild photon sphere to a wide class of metrics.","discovery_kind":"extension","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"No significant objection identified: the central claims are correct as conditional statements, and the acknowledged limitations are the main constraints on scope.","rationale":"The reader's verdict of ACCEPT with HIGH confidence is well-supported. The paper's central claims are correct as conditional statements, and the limitations are acknowledged. The reader's weakest_assumption correctly identifies the energy condition as the key hypothesis, but this is a stated assumption rather than a flaw: the theorem is conditional, and the paper explicitly notes that the converse results need stronger matter bounds. The only technical issue I found is a minor proof gap in Theorem 4.3 (the parallel-propagation claim), which does not affect the theorem's conclusion. The reader's rationale correctly notes that the Fourier decomposition and sum-of-squares arguments check out, and that the IVT arguments are valid given the stated energy conditions. I agree with the verdict.","tokens_in":21666,"tokens_out":22823,"duration_ms":419805,"concrete_test":"Independently verify the MOTS ⇒ lim(1-2m/r)=0 step in Theorem 4.3 by computing Θ_out for the outgoing null normal in a general static spherically symmetric metric (59) and checking that a(r)k^t = 1-2m/r identically under the normalization g(k,n)=1; this bypasses the parallel-propagation claim and confirms the theorem's conclusion.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claims — Definition 2.1/Theorem 2.2 (photon surface iff pure-trace second fundamental form), Theorem 3.1 (area evolution), Theorem 4.1/4.2 (static characterizations), Theorem 4.3 (black hole implies photon sphere), and Proposition 4.4/Theorem 4.5 (converse) — are correct as stated. The only technical issue I found is a minor proof gap in Theorem 4.3: the proof asserts that the outward null normal k is parallelly propagated along the radial geodesics generated by n, but in Schwarzschild ∇_n k^t = m/(r²(1-2m/r)) ≠ 0, so this is false. However, the conclusion lim a(r) = 0 follows directly from Θ_out = 2a k^t/r and a k^t = 1-2m/r, so the theorem stands. The acknowledged limitations (converse results need 4m<r or 24m/7<r rather than 2m<r; spherical symmetry assumed; Kerr requires separate treatment) are correctly stated and do not undermine the central claims.","agreement_with_reader":"partial"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"The paper generalizes the Schwarzschild photon-sphere concept to a coordinate-free notion of a photon surface: an immersed, nowhere-spacelike hypersurface S such that every null tangent vector at every point of S extends to a null geodesic of the ambient space-time that remains in S (Definition 2.1). Theorem 2.2 establishes the central equivalence: for a timelike S, being a photon surface is equivalent to the second fundamental form being pure trace (sigma_ab = 0), equivalently to every affine null geodesic of the induced metric being an affine null geodesic of the ambient metric. The authors then specialize to SO(3)-invariant surfaces in spherically symmetric space-times, deriving an evolution equation for the cross-sectional area (Theorem 3.1) and a first-order locating equation (Theorem 4.2, Eq. (42), reducing to Eq. (54) in static coordinates). Theorem 4.1 gives a static characterization in terms of A(Theta^2 - 3 G_ab n^a n^b) = 12 pi. Theorem 4.3 proves, subject to non-negative density and radial pressure and a marginally outer trapped surface, that a static spherically symmetric metric of the form (59) admits a timelike photon sphere; Proposition 4.4 and Theorem 4.5 provide partial converses under stronger matter bounds. Section 5 works through Schwarzschild, Schwarzschild-de Sitter, the Schwarzschild interior, Reissner-Nordstrom, Janis-Newman-Winicour, charged dilaton, and Vaidya null-dust collapse, the last including a numerical backward evolution of the Schwarzschild","tokens_in":21809,"tokens_out":11728,"duration_ms":224372,"significance":"This is a foundational, definitional paper, and its value lies in the generality and cleanliness of the framework rather than in numerical agreement with data. The definition of a photon surface is conformally invariant and coordinate-free, and Theorem 2.2 converts a statement about families of null geodesics into a local algebraic condition on the second fundamental form, which is exactly the kind of reduction that makes the concept usable in arbitrary space-times. The static locating equation (54) is first order in the metric and reproduces the known Schwarzschild result r = 3m; the black-hole-must-be-surrounded-by-a-photon-sphere theorem (4.3) and the no-go converses (4.4, 4.5) are stated with explicit hypotheses and are proved rather than asserted. There are no free parameters, no fitting, and no tuned quantities: every result is a conditional statement derived from the stated energy conditions and the Einstein equations. The worked examples are genuinely informative, particularly the Reissner-Nordstrom and charged-dilaton cases, where the existence or non-existence of photon spheres tracks the black-hole/naked-singularity transition, and the Vaidya evolution, which shows the","major_comments":[{"comment":"The proof contains an incorrect intermediate assertion. It states that the outward null normal k is 'parallelly propagated along each of the geodesic integral curves of n', and uses this to conclude that lim k is a well-defined, nowhere-zero null field along T_t. This is false in general: in Schwarzschild, with k^t = 1 and k^r = (1-2m/r)^{1/2}, one has (nabla_n k)^t = Gamma^t_{tr} k^r n^r = m/r^2, which does not vanish. The conclusion of the theorem is not endangered, but the argument needs to be restated. With the stated normalization g(k,n) = 1 and n^r = (1-2m/r)^{1/2}, one obtains a(r) k^t = (1-2m/r)^{1/2}, so that Theta_out = 2 a(r) k^t / r = 2(1-2m/r)^{1/2}/r. The condition that T_t be marginally outer trapped then gives lim_{r downarrow r0} (1-2m/r) = 0 directly, which is the first of (71) and is all the rest of the proof requires. The authors should replace the parallel-propagation sentence with this explicit computation and state the normalization of k at the point where it is used.","section":"Section 4, proof of Theorem 4.3","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"As typeset, the first term on the right-hand side of Eq. (17) and of Eq. (25) is dimensionally inconsistent: (1/4) (2)A_s (d/ds (2)A_s)^2 has dimensions of length to the fifth, whereas d^2/ds^2 (2)A_s has dimensions of length. The correct expression, obtained by substituting (2)Theta = d/ds ln (2)A_s into Eq. (24), is (1/4) ((2)A_s)^{-1} (d/ds (2)A_s)^2, i.e. the area carries an inverse power. The derivation in Eqs. (18)-(24) is correct and the fix is a single character, but since Eq. (17) is the displayed statement of the paper's second main result it should be corrected, and the same correction applied to Eq. (25).","section":"Section 3, Theorem 3.1, Eqs. (17) and (25)","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The abstract states that 'any photon sphere must surround a black hole, a naked singularity or more than a certain amount of matter'. The theorems actually proved (Proposition 4.4 and Theorem 4.5) are no-go statements: under lim_{r to 0} m(r) = 0, p <= rho/3, and either 4m(r) < r or (24/7)m(r) < r, no timelike photon sphere exists. The 'naked singularity' branch is not established by these results, and the paper's own Examples 7 and 9 exhibit naked singularities both with and without photon spheres. The abstract should be softened to match what is proved, for instance by saying that a photon sphere forces either a black hole or a violation of the stated matter bounds.","section":"Abstract, sentence [abstract:s3]","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Eq. (93) is typeset as r = b(2 nu + 1)^2. Recomputing from Eq. (54) with g_tt = -(1-b/r)^nu and g_{theta theta} = r^2 (1-b/r)^{1-nu} gives r = b(2 nu + 1)/2, not the square. The stated existence range 1/2 < nu <= 1 is consistent with the linear form, so this appears to be a typesetting error that should be corrected.","section":"Section 5, Example 8, Eq. (93)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The text notes that Eq. (54) 'agrees with an equation obtained by Virbhadra & Ellis [4] on the basis of a different definition'. Reference [4] is listed as 'Phys. Rev. D. To be submitted', i.e. unpublished. Since Eq. (54) is derived here independently from Theorem 4.2, the authors should either update the citation status of [4] or state explicitly that the present derivation is independent of that work, so that the priority of the result is unambiguous.","section":"Section 4, Case 1 and reference [4]","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The sentence 'For 0 <= (e/m)^2 <= 1 the curvature singularity at r = 0 is locally naked but hidden behind an event horizon' is internally confusing, since a singularity hidden behind an event horizon is not normally called naked. Presumably the intended meaning is that the singularity is visible to observers inside the Cauchy horizon while being hidden from infinity. This should be rephrased.","section":"Section 5, Example 7","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"In the step beginning 'From the boundedness of rho(r) and p1(r) on r in (0, infinity)', the hypothesis actually stated in condition (1) is boundedness on (r0, infinity). The argument only needs boundedness on a neighbourhood of r0, so the conclusion is unaffected, but the reference to the hypothesis should match condition (1).","section":"Section 4, proof of Theorem 4.3","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The observation that the Schwarzschild-de Sitter photon sphere sits at r = 3m independently of Lambda is correct and is a nice check, but the remark that this is 'surprising since Schwarzschild and Schwarzschild-de Sitter space-times are not conformally related' would be strengthened by the underlying reason: for any metric with g_tt g_rr = -1 the photon-sphere equation (54) reduces to r = 3m(r) and the tt-component drops out entirely. One sentence to this effect would turn an isolated curiosity into an explanation.","section":"Section 5, Example 5","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"In the version supplied for review, Figures 2, 3 and 5 render as unreadable glyph strings rather than legible plots, so the numerical content of the Reissner-Nordstrom, charged-dilaton and Vaidya diagrams could not be checked against the surrounding text. The authors should confirm that the figure files are correct at proof stage.","section":"Figures 2, 3 and 5","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The manuscript contains pervasive spacing artifacts in the running text (for example 'surf aces', 'spherically symm etric', 'equatio ns', 't he'). These are presumably source-conversion artifacts rather than authorial errors, but they should be cleaned up before the final version, since they make several equations and sentences harder to read than necessary.","section":"Throughout","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"minor_revision","confidential_remarks":"This is a careful and, in my view, correct foundational paper whose definitions and theorems are likely to be used well beyond the examples given. The two items I have listed as major are both local and repairable: an incorrect intermediate sentence in the proof of Theorem 4.3 whose conclusion survives an explicit recomputation, and a missing inverse power in the displayed area-evolution equation. Neither changes any result. The one matter I would ask the editor to watch is the citation of reference [4] as 'to be submitted' for Eq. (54), which is also derived independently in this manuscript; the authors should settle the attribution before publication."},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"Worth your time. The real contribution is Definition 2.1 — a photon surface is a nowhere-spacelike hypersurface on which every null tangent vector extends to a null geodesic that stays on the surface — together with Theorem 2.2, which shows this is equivalent to the second fundamental form being pure trace, sigma_ab = 0, and to affine null geodesics of the induced metric being affine null geodesics of the ambient one. That equivalence is what makes the concept usable, and it is genuinely new relative to [3,4], which defined a photon sphere only through the lensing deflection angle. The Fourier argument in (ii) implies (iii) is correct, and the sum-of-squares in (26), with sigma_10 = sigma_20 = 0 from spherical symmetry, does force sigma_ab = 0. Theorem 3.1's area evolution equation is a tidy piece of work as well. The static half is a careful specialization: photon sphere as SO(3) x R-invariant photon surface, equation (54) as the locating equation, and Theorem 4.3 (a black hole implies a photon sphere, given rho, p1 >= 0 and a marginally outer trapped surface). The examples are the best part — Reissner-Nordstrom, JNW, the charged dilaton, and the Vaidya collapse where the evolved surface encloses the naked singularity. The Schwarzschild-de Sitter result that r = 3m regardless of Lambda is correct and worth flagging to students. Soft spots, in proportion. There is a small proof gap in Theorem 4.3: the proof says k is parallelly propagated along the integral curves of n, which is false — in Schwarzschild the time component of k does not stay constant along those curves. The conclusion still follows from Theta_out = 2a k^t / r with a k^t = 1 - 2m/r, so it is a one-line fix, not a hole. The converse results (Proposition 4.4, Theorem 4.5) need 4m < r or 24m/7 < r rather than 2m < r; the authors say so themselves and admit no counterexample is known. Reference [4] is listed as to be submitted, but they re-derive (54) independently, so nothing rests on it. Who it is for: anyone working on lensing, black hole shadows, or null geodesic structure. Send it out.","headline":"A clean geometric definition of the photon surface, with Theorem 2.2's four-way equivalence doing the real work; the static half is careful specialization with honestly stated limits.","tokens_in":22539,"tokens_out":1899,"would_cite":true,"duration_ms":31221,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":["83C57","83C10","53C50"],"pacs":["04.70.Bw","98.62.Sb","04.20.Cv"],"model":"selfhosted/deepseek-v4.1-flash","headline":"A photon sphere is not a Schwarzschild accident: it is forced on any static, spherically symmetric black hole by non-negative density and pressure, and it can be defined in any spacetime at all.","keywords":["photon sphere","photon surface","null geodesics","gravitational lensing","static spherically symmetric spacetime","second fundamental form","black hole","Vaidya collapse"],"falsifier":"Find an asymptotically flat static spherically symmetric metric with $2m(r)<r$ everywhere, non-negative bounded $ρ$ and $p_1$, and a marginally outer trapped inner boundary, for which $1-3m(r)/r-4πr^2p_1(r)$ never vanishes; such a solution would break Theorem 4.3.","tokens_in":22931,"feed_emoji":"🕳️","tokens_out":3292,"duration_ms":67444,"temperature":0.7,"pith_summary":"The paper asks what a photon sphere really is, once you leave Schwarzschild behind. Its answer is a definition that mentions no symmetry: a photon surface is a surface such that every light ray launched tangent to it stays tangent to it, and Theorem 2.2 shows this holds exactly when the surface's second fundamental form is pure trace (its shear vanishes). A photon sphere is then the symmetric special case, an $SO(3)imes R$-invariant photon surface in a static spherically symmetric spacetime. With that in hand the authors prove the result the paper is built around: if density and radial pressure are non-negative, and the interior ends on a marginally trapped surface, then somewhere outside there must be a radius where light can orbit forever, given by $1-3m(r)/r-4πr^2p_1(r)=0$. Partial converses say a photon sphere cannot sit in a low-mass, non-collapsed spacetime. The authors themselves note in closing that the definition does not carry over unchanged to rotating black holes, where light orbiting with and against the spin circles at different radii.","feed_headline":"Any static black hole must be wrapped in a photon sphere","feed_subtitle":"Given non-negative density and pressure, a trapped surface forces a radius where light orbits forever.","key_machinery":"The load-bearing object is the pure-trace condition $χ_{ab}=Θh_{ab}/3$ on the second fundamental form, equivalently $σ_{ab}=0$. Theorem 2.2 proves the equivalence of four statements: the surface is a photon surface; $χ_{ab}k^ak^b=0$ for every null tangent $k$; $σ_{ab}=0$; and every affine null geodesic of the induced metric is an affine null geodesic of the full spacetime. This converts a statement about families of light rays into a local curvature condition, and it is what makes the rest of the paper tractable. In the static case it collapses further, via Theorem 4.2, to a first-order equation $2g_{θθ}ε^{ab}ε^{cd}r_{;ac}r_{;b}r_{;d}+r_{;a}r^{;a}r_{;c}∂^c g_{θθ}=0$, which for diagonal coordinates reduces to $g_{00}∂_r g_{θθ}=g_{θθ}∂_r g_{00}$ and then to the photon sphere equation $1-3m(r)/r-4πr^2p_1(r)=0$. A separate piece of machinery, Theorem 3.1, gives a second-order evolution equation for the area of an $SO(3)$-invariant photon surface in a non-static spherically symmetric spacetime, with a $1/(4A)(dA/ds)^2$ term, a $G_{ab}n^an^b$ source, and a constant $-4π$ from Gauss–Bonnet.","core_discovery":"The central claim is Theorem 4.3. Write a general static spherically symmetric metric as $g=-(1-2m(r)/r)e^{μ(r)}dt^2+(1-2m(r)/r)^{-1}dr^2+r^2 dΩ^2$, with $m$ and $μ$ continuous and piecewise $C^2$, and suppose density and radial pressure are non-negative and bounded, the spacetime is asymptotically flat in the sense that $4πr^2ρ$ and $4πr^2p_1$ vanish at infinity, $2m(r)<r$ everywhere so no horizon sits at finite $r$, and the inner boundary is a marginally outer trapped 2-sphere. Then there exists $r_1$ with $1-3m(r_1)/r_1-4πr_1^2p_1(r_1)=0$, and $S={r=r_1}$ is a timelike photon surface invariant under the full $SO(3)imes R$ isometry group. The proof is an intermediate value argument: the left-hand side of that equation is at most $-1/2$ at the inner boundary and tends to $1$ at infinity. The converse direction, Proposition 4.4 and Theorem 4.5, shows that under a dominant-energy-style bound $p≤ρ/3$ and a mass bound such as $24m(r)/7<r$, no such surface can exist at all, so a photon sphere must surround a black hole, a naked singularity, or more matter than that bound allows.","pith_inferences":["Because the definition is conformally invariant, any spacetime conformal to a static spherically symmetric one inherits photon surfaces by direct transformation; this is the route the authors take for Robertson–Walker models, and it suggests a cheap way to generate test cases in other conformally flat settings.","The pure-trace condition $σ_{ab}=0$ is a single scalar constraint on an otherwise generic hypersurface, which is why the authors expect photon surfaces to be rare without symmetry. In axisymmetry one would need a pair of surfaces at different radii for prograde and retrograde orbits, so the natural generalization is probably a two-surface or a pair-of-radii condition rather than a single hypersurf","The area evolution equation in Theorem 3.1 is the closest thing here to a stability statement: an area that grows or shrinks according to a Raychaudhuri-type equation with a $G_{ab}n^an^b$ source could be used to ask whether a photon surface in a collapsing spacetime is stable against perturbations, which the paper does not do.","The Vaidya result hints at a horizon-like accounting role for photon surfaces: if every Cauchy surface is cut by the surface, it can serve as a stand-in boundary for a region even when no event horizon exists, which is testable in numerically evolved collapse models."],"forward_implications":["In Schwarzschild–de Sitter the photon sphere sits at $r=3m$ regardless of the cosmological constant, so the radius is insensitive to $Λ$ even though the two spacetimes are not conformally related.","Reissner–Nordström acquires a second timelike photon sphere for $1<(e/m)^2<9/8$, both disappearing above $(e/m)^2=9/8$, so the count of photon spheres tracks whether the singularity is hidden or naked.","In the Janis–Newman–Winicour scalar solution a photon sphere exists only for $1/2<ν≤1$, that is for scalar charge $0≤q^2<3m^2$, matching the range where the deflection angle of an incoming light ray diverges.","For the charged dilaton solution with $β^2=1$, the black hole case $0≤(e/m)^2<2$ has exactly one photon sphere while the naked case $(e/m)^2>2$ has none.","Evolving the $r=3m_1$ Schwarzschild photon sphere backwards through Vaidya null dust produces a surface that runs to a null hypersurface in the past, so the naked singularity is enclosed by it and any partial Cauchy surface reaching spatial infinity is cut in two."],"supporting_citations":[{"why":"Supplies the earlier operational definition of a photon sphere as the limiting radius of unbounded deflection, which this paper replaces with a geometric one.","marker":"[3]"},{"why":"Source of the equation $g_{00}∂_r g_{θθ}=g_{θθ}∂_r g_{00}$ for a general static spherically symmetric metric, which the paper rederives as the diagonal case of its first-order condition.","marker":"[4]"},{"why":"Used for the embedding of de Sitter space as a hyperboloid, and for the type I classification of energy tensors that underwrites the energy condition discussion.","marker":"[5]"},{"why":"Provides the stability bound $m(R)/R<4/9$ for the Schwarzschild interior solution, which fixes where the interior photon spheres can appear.","marker":"[8]"},{"why":"Source of the Janis–Newman–Winicour static spherically symmetric massless scalar solution used as a naked-singularity test case.","marker":"[9]"},{"why":"Supplies the charged dilaton black hole metric whose photon sphere count is worked out in the $β^2=1$ case.","marker":"[12]"}],"fun_headline_variants":["A trapped surface plus non-negative density and pressure forces a photon sphere","Any static spherical black hole must be wrapped in a photon sphere","Photon sphere implies a black hole, naked singularity, or enough matter","A second-order equation tracks photon-surface area in non-static spherical space-time","Under the pressure and mass bounds, no photon sphere can exist"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The existence proof rests on the energy condition that density and radial pressure are both non-negative, since that is what makes $m(r)$ and $μ(r)$ non-decreasing and forces the photon sphere function to change sign between the inner boundary and infinity.","fun_headline_variants_meta":{"raw":{"variants":["A trapped surface plus non-negative density and pressure forces a photon sphere","Any static spherical black hole must be wrapped in a photon sphere","Photon sphere implies a black hole, naked singularity, or enough matter","A second-order equation tracks photon-surface area in non-static spherical space-time","Under the pressure and mass bounds, no photon sphere can exist"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":4995,"prompt_tokens":969,"completion_tokens":4026,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":3948}},"tokens_in":969,"tokens_out":4026,"duration_ms":73160,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":3948,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-10-29T18:13:28.802139+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"falsifier":"Find an asymptotically flat static spherically symmetric metric with $2m(r)<r$ everywhere, non-negative bounded $ρ$ and $p_1$, and a marginally outer trapped inner boundary, for which $1-3m(r)/r-4πr^2p_1(r)$ never vanishes; such a solution would break Theorem 4.3.","supporting_citations":[{"cited_title":"Schwarzschild black hole lensing","cited_arxiv_id":"astro-ph/9904193","evidence_quote":"Supplies the earlier operational definition of a photon sphere as the limiting radius of unbounded deflection, which this paper replaces with a geometric one."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Source of the equation $g_{00}∂_r g_{θθ}=g_{θθ}∂_r g_{00}$ for a general static spherically symmetric metric, which the paper rederives as the diagonal case of its first-order condition."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Used for the embedding of de Sitter space as a hyperboloid, and for the type I classification of energy tensors that underwrites the energy condition discussion."},{"cited_title":"Stephani, General Relativity","cited_arxiv_id":null,"evidence_quote":"Provides the stability bound $m(R)/R<4/9$ for the Schwarzschild interior solution, which fixes where the interior photon spheres can appear."},{"cited_title":"Phys. Rev. Lett. 20, 878","cited_arxiv_id":null,"evidence_quote":"Source of the Janis–Newman–Winicour static spherically symmetric massless scalar solution used as a naked-singularity test case."},{"cited_title":"Phys. Rev. D 46, 1340","cited_arxiv_id":null,"evidence_quote":"Supplies the charged dilaton black hole metric whose photon sphere count is worked out in the $β^2=1$ case."}],"review_version":1}