{"id":"97877190-6e23-44ac-85ab-caf0e4ecc25d","arxiv_id":"2505.08484","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"The zodiac dial ring of the Antikythera Mechanism is reconstructed as a 365-day time scale with 29 to 32 day zodiac months, calibrated to 23 December 178 BC.","lead":"The authors propose that the Antikythera Mechanism's zodiac dial ring was marked with 365 day-ticks rather than 360 degree-marks, making the Sun's uneven yearly speed visible in unequal zodiac month lengths. A general reader should care because this is a test of how much evidence and assumption a famous artifact can bear.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 365-day division of the zodiac ring is not established by the preserved 20%; the two intact months also fit the 360° scheme, and the decisive months are lost (Section 8.3).","rationale":"Read in good faith: the paper is a reconstruction proposal, not a claim to have found new fragmentary evidence. Its strongest claim is the 365-equal-subdivision zodiac ring. What would have to be true is that the preserved subdivisions or their spacing discriminate 365 from 360. The paper does not supply such a measurement. The only two complete signs have 30 marks, which is the signature of the 360° scheme; the 365-day scheme can also produce 30 marks for those signs, but that is a consistency argument, not a test. Section 8.3's own admission that the critical months are lost is the decisive limitation. The reader's focus on Geminus is reasonable but secondary: even granting Geminus, no surviving mark count chooses 365 over 360. Hence my concern is partially aligned with the reader. I would keep the CONDITIONAL verdict: the reconstruction is internally coherent, the authors are candid about missing parts, and a revised version that either supplies a spacing measurement or reframes the 365-day ring as one of several viable hypotheses would be acceptable. I do not see grounds for REJECT: the paper does not fabricate evidence, and its functional model is a legitimate exploratory tool. The missing piece is a quantitative, uncertainty-aware comparison of the two division schemes on the preserved sectors.","tokens_in":26414,"tokens_out":9616,"duration_ms":96846,"concrete_test":"Measure, on the CT volume of Fragment C, the angular span of the 30 preserved subdivision marks in the fully preserved Libra and Scorpio sectors. Under the 365-day hypothesis, 30 subdivisions span 30 × (360/365) ≈ 29.59°; under the 360° hypothesis, they span 30.00°. Fit the common ring center from the preserved bearing-base holes and the Egyptian ring, and propagate the deformation uncertainty that the paper itself stresses (Section 1). If the measured span is consistent with 29.59° and excludes 30.00° at the quoted uncertainty, the 365-day claim gains direct support; if it is consistent with both, the central claim remains underdetermined and should be presented as one plausible hypothesis rather than the finding.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim asserts as fact that the zodiac dial ring was divided into 365 equal day-subdivisions. The surviving evidence does not distinguish this from the traditional 360° division. Section 2.4.1 states that only about 20% of the ring is preserved and that the only fully preserved months, Libra (ΧΗΛΑΙ) and Scorpius (ΣΚΟΡΠΙΟΣ), have 30 subdivisions. Section 8.3 explicitly concedes that the critical sectors — Sagittarius/Capricorn and Taurus/Gemini/Cancer/Virgo — are lost. Since Price's 360° reading and the proposed 365-day reading both predict exactly 30 subdivisions in Libra and Scorpio, the observed count cannot decide between them. The paper's rebuttal (Section 8.3), that five of six best roundings also give 30 subdivisions, only demonstrates consistency; it does not provide positive evidence for 365. The functional preference for a day-scale (parapegma events, Egyptian-ring correlation) motivates a time unit but does not require physically engraving 365 ticks; one could mark the same month boundaries on a 360° ring. The additional assumptions needed for the specific 29-32 day month lengths (Geminus season durations, rounding, uniform increment from Sagittarius to Gemini) are a second layer of underdetermination, but the 365-vs-360 question is logically prior and remains open. The abstract's wording therefore outruns the evidence, though the reconstruction itself is coherent and honestly reports the missing sectors.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a reconstruction of the Antikythera Mechanism's central front dial, arguing that the zodiac dial ring was divided into 365 equal day-subdivisions rather than the traditional 360 degrees. The authors present a bronze reconstruction based on photographs and CT scans of Fragment C, and they use ancient season durations (particularly those of Geminus) to assign unequal day counts (29--32) to the twelve zodiac months. They further claim that the ring was calibrated to 23 December 178 BC via the fiducial line and the first subdivision of Capricorn. The paper explicitly acknowledges that only about 20% of the zodiac ring is preserved and that the decisive sectors are lost (Section 8.3), but the abstract and conclusions nevertheless present the 365-day division as an established result.","tokens_in":26833,"tokens_out":4133,"duration_ms":41272,"significance":"If the 365-day division were established, the paper would contribute a concrete physical reconstruction and a novel reading of the zodiac dial as a time scale rather than an angular scale. The authors deserve credit for the detailed photographic and CT-based documentation of Fragment C, for building working bronze reconstructions, and for transparently reporting the missing sectors that limit inference. However, the central evidential claim is not currently supported: the preserved 30-subdivision months of Libra and Scorpius are compatible with both the 360-degree and 365-day schemes, and the paper's own Section 8.3 concedes that the critical sectors (Sagittarius/Capricorn and Taurus/Gemini/Cancer/Virgo) are lost. The additional assumptions about Geminus' season durations, the rounding rule, and the calibration date are not independently justified, so the specific month lengths and the 365-day ring remain a plausible hypothesis rather than a demonstrated reconstruction.","major_comments":[{"comment":"The preserved evidence does not distinguish between a 360-part and a 365-part zodiac ring. The authors concede in Section 8.3 that the critical sectors (Sagittarius/Capricorn and Taurus/Gemini/Cancer/Virgo) are lost, and the two fully preserved months, Libra (ΧΗΛΑΙ) and Scorpius (ΣΚΟΡΠΙΟΣ), each have 30 subdivisions under both schemes. The statement that \"five of the six 'Best rounding' options also presented Libra and Scorpio with 30 subdivisions\" shows only consistency, not positive evidence for 365. The abstract's assertion that the ring \"was correlated with 365 equal subdivisions-days\" therefore outruns the evidence; at most the paper can claim that a 365-day division is compatible with the preserved fragments.","section":"Abstract and Section 8.3"},{"comment":"The derivation of specific month lengths depends entirely on adopting Geminus' season durations (Autumn 88⅛, Winter 90⅛, Spring 94½, Summer 92½). The authors justify this choice by invoking \"a strong correlation between Geminus' astronomy and the Antikythera Mechanism,\" but they also admit it is unclear whether Geminus' numbers came from his own observations or from earlier astronomers. If the values of Meton, Euctemon, Callippus, or Eudoxus were used instead, the proposed month lengths change, and even the necessity of a 365-day ring is affected. No independent argument is given that the Mechanism's builder specifically encoded Geminus' fractional season lengths, so the 29--32 day month assignments are not established by the evidence presented.","section":"Section 2.4.1 and Table 3"},{"comment":"The calibration date of 23 December 178 BC is imported from the authors' earlier work (Voulgaris et al., 2023a) and then used to assign concrete dates to each zodiac month in Table 4. The present paper does not re-derive or test this date, and the fiducial-line argument shown in Figure 12 depends on assumptions about the original ring positions that are not independently verified. If the earlier dating is insecure, the month-date assignments in Table 4 inherit that uncertainty, and the claim that the design \"directly connects\" the Mechanism to that specific date is stronger than the evidence supplied here.","section":"Section 3 and Table 4"},{"comment":"The functional preference for a day-scale (parapegma events, Egyptian-ring correlation) does not logically require physically engraving 365 day-ticks on the zodiac ring. One could mark the same month boundaries on a 360-degree ring and still interpret the subdivisions as time units, as the authors themselves note the possibility of unequal degree divisions. The paper conflates the unit of measurement with the number of physical subdivisions. To establish the 365-day division, positive evidence of 365 ticks (or of months with 31 or 32 subdivisions) is needed, not merely agreement of the preserved 30-subdivision months with a reconstructed 365-day scheme.","section":"Section 2.2 and Section 8.3"}],"minor_comments":[{"comment":"The abstract contains typos and formatting errors, including \"Mechanism s Central\" and \"scacs,\" which should be corrected before publication.","section":"Abstract"},{"comment":"In the 'Best rounding 5' row, the season-duration line appears to omit the Autumn value (showing only \"90 94 93\"); please check and complete the entry.","section":"Table 3"},{"comment":"The zodiac dial row lists \"360 degrees or 365 days\" as alternatives, but the paper later considers a 360-degree ring with unequal subdivisions as a third possibility; the table should be revised to reflect the full range of proposals.","section":"Table 1"},{"comment":"The statement that the line of solstices does not cross the center because of the unequal season lengths is potentially confusing, since the line is a chord between two points on the ring; please add a sentence explaining the geometry.","section":"Figures 8 and 9"},{"comment":"The claim that \"the solar anomaly does not have to be represented\" on a 360-degree division is inaccurate, because Evans et al. (2010) proposed 360 unequal period subdivisions precisely to represent the anomaly; the contrast should be stated more carefully.","section":"Section 8.1"},{"comment":"Some online sources include access dates while others do not; please standardize the reference format for URLs.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is a reconstruction proposal with an overstrong central claim. The authors honestly report the missing sectors, but the abstract and conclusion present the 365-day division as established despite the evidence being compatible with the traditional 360-degree reading. This is fixable by reframing the contribution as a hypothesis and by softening the language, but the current framing would mislead readers. The heavy reliance on the authors' own previous papers for the calibration date and the Geminus correlation should also be reviewed carefully."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe thing to know is that this is an honest, physically grounded reconstruction that oversells its central result in the abstract. The authors build a bronze replica of the front dial, give detailed tables of possible daily subdivisions per zodiac month, and explicitly admit in Section 8.3 that the decisive parts of the ring are lost. That candor deserves credit. What they have not established is that the ring was divided into 365 days rather than 360 degrees. The preserved evidence—two months with 30 subdivisions each—fits both schemes. Their own Table 3 shows five of six rounding scenarios produce 30 for Libra and Scorpio, so that's consistency, not positive evidence.\n\nThe day-count tables are the real new contribution, along with the Eudoxus-based scenarios and the physical bronze reconstruction. But those tables are fitted, not derived. The Geminus season durations are selected because the authors assert a correlation between Geminus and the Mechanism, while admitting it's unclear whether Geminus borrowed from earlier astronomers. Using Meton or Euctemon instead would shift the month lengths. The 'uniform increment' assumption from Sagittarius to Gemini is a modeling choice with no surviving support. And the calibration date of 23 December 178 BC is imported from their earlier work, so the new claims inherit that dependence.\n\nSoft spots in proportion: the abstract says the solar anomaly and unequal seasons 'were well represented' as if established; the paper's own caveat in Section 8.3 is more careful. Some reconstruction details—the locking pin, the square pin-drivers, the sliding catches—are presented as definite when they are hypotheses. The appendix's construction protocol and the airplane anecdote are tangential to the main argument.\n\nWho should read this: anyone working on the Antikythera Mechanism or on ancient instrument reconstruction. It is a serious attempt that advances the physical understanding of the front plate, even if the 365-day division remains one plausible hypothesis alongside 360. I would send it to peer review, with the expectation that the authors qualify the claim and quantify measurement uncertainty. A good referee can push them toward presenting the 365-day scheme as a testable possibility, not a conclusion.","headline":"A serious, physically grounded reconstruction whose central 365-day claim is underdetermined by the preserved evidence and oversold in the abstract.","tokens_in":27314,"tokens_out":2449,"would_cite":false,"duration_ms":25350,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper argues that the Antikythera Mechanism's zodiac dial ring was a 365-day time scale, not a 360-degree angular ring.","keywords":["Antikythera Mechanism","zodiac dial ring","Egyptian dial ring","solar anomaly","365-day year","Geminus season lengths","Fragment C","fiducial line"],"falsifier":"A direct observation that would settle the claim is to locate and measure any preserved subdivision of the missing zodiac months, Sagittarius/Capricorn or Taurus/Gemini/Cancer/Virgo, in the CT volume of Fragment C. If a complete month is found to contain exactly 30 equally spaced marks, or if the preserved Libra and Scorpio arcs turn out to be exact $30^\\circ$ angular sectors with no adjacent month deviating from 30, the 365-subdivision claim fails; if Gemini and Cancer show 31 or 32 equal marks and Sagittarius and Capricorn show 28 or 29, it stands.","tokens_in":26198,"feed_emoji":"☀️","tokens_out":11074,"duration_ms":110174,"temperature":0.7,"pith_summary":"The paper argues that the zodiac dial ring on the Antikythera Mechanism's front face was not a 360-degree angular scale but a 365-day calendar ring. On that reading, each of the twelve zodiac month sectors contains an unequal number of day-subdivisions, from 29 to 32, so the engraved ring itself encodes the solar anomaly, the Sun's varying apparent speed, and the unequal durations of the seasons. The authors support the claim by reconstructing the four parts of the central front dial from same-scale front/back images and CT scans of Fragment C, then matching their reconstruction to the preserved 30-subdivision months of Libra and Scorpio. The result matters because it makes the front dial a coherent time-measuring instrument, calibrated to an absolute starting date of 23 December 178 BC, without invoking extra gears or complex spacing assumptions.","feed_headline":"Antikythera zodiac ring used 365 days, not 360 degrees","feed_subtitle":"Unequal 29-to-32-day zodiac months let the ancient calculator track the Sun's varying speed and season lengths.","key_machinery":"The load-bearing object is the 365-subdivision zodiac dial ring: 365 equally spaced marks around the circle, assigned to twelve zodiac months whose boundaries fall between marks so that month lengths are unequal. The assignment of lengths is carried by a rounding table built from Geminus' season durations, with the constraints that Sagittarius, the month containing perihelion, is shortest and Gemini, the month containing aphelion, is longest. The calibration geometry is carried by the fiducial line and the vertical line VLb, which put the 1st of Capricorn opposite the 18th of Hathyr and set the epoch. This machinery lets a single uniformly engraved ring do double duty: days for the Sun and, at about 1.8 hours per subdivision, a timing scale for lunar motion.","core_discovery":"The paper's central claim is that the zodiac dial ring was divided into 365 equal subdivisions representing days of a rounded solar tropical year of $365.25$ days, not into 360 degrees; hence each zodiac month occupies an unequal number of subdivisions, from 29 for Capricorn and Sagittarius to 32 for Gemini and Cancer under the preferred rounding. The unequal arcs are not an error or a crude approximation but the deliberate representation of the solar anomaly and of the season durations reported by the Hellenistic astronomer Geminus: autumn $88\\frac{1}{8}$, winter $90\\frac{1}{8}$, spring $94\\frac{1}{2}$, and summer $92\\frac{1}{2}$ days. The paper also claims the ring was calibrated by making the first subdivision of Capricorn point at the fiducial line, which places the initial date at 23 December 178 BC, with the 18th of Hathyr on the Egyptian ring aligned to the same line. In this design the front dial is a time-measuring device: the solar pointer crosses day-subdivisions, the parapegma star-event letters fall on days, and the Egyptian ring shifts by one subdivision every four solar years to track the quarter-day excess.","pith_inferences":["A testable extension: if the 365-day division is correct, future imaging of the missing Gemini/Cancer and Sagittarius/Capricorn sectors should reveal 31 or 32 subdivisions for the first pair and 28 or 29 for the second, rather than 30 equal marks.","The same criterion could be applied to other ancient dials: whenever a graduated ring sits next to a calendar scale and is marked with event letters that refer to dates, counting its subdivisions as days rather than degrees is the hypothesis with the fewest extra assumptions.","The one-subdivision counter-rotation of the Egyptian ring every four years is the same correction that the Julian calendar codified as a leap day; if this reconstruction is right, the Antikythera Mechanism is an early mechanical embodiment of that correction, though the paper does not draw the comparison."],"forward_implications":["The two fully preserved zodiac months, Libra and Scorpio, each with 30 subdivisions, cease to be evidence for a 360-degree ring: five of the six best roundings in Table 3 keep those months at 30 day-subdivisions, so the surviving fragments are consistent with a 365-day ring.","Because the parapegma letter-events are date events, putting them on a day-subdivision ring makes the star calendar and the zodiac ring share the same unit, so a pointer reading on the zodiac ring corresponds directly to a parapegma date.","A user can convert any date between the Egyptian and zodiac calendars by reading the two concentric rings together; after four full solar turns, the Egyptian ring is shifted one subdivision, mechanically reproducing the $365.25$-day tropical-year excess over the 365-day Egyptian year.","With the preferred rounding and the 23 December 178 BC calibration, the paper generates absolute date ranges for all twelve zodiac months, placing perihelion in Sagittarius around 29 November and aphelion in Gemini around 30 May for that epoch.","The front dial can be reconstructed as a time-measuring device with fewer added components than a 360-degree unequal-spacing scheme, because the solar anomaly is carried by the unequal month lengths rather than by irregularly spaced degree marks."],"supporting_citations":[{"why":"It documents the front dial and parapegma inscriptions, including the preserved 30 subdivisions of Libra and Scorpio and the solar-ray pointer, which any zodiac-ring division must fit.","marker":"Bitsakis and Jones, 2016a"},{"why":"It established the 360-degree reading of the zodiac ring and first described the Egyptian dial ring and the 365-hole bearing ring that the paper's reconstruction builds on.","marker":"Price, 1974"},{"why":"It introduced the solar-anomaly requirement and the perihelion/aphelion dates that the paper uses to decide which zodiac months should have the most and fewest day-subdivisions.","marker":"Evans et al., 2010"},{"why":"It supplies the 23 December 178 BC initial calibration date and the fiducial-line geometry that places the 1st subdivision of Capricorn opposite the 18th of Hathyr.","marker":"Voulgaris et al., 2023a"},{"why":"It is the edition of Geminus' astronomical text from which the paper takes the four season durations used to compute the 29-to-32 day month lengths.","marker":"Manitius, 1880"},{"why":"It publishes the Eudoxus papyrus with 30-day zodiac months plus solstice extra days, which provides the alternative Scenario 2 for the ring division.","marker":"Blass, 1887"}],"fun_headline_variants":["365 days, not 360: Antikythera's zodiac ring secret","Unequal month lengths on Antikythera mirror Sun's speed","Ancient Greek ring used days to track seasons, not degrees","Antikythera zodiac dial: a calendar, not a protractor","How Antikythera's 365-day ring captured solar anomaly"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the season lengths the builder encoded are the ones attributed to Geminus (autumn $88\\frac{1}{8}$, winter $90\\frac{1}{8}$, spring $94\\frac{1}{2}$, and summer $92\\frac{1}{2}$ days); the paper itself notes in Section 2.4.1 that it is unclear whether these were Geminus' own values or borrowed from earlier astronomers, so if a different ancient season table was used, the 29-to-32 day month pattern and the calibration dates would shift.","fun_headline_variants_meta":{"raw":{"variants":["365 days, not 360: Antikythera's zodiac ring secret","Unequal month lengths on Antikythera mirror Sun's speed","Ancient Greek ring used days to track seasons, not degrees","Antikythera zodiac dial: a calendar, not a protractor","How Antikythera's 365-day ring captured solar anomaly"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000325,"raw_usage":{"total_tokens":1834,"prompt_tokens":971,"completion_tokens":863,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":587,"completion_tokens_details":{"reasoning_tokens":771}},"tokens_in":587,"tokens_out":863,"duration_ms":8563,"temperature":1.0,"reasoning_tokens":771,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:53:47.156947+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct observation that would settle the claim is to locate and measure any preserved subdivision of the missing zodiac months, Sagittarius/Capricorn or Taurus/Gemini/Cancer/Virgo, in the CT volume of Fragment C. If a complete month is found to contain exactly 30 equally spaced marks, or if the preserved Libra and Scorpio arcs turn out to be exact $30^\\circ$ angular sectors with no adjacent month deviating from 30, the 365-subdivision claim fails; if Gemini and Cancer show 31 or 32 equal marks and Sagittarius and Capricorn show 28 or 29, it stands.","supporting_citations":[],"review_version":1}