Measurement Quantization (MQ)

A First-Principles Research Program Developed by Jody A. Geiger at the Informativity Institute

Measurement Quantization (MQ) is the research program developed by Jody A. Geiger, cofounder and researcher at the Informativity Institute. It proposes a first-principles physical framework in which the configurations underlying measurement are discrete and count-based, while observable quantities are continuous physical realizations of those configurations.^[1]

MQ describes these two representations as the Internal Frame and the System Frame. The Internal Frame is dimensionless and vectorless. Physical length, mass, time, momentum, charge, gravitation, and other measurable quantities arise only through the Frames mapping into the observable System Frame.

The program develops mathematical constructions and quantitative comparisons spanning fundamental measurement, electromagnetism, gravitation, quantum mechanics, galactic dynamics, and cosmology. Those results have different evidential roles. Recovering an established relation tests the construction under its assumptions; a comparison with independently collected measurements can test a specific prediction when its calibration, uncertainty, and physical interpretation are specified. The 199 Discoveries overview below organizes the Institute’s catalog rather than counting independent experimental confirmations.^[6]

A Prediction Enters the Scientific Record

The Informativity differential is MQ’s finite-count departure from continuous length normalization. It is distinct from velocity-dependent Lorentz contraction. In the gravitational construction, finite electromagnetic and upper-count realizations give different model values of G. That is a mathematical consequence of the specified realization relations; assigning a laboratory mode to one of those values is a further physical hypothesis.^[1][7]

The general realization distinction predates the 2026 NIST result. In Measurement quantization, Appendices M and N and Table N.1 present finite-count and upper-count gravitational constructions using the calibration then adopted.^[1]

Yan-Jie Zhang and colleagues’ Progress on measurement of gravitational constant G, published in Measurement Science and Technology in 2025, reviews the methods, precision, and persistent dispersion of laboratory determinations.^[3]

Reference 154 in Progress on measurement of gravitational constant G identifies Measurement quantization. This establishes a traceable citation before the NIST publication. It does not establish that the NIST experiment adopted MQ’s premises or that a specific free-deflection/servo assignment was fixed in advance.^[3]

Laboratory Measurements from NIST

In Redetermination of the gravitational constant with the BIPM torsion balance at NIST, Schlamminger and colleagues report measurements using the original International Bureau of Weights and Measures (BIPM) balance with documented modifications at NIST. Equation (64) gives their combined result, with standard uncertainty, of 6.67387 ± 0.00038 in units of 10⁻¹¹ m³ kg⁻¹ s⁻².^[4]

Tables 16–18 of Redetermination of the gravitational constant with the BIPM torsion balance at NIST give the individual determinations, uncertainty budgets, and correlations. Free deflection infers gravitational torque from mechanical motion; electrostatic servo balances it with electrical torque. Both copper and sapphire campaigns gave a higher free-deflection value. The accompanying artwork is illustrative; the proposed MQ experiment below is a separate design.^[4]

The paper reports an unresolved mode offset and treats unidentified configuration effects in its combined analysis. It regards instrumental systematics as the more plausible explanation of the wider dispersion of G measurements. Those effects must remain in the interpretation of an MQ comparison.^[4]

For MQ, the useful observable is the signed difference between paired modes after their shared uncertainties are included. A general acknowledgment that new physics is possible does not favor a particular mechanism. The numerical comparison, its calibration, and its apparatus assignment determine the support available.

The NIST study also identifies Clive Speake’s involvement in both the BIPM and NIST efforts. The measurements are external to MQ, but the BIPM and NIST determinations are not wholly independent replications of one another. Their shared apparatus history matters when synthesizing them.^[4]

The Institute’s synthesis of mature BIPM and NIST pairs reports a common gap of 60.994 ± 26.930 ppm under a specified covariance model. This is an MQ-associated reanalysis, not the NIST paper’s result. It assumes a common physical contrast and particular treatment of cross-campaign dependence and unexplained variation. The observed common-gap estimate and its interpretation must be kept separate.^[7]

One Principle Across the Physical Sciences

MQ describes finite whole-unit counts through the Frames mapping into continuous physical observables. Its length-normalization departure becomes small in the upper-count limit. The proposed physical effects depend on how this count construction is related to the particular observable and measurement regime.^[1]

The Informativity differential introduced above is one part of this shared mathematical structure.

The framework applies that structure at electromagnetic, gravitational, galactic, and cosmological scales. A shared formula or symbol does not make the empirical tests independent, and agreement in one regime extends to another only through the relevant dependencies.^[9]

This provides the organizing principle connecting MQ's principal results:

• fundamental length, mass, and time;
• the fine structure constant and elementary charge;
• the electric, magnetic, and Coulomb constants;
• the gravitational constant;
• Newtonian gravitation and the equivalence principle;
• general relativity and the cosmological term;
• quantum probability and measurement;
• galactic rotation and gravitational lensing;
• the apparent dark matter phenomenon;
• cosmological domain fractions;
• the Hubble tension;
• cosmic expansion and spatial flatness;
• the cosmic microwave background;
• and the physical development of the universe from a pre-physical configuration domain.

The same distinction between discrete configuration and continuous realization is retained throughout.

Physical Constants from a Common Measurement Structure

Precisely measured constants provide exacting comparisons for MQ’s proposed measurement structure. Their evidential role depends on which quantities are measured inputs, exact SI definitions, inferred calibrations, and outputs of the calculation.^[10]

MQ relates fundamental length l_f, mass m_f, and time t_f through a common count construction. These symbols denote the proposed fundamental measures. Numerical realizations of that construction also use specified empirical and SI inputs; “first principles” does not by itself establish independence from measured constants.^[1][9]

Among the quantities treated within this construction are the elementary charge e, electric constant ε₀, magnetic constant μ₀, Coulomb constant k_e, and gravitational constant G.

In the electromagnetic calibration, the dimensionless interaction-demarcation parameter n_L has a reported rounded value of 84.60055. MQ calls this boundary the Fine Structure demarcation. Its role is to connect the fine-structure, charge, and action constructions under their common calibration.^[9]

Agreement among quantities connected by exact identities or shared calibration establishes consistency of that input chain. A new empirical test requires an output not already fixed by the comparison input, together with the relevant uncertainty and covariance.

The Institute’s Physical Support for Measurement Quantization (MQ) separates these cases. Its historical aggregate “concordance” scores are not used here as detection significances or combined evidence for the framework. The complete joint covariance, calibration dependence, and statistical transformation would have to be specified before such an aggregate could support that interpretation.^[9]

Explore the physical support for Measurement Quantization

A Prospective Test of the Gravitational Realization

The laboratory comparison above identifies a specific measurement problem. Precision G determinations retain unexplained dispersion, while a controlled paired-mode experiment can test a differential target without relying only on an absolute calibration.^[4][7]

Under the specified MQ mode assignment, the target remains a positive 34.9284 ppm servo-normalized separation. The distinction between a finite electromagnetic realization and its macroscopic upper-count limit supplies the model rationale. The experiment must establish that its two modes actually realize that distinction.^[7]

A prior general proposal and a prior apparatus-specific prediction are different evidential claims.

The published 2023 construction documents the former. The sources reviewed here do not establish a dated advance record of the particular NIST mode assignment. The present comparison is therefore treated as retrospective numerical compatibility, while a future frozen protocol could provide a prospective test.^[1][7]

For reference, Equation (64) of Redetermination of the gravitational constant with the BIPM torsion balance at NIST gives the combined absolute determination shown below. It summarizes that paper’s configuration-effect analysis; it is distinct from an estimate of a mode gap.^[4]

The Institute’s covariance-aware common-gap estimate for the selected BIPM/NIST pairs is 60.994 ± 26.930 ppm. The quoted uncertainty belongs to that synthesis, not to an independent repetition of the experiment.^[7]

Under that common-gap model with the quoted covariance, the fixed-target likelihood is about 8.1 times the likelihood at exactly zero gap. The Institute reports about 5.4 when additional unexplained heterogeneity is admitted. These ratios compare two specified hypotheses under different error models; they are not posterior probabilities or odds against all instrumental explanations. The pooled analysis also inherits shared BIPM/NIST apparatus history and uncertain cross-campaign effects.^[7]

The target, instrument sensitivity, and residual systematic effects determine what a future result could establish. A sufficiently precise result near zero could constrain this mode assignment; a result compatible with the target would support it to the extent that competing instrumental and physical explanations are discriminated.

The useful next step is a published protocol and a reproducible differential uncertainty budget, followed by measurements assessed under the predeclared interpretation.^[8]

Explore the MQ description of the gravitational constant

Gravity from Measurement Geometry

Within MQ, gravitational response is constructed from continuous physical description and finite-count realization rather than introduced as a separate primitive interaction. That is the framework’s proposed mathematical account.^[1]

The construction relates the length-normalization departure to a realized gravitational response. Evaluating it requires the applicable scale, source geometry, and realization assumptions; the same numerical correction cannot simply be transferred between unrelated observables.

The classical constructions recover Newtonian relations, and the continuum program seeks an Einstein-field description under specified mapping, action, and source assumptions. Such recovery demonstrates consistency with an established description where the assumptions apply. It becomes a distinct empirical discriminator only where an additional observable prediction differs from the alternatives.^[1][9]

This distinction is important. Curved spacetime remains a valid observable description within the System Frame. MQ proposes, however, that curvature is not fundamental to the underlying Internal Frame. The latter contains dimensionless count relationships rather than physical lengths, vectors, or an independently existing curved spacetime.

This gives a proposed route between discrete measurement geometry and continuous gravitation. The mathematical construction and its observed physical realization remain separate questions, assessed through their respective derivations and measurements.

Galactic Rotation Without a Fitted Non-Baryonic Halo

Galactic rotation supplies a population-level comparison for the MQ response law, with both favorable residual reductions and material source-model sensitivity.^[13]

Published rotation curves often exceed the velocity expected from a specified Newtonian baryonic reconstruction. The size of that discrepancy depends on the gas distribution, stellar mass-to-light normalization, distance, inclination, and adopted geometry.^[14]

MQ transforms the adopted Newtonian baryonic acceleration into a realized gravitational response. The calculation reported here adds no galaxy-specific MQ acceleration parameter or dark halo. It still inherits the empirical source reconstruction and the cosmological normalization used to set its acceleration scale.^[13]

The symbol a_MQ denotes the fixed acceleration scale used in these galaxy calculations, in metres per second squared. The source analysis adopts the displayed value before its catalog comparisons.^[13]

The scale follows from the stated MQ cosmological construction rather than a separate fit to each galaxy. Its upstream normalization and the adopted baryonic inputs remain part of the physical hypothesis being tested.

Jody A. Geiger’s A First Principles Derivation of the Galactic Radial Acceleration Relation from Measurement Quantization reports the following analyses of external astronomical data.^[13]

Section 4.3 of A First Principles Derivation of the Galactic Radial Acceleration Relation from Measurement Quantization compares 38 Milky Way measurements from Eilers and colleagues’ The Circular Velocity Curve of the Milky Way from 5 to 25 kpc with the baryonic part of McMillan’s The mass distribution and gravitational potential of the Milky Way. The reported root mean square (RMS) residual falls from 69.03 to 8.90 km s⁻¹, an 87.1% reduction relative to that baryonic Newtonian comparator with its dark-halo force omitted. This is a substantial improvement for the stated comparison, not a comparison with an optimized dark-halo model.^[11][12][13]

With the specified propagated McMillan source covariance, the same analysis reports chi-square 39.57 for 38 degrees of freedom and a right-tail probability of 0.400. A diagonal-error treatment rejects the complete sample. The full point-to-point observational covariance and complete source posterior are not jointly propagated, so the favorable probability remains conditional on the chosen error model.^[13]

The author’s complete SPARC comparison uses 175 galaxies and 3,366 radial measurements from Lelli, McGaugh, and Schombert’s SPARC: Mass Models for 175 Disk Galaxies with Spitzer Photometry and Accurate Rotation Curves. The ratio of point-weighted mean MQ and observed velocities is 1.0001888.^[13][14]

This is an aggregate normalization offset of about 0.019%. The same analysis reports a pooled RMS residual of 22.82 km s⁻¹ and mean absolute residual of 16.20 km s⁻¹. The small signed aggregate offset can coexist with substantial pointwise departures.^[13]

For the 99 galaxies in the catalog’s highest-quality category, the reported galaxy-weighted geometric ratio is 1.0042, about 0.42% above unity. This is a different weighting and subset from the full-catalog ratio of means, so the two numbers are not interchangeable measures of precision.^[13]

The complete-catalog analysis also reports velocity-dependent residuals and sensitivity to source reconstruction. Radial points within a galaxy share distance, inclination, and mass-model quantities; they are not 3,366 independent systems. The results support assessing the fixed law across a broad population while leaving a covariance-complete population likelihood and matched comparison with alternatives unresolved.^[13]

Explore the effective gravitational mass of a galaxy

A Test Beyond the SPARC Galaxy Sample

A useful physical model must survive application to data outside the population on which its behavior was initially characterized.

The same MQ paper also analyzes 15 galaxies and 288 figure-extracted radial comparisons from the separate LITTLE THINGS population. Its observational basis includes Oh and colleagues’ High-resolution Mass Models of Dwarf Galaxies from LITTLE THINGS. The reported galaxy-weighted geometric ratio is 0.9907, with a 95% galaxy-block bootstrap interval of 0.868–1.130.^[13][15]

That interval includes unity, supporting compatibility of the population normalization under the stated reconstruction and bootstrap. It does not establish precise pointwise agreement, and the conditional velocity ordering seen in parts of SPARC was not independently reproduced in this smaller sample. Figure extraction and reconstructed baryonic sources further limit its precision. The population is external to SPARC; the MQ reanalysis itself is author-reported rather than an independent replication.^[13]

The combined galaxy program therefore provides a direct observational question: can baryonic matter, transformed only by the MQ realization geometry, reproduce the gravitational response conventionally attributed to non-baryonic dark matter?

That question can be tested with increasingly precise baryonic reconstructions, rotation curves, gravitational lensing, galaxy clusters, and high-redshift observations.

Cosmological Fractions from Geometry

MQ also approaches the apparent composition of the universe differently from ΛCDM.

MQ partitions its realized expanding geometry into dark, observable, unobserved, and visible referenceability domains. The fixed fractions below are conditional geometric outputs, not four independent measurements of cosmological substances. The observable fraction contains the visible and unobserved fractions.^[1][13]

In the list below, upright Ω labels a fixed geometric fraction. The subscripts dk, uobs, obs, and vis identify the dark, unobserved, total observable, and visible domains, respectively. These are distinct from independently fitted ΛCDM density parameters.

Their proximity to some observationally inferred density fractions is a numerical correspondence requiring a physical mapping between domain membership and measured cosmological observables. The four displayed percentages share their coefficient calibration and include subset relations; they cannot be counted as four independent successful tests.^[13]

Instead, they arise as fixed geometric partitions associated with what can be visible, observable, unobserved, or dark within an expanding frame.

A quantitative cosmological assessment must propagate the geometric construction into expansion, perturbations, and radiation observables before comparing like quantities. Domain percentages alone do not provide that likelihood.^[9]

Two Realizations of Cosmic Expansion

The Hubble tension is a comparison between differently inferred expansion rates. MQ proposes to relate it to the Frames mapping; evaluating that proposal requires both the shared age normalization and a consistent mapping to the observational analyses.^[2][9]

Within that proposal, H_sys and H_int denote two physically realized comparison rates associated with radiation-anchored and matter-anchored observations. Their subscripts label MQ realization branches. They do not assign an SI rate to the primitive dimensionless Internal Frame.

The October 2026 calculation gives the following conditional, rounded rates in kilometres per second per megaparsec. Parentheses give propagated input standard uncertainties in the final digits, excluding unquantified model discrepancy. Both branches share a temperature-derived age normalization whose absolute time prescription and radiation history require physical justification.^[9][16]

For external comparison, Camphuis and colleagues’ SPT-3G D1: CMB temperature and polarization power spectra and cosmology from 2019 and 2020 observations of the SPT-3G main field reports 67.19 ± 0.38 km s⁻¹ Mpc⁻¹ for its combined CMB analysis under ΛCDM. Breuval and colleagues’ Small Magellanic Cloud Cepheids Observed with the Hubble Space Telescope Provide a New Anchor for the SH0ES Distance Ladder reports 73.17 ± 0.86 km s⁻¹ Mpc⁻¹ from its distance-ladder analysis.^[17][18]

The displayed MQ central values are 1.069 and 0.338 km s⁻¹ Mpc⁻¹ above those two estimates. Dividing by the quoted observational uncertainties alone gives offsets of 2.81 and 0.39 standard uncertainties. This retains both the lower-branch discrepancy and upper-branch proximity. These fixed-central-value diagnostics do not include the full shared model uncertainty or translate a ΛCDM CMB likelihood into MQ. The common calibration and age also prevent treating the branches as independent tests.^[9][17][18]

The proposed two-branch interpretation can therefore be examined quantitatively, but discriminating support requires a consistent cosmological likelihood and its shared dependencies. Matching the separation of two estimates cannot by itself resolve the discrepancy of either absolute branch.

Explore the MQ description of the Hubble tension

The Cosmic Microwave Background from Expansion Geometry

The CMB monopole temperature, its angular spectrum, and cosmic expansion test different aspects of a cosmological construction. Agreement with the monopole does not automatically assess acoustic peaks or perturbation dynamics.^[19]

Section 3.11, Equation (118), and Table 6 of A First Principles Derivation of the Galactic Radial Acceleration Relation from Measurement Quantization give a forward temperature calculation using an adopted Planck 2018 cosmic age and the blackbody radiation relation. Here T_CMB denotes the CMB monopole temperature. The source reports^[13]

Fixsen’s The Temperature of the Cosmic Microwave Background gives the consolidated observational determination shown below. The source-calculation and measurement uncertainties have different origins.^[19]

The same forward construction reports a radiation energy density, denoted ρ_γ, of 4.1723 × 10⁻¹⁴ J m⁻³. It is linked to the temperature through the blackbody relation, rather than an additional independent confirmation.^[13]

The central predicted and measured temperatures differ by 0.00039 K. The predicted uncertainty above is the source’s age-propagated uncertainty for this forward route; it does not encompass the unquantified physical normalization and radiation-history uncertainty.^[13]

The forward route does not fit the measured monopole temperature, but it adopts a cosmic age inferred within a CMB-based cosmological model. Its numerical agreement is consequently a conditional cross-scale consistency result. Inverting the relation to infer age from the measured temperature and then recovering that temperature would instead test calibration closure. Those two routes must not be combined as independent evidence.^[9][13]

Explore the MQ description of the CMB power spectrum

A Universe Without Fundamental Spatial Curvature

Planck’s cosmological analyses, including measurements combined with baryon acoustic oscillations, constrain large-scale spatial curvature to be close to zero within their assumed cosmological model.^[20]

MQ reaches flatness from a different starting point.

The Internal Frame is dimensionless and vectorless; physical spatial curvature is therefore not a primitive property of that underlying configuration domain. Spatial geometry appears only when count relationships are realized through the Frames mapping into the physical System Frame.

MQ proposes a conditional explanation of realized spatial flatness from its referenceability and mapping premises. The absence of a primitive metric in the Internal Frame is not itself an observational measurement of zero curvature. Relating the construction to measured curvature requires its realized metric and cosmological dynamics, with the same observational assumptions exposed.^[21]

Explore the MQ description of a universe without curvature

From a Pre-Physical Configuration to a Physical Universe

MQ extends the same framework to the origin and development of the universe.

It does not begin with physical length, mass, time, or spacetime already in existence. Instead, it distinguishes a pre-physical configuration domain from the later Internal Frame and physically realized System Frame.

The quantum-epoch construction develops a count scale before physically realized clocks and distances are available. Its minimum three-unit referencing geometry uses an assigned positive orthogonal norm in the realized description; the dimensional interpretation therefore depends on those mapping premises.^[22]

The reported scale near 363,312 years is an SI-equivalent encoding of the pre-referenceable count construction. It is not elapsed time recorded by a physical clock before realization, nor the same observable as the standard cosmological age at photon decoupling. Its dimensional normalization is a premise to justify, not a precision measurement of an early epoch.^[22]

The Frames mapping supplies the physically meaningful length, mass, and time of the System Frame. MQ then seeks a connected account of expansion, mass allocation, radiation, and structure. Assessing that history requires a consistent dynamical and observational model.^[22]

This approach removes the need to assign physical dimensions to a state that, within MQ, precedes physical measurement itself. It also distinguishes the geometric allocation of mass during cosmic development from the later gravitational accretion of that matter into stars and galaxies.

Explore the MQ quantum epoch

Quantum Mechanics as Realization

At the opposite end of physical scale, MQ applies the same distinction between configuration and observable realization to quantum mechanics.

Within the Internal Frame, admissible configurations are discrete. Observable probabilities arise when the availability of those configurations is represented in the continuous System Frame.

MQ studies normalized realization availability as an account of Born-rule probabilities and develops continuous amplitude and Schrödinger-evolution descriptions under specified update and normalization assumptions. A recovery of standard quantum statistics under those assumptions is a mathematical consistency result.^[23]

In the proposed measurement account, an outcome is a stable realized record associated with admissible configurations. Establishing the physical record-selection mechanism requires more than reproducing the probability formula.^[23]

Where MQ and conventional quantum mechanics predict the same measured probabilities, the experiment supports their shared observable description without selecting the proposed underlying interpretation. A discriminating experiment would need a separately specified difference in observable behavior.^[23]

Independent Tests, Not One Combined Significance

The breadth of MQ makes statistical interpretation especially important.

Physical constants, laboratory G measurements, galaxy rotation curves, expansion estimates, and the CMB have different calibration and error structures. The comparisons on this page are therefore assessed within their stated domains.

Some calculations share measured inputs or exact identities; some reuse related observations; others use distinct external populations. The source and input chain determine whether each result supplies calibration closure, conditional numerical agreement, or a new independent constraint.

No joint framework-wide probability is calculated here. Combining these comparisons would require specified hypotheses, the full dependence structure, selection and fitting rules, and appropriate uncertainty models. A missing joint analysis leaves that assessment unresolved; it does not erase the support or constraints available for the particular results described above.

The continuing research question is

Does the same underlying MQ structure continue to survive independent quantitative tests as it is applied to new physical systems and new measurements?

That is the standard against which the research program should be evaluated.

What Jody A. Geiger and Measurement Quantization Have Developed

The MQ program developed by Jody A. Geiger connects proposed count-based constructions across fundamental measures, quantum mechanics, electromagnetism, gravitation, astrophysics, and cosmology. The following summary includes formal recoveries, conditional calculations, and observational comparisons. Their evidential roles are those specified in the preceding sections.^[1][9]

Its principal results now include:

• discrete derivations of fundamental physical measures;
• first-principles descriptions of multiple physical constants;
• a common electromagnetic interaction demarcation;
• a measurement-geometric origin for the Informativity differential;
• recovery of Newtonian gravitation and the continuum gravitational field description;
• a realization-dependent prediction for precision measurements of G;
• a fixed-form galactic gravitational realization law;
• application of that law to 175 SPARC galaxies and an independent dwarf-galaxy population;
• a geometric interpretation of the apparent dark matter phenomenon;
• fixed cosmological domain partitions;
• two predicted realizations of the Hubble expansion;
• a derived present-day CMB temperature;
• a flat large-scale spatial geometry;
• a finite quantum epoch and physical transition to the System Frame;
• and a realization-based physical interpretation of quantum probability and measurement.

The strongest assessment is claim-specific. The Milky Way residual reduction, the conditional SPARC population comparisons, and the NIST mode compatibility deserve the credit justified by their stated comparators and uncertainty models. Shared calibrations, adverse residual structure, and incomplete dynamical or covariance treatments limit how far that support transfers to other claims.

A Framework Designed to Be Tested

A useful test identifies a prediction, the conditions under which it applies, and an outcome capable of constraining it.

The Institute’s research preprints provide the detailed proposals and calculations. Peer-reviewed papers supply the measurement and theory sources identified in the footnotes. Author-reported analyses and independent data remain distinguished throughout this page.

Further tests include a predeclared same-apparatus G comparison; galaxy analyses with independently constrained baryonic sources and fuller covariance; matched lensing and cluster predictions; and cosmological calculations that propagate the shared normalization into radiation and expansion observables. Each test should state its sensitivity, parameter freedom, calibration inputs, and relevant alternatives.

Successful comparisons can support particular predictions; precise disagreements can constrain them within the measured regime. Neither conclusion requires an all-or-nothing verdict on every part of MQ.

That is the continuing test of Measurement Quantization.

Explore Measurement Quantization

Follow the reported evidence, its sources, and the assumptions connecting each calculation to its observable.

• Physical Support for Measurement Quantization
• The Gravitational Constant
• Effective Mass of a Galaxy
• The Hubble Tension
• CMB Power Spectrum
• Universe Without Curvature
• Quantum Epoch
• Early Galaxy Formation at 280 Million Years