Cross-paper connection · Molecular-dimensions companion preview

Three routes toward a molecular number.

Radiation, visible particle motion, and the viscosity of a solution ask different questions. Compare what each needs before it can supply a number.

The dissertation is a companion to the four papers, not a fifth paper in the edition. This preview makes the mathematical comparison interactive without presenting illustrative settings as historical observations.

Return to connections among the papers →

Three methods · One comparison

What information fixes the molecular number?

Companion preview · Host calculation

The radiation row reconstructs a historical calculation. The other rows use authored illustrative inputs and modern SI constants, so they are consistency checks, not independent counts or historical measurements. No Bancelin dataset is claimed here.

Predict before calculating

Can displacement alone distinguish a large particle from a large molecular number? With the same viscosity and diffusion observations, will correcting the viscosity coefficient increase or decrease the inferred radius?

Inputs, not fitted answers
Shared solvent conditions for the two diffusion routes
1. Radiation: vary α

One means the existing owner's corrected reference α = 6.1 × 10⁻⁵⁷ in its historical CGS convention. β, R and the speed of light remain fixed. Perturbations are sensitivity tests, not measured historical constants.

2. Brownian displacement

The interval requires independent Gaussian coordinate increments, known zero drift, exact timing and calibration, and no localization noise, exposure blur, overlap or censoring. Selecting “No” refuses this model rather than narrowing an unjustified interval.

3. Viscosity plus solute diffusion

The tracer radius in panel 2 is not the solute radius in this panel. Both use the declared solvent conditions. No particle radius or target molecular number is supplied to this joint inversion.

One accepted revision: molecular number per mole
MethodResultWhat this means
Radiation constants6.1705e23 mol^-1Historical reconstruction; α scaled by 1. No confidence interval is asserted.
Brownian displacement6.0143e23 mol^-1
Conditional 95% interval: 4.4639e23 to 7.7922e23 mol^-1
Modern consistency check conditional on an independent radius and the admitted observation model.
Viscosity and solute diffusion6.0188e23 mol^-1Illustrative dilute-sphere inversion with coefficient 2.5. Not a historical dataset or uncertainty interval.
Modern reference6.0221e23 mol^-1Exactly 6.02214076 × 10²³ mol⁻¹ by definition, not measured by this laboratory.

Do not average these rows. They have different provenance and assumptions. The modern gas constant is defined using Nₐ and kᵦ; agreement in the illustrative diffusion rows is not independent evidence for either constant.

What diffusion cannot identify by itself

Accepted mean-square displacement: 0.86 µm² over 1 s; 100 independent coordinate increments. The Brownian owner determines the product aN: 3.0071e17 m/mol.

Remove the independent radius above to reveal the family instead of a spurious unique answer. Radius, viscosity, temperature and calibration uncertainty are held exact in the displayed conditional interval. The detailed inference lab supports additional measurement models.

Read Brownian motion §5 · Open the full inference laboratory

What the viscosity correction changes

Jointly inferred solute radius: 4.2968e-10 m. Inferred volume fraction: 0.004 1.

At fixed observations, switching the coefficient from 1 to 2.5 divides the radius by √2.5 and multiplies N by √2.5. This is not a reproduction of the entire historical revision, which also involves the choice of data. The dilute-sphere model is refused above the explicitly chosen 5% volume-fraction ceiling; that ceiling is not an accuracy guarantee.

Setting both concentration and viscosity increment to zero removes the extra information: the solute diffusion then determines only aN, not a unique radius and number.

Bookmark these accepted settings · Inspect the historical radiation calculation · Read light quanta §2

Assumptions, provenance and the code behind the numbers

These are host reference calculations, not FrankenSim/WASM execution. Hydrodynamic spheres, dilute solutions, the Stokes drag law and uniform solvent conditions are idealizations. Solvation, molecular shape, concentrated-solution interactions, measurement error and historical data fitting are not modeled. Shared controls do not make the two sets of illustrative observations a single experiment.

Every displayed quantity comes from one instance-scoped accepted snapshot. Invalid drafts preserve that snapshot. This companion preview does not claim the full catalogue's launch acceptance or source review.

Historical radiation owner · Brownian inference owner · Joint viscosity–diffusion owner · Snapshot composition