Lodge-O: Mathematician review (Step 2)

Reviewer role: PhD mathematician. I check derivations, notation, units and hidden assumptions. I leave mainstream-physics judgement to the physicist and the experimental-evidence case to the contrarian, but I answer a-d for each claim as the prompt requires.

Project: Oliver Lodge, "The Ether of Space", North American Review 187 (May 1908) 724-736 [read: project PDF]. The Investigator then found the missing derivation (Lodge, Phil. Mag. 13, April 1907, pp. 488-506, and the Ether of Space book of 1909, Appendix 2). So there is no source gap on the density derivation any more. I read those passages myself [read: research/files/philmag-1907-ser6-v13-djvu.txt; lodge-1909-ether-of-space-gutenberg40911.txt, Appendices 1-2; thomson-1904-electricity-and-matter-djvu.txt pp.25-27; project PDF p.5 as a page image]. All numbers marked [computation] were recomputed by me with NumPy (constants: c = 2.998e10 cm/s, e = 4.803e-10 esu, m_e = 9.109e-28 g, G = 6.674e-8 c.g.s.).

Source tags: [read] from a document; [web] with URL; [memory] with confidence; [computation] my arithmetic.

---

0. Reader's toolkit (symbols used below)

• rho: mass density of the ether (g/cm^3). a: electron radius (cm). e: electron charge. u: speed of the electron. c: speed of light.
• H: magnetic field strength. w: the speed of the imagined "circulation" of ether that Lodge says a magnetic field is. w0: that circulation speed at the electron's equator. mu: magnetic permeability of free space (equal to 1 in the c.g.s. electromagnetic system).
• "c.g.s.": centimetre-gram-second units. A stress or energy density of 1 dyne/cm^2 is the same as 1 erg/cm^3.
• Energy density of a magnetic field: H^2 mu / 8 pi (Lodge's convention, Gaussian-type units with no stray 4 pi in Biot-Savart).

---

1. What the 10^12 g/cm^3 derivation really is (the central mathematics)

Lodge's recipe, as printed in 1907 section 12 and 1909 Appendix 2 [read]:

1. A moving electron of radius a, charge e, speed u has magnetic field H = e u sin(theta) / r^2 outside it (Biot-Savart; theta is the angle from the direction of motion).
2. Lodge reads the field as a "circulation" of ether at speed w along the field lines, with kinetic energy density equal to field energy density: (1/2) rho w^2 = mu H^2 / 8 pi.
3. Hence w / u = sqrt(mu / 4 pi rho) e sin(theta) / r^2, and at the equator (r = a): w0 / u = sqrt(mu / 4 pi rho) e / a^2.
4. Integrating the kinetic energy over all space outside the sphere gives (4/3) pi rho a^3 w0^2. Setting this equal to (1/2) m u^2 gives m = (8/3) pi rho a^3 (w0 / u)^2.
5. He assumes w0 = u ("most reasonable ... same order of magnitude"). Then e = a^2 sqrt(4 pi rho / mu), and m is twice the mass of an ether sphere the size of the electron.
6. Inserting "diameter 10^-13 cm, mass 10^-27 g" gives rho of order 10^12 g/cm^3.

1.1 Algebra check: correct [computation]

I redid the integral. With w = w0 a^2 sin(theta) / r^2 and volume element 2 pi r^2 sin(theta) dr dtheta:

• radial integral: integral from a to infinity of r^2 / r^4 dr = 1/a
• angular integral: integral of sin^3(theta) from 0 to pi = 4/3
• energy = (1/2) rho w0^2 a^4 * 2 pi * (1/a) * (4/3) = (4/3) pi rho a^3 w0^2. This matches Lodge.

Lodge's side remark that the hydrodynamic estimate (sphere in perfect fluid: effective mass 2 pi rho a^3) agrees "if w0 = (sqrt 3 / 2) u" is also arithmetically right: (8/3)(3/4) = 2. The "2 pi rho a^3" figure itself is 1.5 times the sphere's volume times rho: the electron's own mass (rho times volume, taken as ether inside it) plus the standard added mass of half the displaced fluid [memory, high: textbook result for a sphere in ideal fluid].

1.2 The hidden fact: rho cancels out of the inertia

Substitute step 3 into step 4:

m = (8/3) pi rho a^3 * mu e^2 / (4 pi rho a^4) = (2/3) mu e^2 / a.

The ether density drops out. This is the classical electromagnetic mass of a charged shell, m = 2 e^2 / (3 a c^2) in Gaussian units (mu = 1/c^2) [memory, high; the classical-electron-radius result]. The energy argument fixes m in terms of e and a only. Density enters solely through the extra assumption w0 = u, which is what converts "m, a" into a number for rho.

So the physical content of "rho = 10^12" is:

rho = mu e^2 / (4 pi a^4) * (u / w0)^2.

With w0 = u this is a number set by e and a alone. It is not a separate measurement of anything. The assumption w0 = u was chosen by Lodge ("a relation suggested, though not enforced, by gyrostatic analogies" [read: 1909 App. 2]). If the circulation at the surface were 10 times faster than the electron, rho would be 100 times smaller; if 10 times slower, 100 times larger. Lodge says only that "a reduction ... would only go to increase" rho. That is true, but the opposite direction (w0 > u, lower rho) is excluded only by his remark that "it is difficult to suppose" it. That is an argument from plausibility, not a bound.

1.3 Sensitivity to the size of the electron [computation]

The electron's radius is not measured; it is whatever makes the field-energy mass equal to m_e. Using rho = (e/c)^2 / (4 pi a^4) (w0 = u, EMU with e in abcoulomb = e_esu / c) and for comparison the plain ratio m_e / volume:

Three things follow.

1. 10^12 is reached only with Lodge's choice of the smaller radius. Using the textbook classical radius gives about 10^10. Between these, the "order of magnitude" moves by two to three powers of ten. Lodge's sentence that "the estimate ... cannot differ as regards order of magnitude" holds only for a fixed radius [read: N. Am. Rev. p.14].
2. Density scales as a^-3 to a^-4. Scattering experiments later found no electron size down to very small scales (bounds below about 10^-16 cm are quoted; [memory, medium]). If a is taken to zero, rho diverges. This is the same divergence as the infinite self-energy of a point charge. The "density" is therefore a measure of how hard we pack the electron's field energy into an assumed ball, not an independent property of space.
3. The two famous numbers are the same number. The last column: the electromagnetic energy density at the electron's surface is about 1.5e33 erg/cm^3. This equals rho c^2 (about 9e32 for rho = 10^12). So Lodge's "ether density 10^12 g/cm^3" and "ether elasticity 10^33 c.g.s." are the electron's surface field energy density written in mass units and energy units. The step from density to elasticity (below) is the identity c^2 = elasticity / rho, which holds by definition once light is a wave in the medium. It is not a second line of evidence.

Where Lodge extends this value "everywhere in space" he adds a new assumption (uniformity of the ether, incompressible so "the density of fluid inside and outside an electron must be the same" [read: 1909 App. 2]). The calculation itself only describes the neighbourhood of an electron.

Verdict on the derivation (answer a): internally consistent algebra; conditional on (i) a chosen electron radius, (ii) w0 = u, (iii) extension to all space. The number is a definition-dependent estimate, not a determination.

1.4 Elasticity, mass and energy per cubic millimetre [computation]

• Elasticity: rho c^2 = 1e12 * (2.998e10)^2 = 9.0e32, i.e. "of order 10^33 c.g.s." Correct, via c^2 = elasticity / rho. Lodge's own text confirms the 10^33 dyne/cm^2 figure [read: 1907 section 12; 1909 App. 1].
• Mass of 1 mm^3: 1e12 g/cm^3 * 1e-3 cm^3 = 1e9 g = 1,000 metric tons. Correct.
• Energy of 1 mm^3: rho c^2 * 1e-3 cm^3 = 9.0e29 erg = 9.0e22 J = 2.5e16 kWh.
• Lodge states: 1907: "3 x 10^17 kWh ... a million-kilowatt power station for thirty million years" (the exponent is garbled in the OCR as "10!", but the stated equivalent fixes it: 1e9 W * 30e6 yr = 9.5e23 J = 2.6e17 kWh) [read: 1907 section 12; computation]. 1908/1909: "million horse-power for forty million years" = 9.4e23 J [read: N. Am. Rev. p.11; 1909 text; computation].
• Both are about 10 times the energy that follows from his own 10^33 erg/cm^3. Check: 10^33 erg/cm^3 = 10^30 erg/mm^3 = 10^23 J = 2.8e16 kWh, i.e. the correct stations would run about 3 million years (million kW) or 3.8 million years (million hp), not 30 and 40 million. The Investigator's lead (the factor of 10 is already in the 1907 paper) is confirmed; the cause is an arithmetic slip of about a factor 10 that carried unchanged from 1907 into 1908 and 1909. The "100 foot-lbs per atomic volume" line is consistent with 10^33 erg/cm^3 (74 ft-lb for 10^-24 cm^3) [computation], so the slip is in the kWh/station conversion.
• Kragh's quoted "10,000 tons per cm^3" [web: https://arxiv.org/abs/1111.4623 via Investigator] is 100 times below 1e12 g/cm^3 = 1e6 tonnes/cm^3 [computation]. Lodge's own "thousand tons per mm^3" is right.

1.5 Consequences Lodge draws: all reproduce [computation]

These come from the 1909 Appendix 2 and are a good check that the model hangs together. They all follow from rho = 10^12 by simple formulas.

Important consequence for claim 3 and the experiment: Lodge's model predicts a drift of only 4e-4 cm/s at 1400 gauss. His 1893 null result allowed up to 30 cm/s. So the experiment was about 75,000 times too insensitive to see the effect predicted by the density he later adopted. The null result is compatible with rho = 10^12, but it says only rho > 173 g/cm^3. It does not support 10^12 [computation].

---

2. Claim-by-claim review

Claim 1. No action at a distance; all strain, cohesion and potential energy live in a continuous ether, which is absolutely stationary.

a. Mathematics. This is a philosophical and structural claim, with little mathematics in the article. One internal tension is real. Claim 2 (and the Thomson quote Lodge cites) says mass is "mass of the ether ... carried along by the Faraday tubes", with the ether's kinetic energy being the energy of motion of the carried ether. Claim 1 says the ether is "the most stationary body in existence" and "cannot be moved by mechanical means" [read: 1907 section 13; 1909 App. 2]. The inertia calculation in section 1 above treats the ether near an electron as flowing at speed w about it. A medium cannot both circulate with speed w0 = u near a moving electron and have zero motion. Lodge reconciles this by saying it is a pattern that moves (the electron is a singularity in the ether, not a lump of it), and by treating the "circulation" as hypothetical. But the energy and momentum books are kept with real fluid velocities, so the reconciliation is not carried out mathematically. Larmor's remark that a dense ether "would require the ether to be absolutely stationary, except in a magnetic field, in order to avoid hydrodynamical systems of force between moving bodies" [read: 1907 App. A, quoting Larmor, Phil. Trans. 185 (1894)] explains why Lodge needs stationarity: a body moving through a dense perfect fluid would feel strong forces unless it did not displace the fluid. Larmor's later suggestion (Investigator lead, 1907 App. B, exponent garbled in OCR) that circulation is shed as a wake is an argument, not a calculation.

A positive technical note: a uniformly moving sphere in an ideal fluid feels zero drag (d'Alembert's paradox) [memory, high], which helps Lodge for steady motion, though not for accelerations (the added mass is a real force).

b. Experiments. Absence of a detectable ether wind: Michelson-Morley 1887 and successors [memory, high]. A stationary ether predicts a positive fringe shift unless length contraction is added (Fitzgerald, Lorentz) [memory, high]. Lodge's own spinning-disc experiments (1893, 1897) found no dragging of the ether by rotating matter, which he treats as showing it is stationary relative to matter's motion [read: N. Am. Rev. p.9; Phil. Trans. 184 (1893) and 189 (1897) per Investigator]. Local field theories (Maxwell's equations, with fields acting locally) are Established; the ether as a substance with its own state of rest was dropped because relativity gives no preferred frame [memory, high].

c. Tested? Lodge himself (disc experiments, 1893-97 [read: 1909 ch. V per Investigator; Phil. Trans. 189 abstract]); Michelson and Morley (1887) and later repeats [memory, high]. The disc experiment as a separate geometry has no independent repeat that I could find; Hunt (1986, HSPS 16:111-134) is the history and was not opened (paywalled) [Investigator].

d. If true. Special relativity's rejection of a preferred frame would have to be reinterpreted. Lorentz-ether theory is empirically equivalent to relativity for these tests [memory, high], so the "stationary ether" claim cannot be separated from relativity by any experiment on record.

Verdict (split). - "No action at a distance; effects propagate through local fields": Established (as field theory, without a material medium). - "Matter acts via a continuous, absolutely stationary material ether": Historically superseded but instructive. Not refuted by mathematics, and no experiment on record can distinguish it from relativity-plus-fields. The internal tension between "stationary" and "carries the momentum" is left unresolved by Lodge and noted as a gap.

Claim 2. Matter is made of ether: the electron is a singularity of a uniform-density ether; all mass, momentum and kinetic energy are the ether's.

a. Mathematics. Section 1 is the supporting calculation. Findings: - The inertia derived is the electromagnetic mass (2/3) mu e^2 / a with rho eliminated (1.2). Thus the maths supports "some of the electron's inertia could be field energy". It does not require an ether fluid. - The "three methods agree" claim (1907 section 12) is agreement of coefficients of order unity once w0 is tuned (3/4 vs 1). Weak confirmation. - The classic problem that electromagnetic energy and momentum give different masses (the "4/3 problem": mass from energy is 3/4 of mass from momentum) is not noticed. Lodge uses the energy route only. [memory, high; this is the known weakness of the Abraham-Lorentz electromagnetic-mass programme.] Within a factor of order one he is fine. - "Singularity of uniform density" needs a definition of the pattern (a vortex knot? a source?) and none is given; Lodge's 1920 note says "how the ether can be tied into the knots which we call electrons ... remains to be discovered" [read: Phil. Mag. 39 (1920) 170-174, per Investigator].

b. Experiments. Mass-velocity measurements of fast electrons (Kaufmann 1901-06, Bucherer 1908) were then read as supporting purely electromagnetic mass; later relativistic dynamics fit them [memory, high]. Electron scattering shows no structure at the size Lodge assumes [memory, medium]. In the Standard Model the electron's rest mass is attributed to the Higgs coupling, not to field energy [memory, high].

c. Tested? The electromagnetic-mass hypothesis was tested by Kaufmann and others around 1901-08 with ambiguous results at the time, then superseded [memory, medium]. The ether-fluid version was not tested (no distinguishing measurement exists). Reasoning by Thomson (1904) and Lodge (1906, Electrons) is the extent of the work.

d. If true. Rest mass would be pure field energy; the Higgs mechanism would play no part for the electron; a point-like electron would not exist.

Verdict (split). - "Part of an electron's inertia is the inertia of its own field (electromagnetic mass)": Historically superseded but instructive. A reasonable position in 1904-08; the algebra is correct. - "The ether is a literal fluid that carries the momentum, and the electron is a knot in it": Not testable as stated (no measurable difference from field energy and momentum, which the formulas already give).

Claim 3. Ether density about 10^12 g/cm^3; elasticity about 10^33; whirl speed about c; one cubic millimetre equals 1,000 tons and 40 million years of a million-hp station.

a. Mathematics. Details are in section 1. Summary: - Derivation of 10^12: algebra correct; value depends on the chosen radius and on w0 = u; density cancels in the inertia (so the density is not independently determined). - Elasticity 10^33 = rho c^2: correct and by definition. - Mass per mm^3: correct. - Energy per mm^3: a factor of about 10 too large (1.4); this is an arithmetic slip repeated in 1907, 1908 and 1909. Lodge's "10^33 erg/cm^3" is itself right for rho c^2. - Whirl speed about c: the cord analogy is correct: on a loop of cord moving along itself at speed v with tension purely from the motion, T = lambda v^2, and transverse pulses travel at sqrt(T/lambda) = v [memory, high]. But applying it to a three-dimensional medium needs an actual model. A plain rotating fluid supports "inertial waves" whose frequency depends on direction and rotation rate (omega = 2 Omega cos(angle)), independent of wavelength; these do not give a fixed speed c for all colours [memory, high]. Kelvin's gyrostatic and vortex-sponge ethers do have a wave speed, and the Investigator notes Kelvin's laminar vortex ether has wave speed about 0.47 of the turbulent speed [Investigator lead; not checked by me]. So "of the same order as c" is plausible in order of magnitude but not shown. - Units problem in the "disruption" argument (project PDF p.5, read as page image): Lodge compares the Newtonian "ether tension" with a disruptive stress "10^33 dynes per square centimetre". In Appendix 1 of the 1909 book, T = gamma M / r, which is the gravitational potential, with units cm^2/s^2 (energy per gram), not dyne/cm^2. Its value at Earth's surface (6.2e11) is correct as a number: g R = 981 * 6.37e8 [computation]. Comparing T to a stress requires multiplying by a density: rho_ether * T. With rho = 10^12, "disruption" at 10^33 would correspond to T about 10^21, about c^2 (9e20), which is the free-fall-at-light-speed condition Lodge also quotes. As printed, he sets the threshold at T = 10^33, which is about 10^12 times too high. The dimensionless quantity is T / c^2 (7e-10 at Earth's surface [computation]). This is a genuine unit inconsistency, though it affects only the speculative gravitational-disruption aside. - Also in Appendix 1: "a globe of the density of the earth would have to have a radius of eight light years" to reach T = 10^33. I get 2.55e19 cm = 27 light years = 8.3 parsecs [computation]. Lodge's other figures in that appendix (free fall at c at about the earth-sun distance; 10^13 and 10^7 light years for the visible-universe density) all check [computation]. So "eight light years" is probably "eight parsecs" mislabeled.

b. Experiments. The ether density has no direct measurement; Lodge's tests cannot reach it (1.5). The only observed quantity involved is the speed of light. The energy density 10^33 erg/cm^3, if the vacuum gravitated, would be about 10^41 times the observed dark-energy density (about 5e-9 erg/cm^3) [memory, medium: derived from Lambda-CDM critical density 9e-30 g/cm^3 times 0.7 times c^2]; Lodge avoids this by saying the ether is "not felt mechanically", i.e. by assumption, not calculation. Kragh and others (Investigator) link the figure to zero-point-energy estimates, which are famously far larger than the observed vacuum energy [memory, high]. That is a resemblance in size, not a derivation.

c. Tested? The magnetic-drift null experiment (1893, unpublished until 1907) set rho > 173 g/cm^3 and cannot reach 10^12 (1.5). Lodge's 1920 restatement (10^30 to 10^33 erg/cm^3 with 10^-30 to 10^-33 cm grain [Investigator]) is the same number. No replication of any test of the density exists; there is no way to measure it. Critics' responses: Larmor (letter in 1907 appendix, accepted a large density and suggested more); FitzGerald (1896) thought "greater than platinum" with turbulent energy about 10^22 erg/cm^3 [Investigator]. Neither is a test.

d. If true. The vacuum would carry an energy density 10^40 times what cosmology allows unless it is decoupled from gravity; quantum field theory's treatment of the vacuum would have to be replaced by a mechanical medium.

Verdict (split). - The arithmetic from rho = 10^12 to 10^33 erg/cm^3 and 1,000 tons/mm^3: Established as arithmetic. - The value 10^12 g/cm^3 as a property of space: Not testable as stated. It is the electron's surface field energy divided by c^2, dependent on an assumed radius (spread of 10^9.5 to 10^12.5) and on w0 = u. - The "forty million years of a million-hp station": Internally inconsistent with Lodge's own 10^33 erg/cm^3 by a factor of about 10 (should be about 4 million years). - The "disruptive stress" comparison of gravitational potential with a stress: Internally inconsistent (units).

Claim 4. The ether spins near a charge and a magnetic pole with moment of momentum e m (J. J. Thomson); Faraday's rotation is the gyrostatic evidence; Current, Magnetism and Motion form three perpendicular vectors (E, H, V) that may underlie the three dimensions of space.

a. Mathematics. I checked each part.

• Thomson's e m. The source (Electricity and Matter, 1904, pp.25-27 [read]) takes the momentum density N B sin(theta) (tubes times induction), notes it points along circles about the line joining a point charge and a pole ("a top spinning around AB"), and states the total moment of momentum is "e m", independent of how far apart they are. In modern Gaussian units the field angular momentum is L = e g / c, with e the charge and g the monopole strength [memory, high]. I verified numerically that the integral of r x (E x B)/(4 pi c) over all space for a charge and a pole 0.5, 1 and 2 units apart gives 0.997, 0.997 and 0.994 times e g / c (grid error under 1%; exact value 1) [computation]. So the result is correct and the independence from separation is right. This was reproduced for the 1904 text and the claim in the 1908 article.
• Units: Lodge says the charge is "measured electrostatically and the pole magnetically". In Gaussian units L = e_esu g_emu / c, so with the charge in electrostatic units and the pole in electromagnetic units a factor 1/c remains; "e m" exactly holds when both are in one consistent system with c absorbed (for example, charge in electromagnetic units). Step 1's deep read said leaving out c "is consistent with mixing units"; mixing units does not remove c, it requires it (see Corrections). This does not hurt the physics but it is a units slip in the sentence as printed.
• Learner's worked example: Dirac's condition says L = e g / c = n hbar / 2 for integer n. With g = (137/2) e for n = 1: e^2 * 68.5 / c = 5.27e-28 erg s = hbar / 2 [computation]. This shows that "spinning momentum e m" is the origin of the later Dirac monopole quantization [memory, high].
• "Excessive slowness". If the field momentum is carried by ether of density rho, the velocity is momentum density divided by rho; with rho = 10^12 this is tiny (compare 6.7e-23 cm/s above). Consistent with his statement.
• Gyrostatic argument. A spinning top responds to a force at right angles. Lodge argues the charge-pole system "sets off at right angles to the force", hence hidden spin. Mathematically this is the conservation of the total angular momentum (field plus mechanical), as Thomson's derivation uses: moving the charge changes the field's angular momentum, so the pole and charge must gain equal and opposite momentum at right angles [read: Thomson 1904 pp.26-28]. The analogy is sound as a restatement; it adds no new equation.
• Three perpendicular vectors. For a current of strength I in a field B, the force is I L x B (Fleming's rule); EMF = v x B; Poynting flux E x H [memory, high]. Each is a cross product, which gives a vector perpendicular to the two inputs. The statement "any two imply the third" is true as direction-and-magnitude bookkeeping only. A note for the reader: the cross product of two vectors as a vector exists only in 3 and 7 dimensions [memory, high], which gives a genuine mathematical grain behind Lodge's remark about three-dimensional space, although the physics link he draws (EM and dynamos imply three dimensions) is not derived. Lodge never defines E, H, V and says "E" for Current, so the symbols are not the standard electric field.

b. Experiments. Faraday's 1821 rotation of a pole about a current-carrying wire (and vice versa) is real [memory, high], and the standard account is the Lorentz force; Lodge himself admits "usually worked out in another way" [read]. Field angular momentum of a charge-and-monopole system is standard textbook electromagnetism (the Feynman-type paradox, Dirac quantization) [memory, high]. No experiment is cited or exists that detects ether spin separately from field angular momentum. Magnetic monopoles have never been observed [memory, high], so the charge-pole case is a thought experiment; a pole fixed in a magnet next to a moving charge is the realizable version.

c. Tested? The cross-product facts: tested continuously (dynamos, motors). Thomson's e m: derivation only; the quantitative form corresponds to standard results. "Ether spin": untested and with no distinguishing prediction.

d. If true. Nothing in mainstream electromagnetism changes if the field angular momentum is called ether spin. If a separate fluid spin existed with a different value from e g / c, it would contradict Maxwell's equations with the Lorentz force.

Verdict (split). - Field angular momentum of a charge and a pole is e m (in the right units) and is independent of separation: Established (verified by me, with a units caveat). - The E, H, V perpendicular-vector rule as stated for dynamos and motors: Established. - That this is evidence of a spinning ether fluid: Not testable as stated. - That it underlies the three dimensions of space: Not testable as stated (mathematically suggestive only in the sense of the 3-dimensional cross product).

Claim 5. A moving fluid can imitate solid rigidity and elasticity (Kelvin's kinetic elasticity), so a perfect-fluid ether can make solid matter; Lodge's own spinning-disc test found no effect on light.

a. Mathematics. Kinetic elasticity: a fluid with internal motion can transmit transverse disturbances, shown by the cord analogy and by Kelvin's gyrostatic medium [memory, high]. This is mathematically fine. What is missing is a model that reproduces the specific properties of light (non-dispersive transverse waves, no longitudinal waves, Maxwell's equations). Nothing in the article does so, and the inertial-wave dispersion above shows that an ordinary rotating fluid fails to be non-dispersive. Lodge's 1909 treatment of Fresnel's law as a special case (Appendix 3, Investigator) was not reviewed by me in detail.

Disc experiment arithmetic: a disc one yard (0.914 m) in diameter at 4,000 rpm has a rim speed of 191 m/s [computation]. Lodge's 1909 bound is that the speed of light between the discs is unchanged by more than 1/1000 of the disc speed (that is, about 0.19 m/s) after tightening from 1/200 in 1893 [Investigator, from 1909 book ch. V]. The 1897 version used about 3,000 rpm, a ton of iron, and magnetization [Investigator, Phil. Trans. 189 abstract].

b. Experiments. The disc result agrees with the general null results for ether drag (Michelson-Morley 1887 and descendants) and with relativity, where moving matter does not drag a medium [memory, high]. The Fizeau experiment (1851) finds partial dragging of light by moving water with coefficient 1 - 1/n^2 [memory, high], which Lodge's Appendix 3 treats via the loading fraction (mu^2 - 1)/mu^2 [read]. A steel disc in vacuum or air is not expected to show a large effect in mainstream theory either.

c. Tested? Yes for the dragging question: by Lodge himself, 1893 and 1897, with a null result, never independently repeated in that geometry as far as I found. Not tested: whether a vortex or turbulent fluid can reproduce electromagnetism, since no such model has been shown to work.

d. If true. A fluid-mechanical derivation of Maxwell's equations from a vortex medium would be needed; mainstream physics took the field as fundamental instead.

Verdict (split). - The disc experiment (no ether drag from rotating matter at the 1/1000 level): Tested, not supported for the idea that spinning matter drags the ether; consistent with every other null result. - "A perfect-fluid ether can make solid matter and light by kinetic elasticity": Never properly tested (no explicit model was ever built that reproduces Maxwell's equations plus the electron; this was not dismissed without testing so much as left as a program; Lodge conceded in 1920 that how the knots are tied "remains to be discovered").

Claim 6. Matter is extremely rare compared with ether (visible cosmos about 10^-23 of water density), so the ether being immensely denser is consistent; the ether may have psychical significance.

a. Mathematics. Kelvin's numbers: parallax 1/1000 arcsec means distance 1,000 parsecs = 3.09e16 km; sphere volume 1.13e59 m^3; mass 1.5e36 tons gives density 1.34e-23 g/cm^3 (1.6e-23 in the 1909 appendix; both "of the order 10^-23") [computation; consistent with deep read]. 1e12 / 1.34e-23 is about 7e34, so the ether is about 10^35 times denser than the "visible cosmos". The comparison of vacuum to lead (both 10^12) works [deep read, recomputed]. Kelvin's solar mass of 1.5e27 tons is about 25% low versus the modern 2e27 [memory, high]. All this is fine arithmetic.

An important point of context. A sphere of radius 1,000 pc is not "the visible cosmos" in any modern sense; it is the Sun's neighbourhood in our Galaxy. The local density of stars is about 0.1 solar masses per cubic parsec [memory, high], which gives roughly 7e-24 g/cm^3 [computation]. So Kelvin's number is a good local galactic density, but the cosmic mean density of ordinary matter is much smaller (about 4e-31 g/cm^3 [memory, high]). This does not hurt Lodge's argument (rarer matter only makes the ratio more extreme) but the label "cosmos" is a misnomer.

The logic: "ether denser than matter, therefore consistent with matter being a thin aggregate in it" is a restatement of the assumption that matter is made of the ether (claim 2). The Hooke quote (1705) is a statement of the same assumption, as Lodge himself says ("must have been only a speculation") [read: p.13-14].

b. Experiments. The density of ordinary matter is measured; the ether density is not. Psychical significance: Lodge's own addition, tied to his psychical-research work (Raia 2007, Noakes 2005 per the Investigator, not read).

c. Tested? The astronomical numbers: measured (and revised). The psychical idea: no test is described.

d. If true. For the density: nothing beyond claims 2-3. For psychical significance: a physical basis for mind would be needed.

Verdict (split). - Arithmetic of the matter density and the ratio to the ether figure: Established as arithmetic; the ratio inherits the weakness of claim 3's number. - "The ether has psychical significance": Not testable as stated.

---

3. Investigator's findings: what I checked, what I keep

I read the key sources myself where they bear on the mathematics. All pertinent Investigator leads stay in the record whatever my verdict.

1. The missing derivation (Phil. Mag. 13, 1907; book App. 2): read and verified (section 1). It removes the Step 1 caveat that the density figure rests on the article alone. It also shows the derivation's real structure (1.2), which the project article hides.
2. Energy figure: Investigator's note that the factor of 10 is already in 1907 is correct (1.4). Not an error of transcription in the 1908 article.
3. Unpublished 1893 experiment: arithmetic of the "eight times platinum" bound confirmed (1.5). It is a true null with a lower bound only; the later density value is five orders beyond what it could reach. The 13-year delay (absence-of-evidence item): competing explanations the Investigator lists (null result judged unpublishable; waiting for a theory that gave it meaning; loss of details). My arithmetic supports the second in one respect: once Lodge had derived 10^12 g/cm^3 in 1907 he could say the experiment "ought to be recorded" because the result was consistent, but it constrains nothing beyond 173 g/cm^3. The Liverpool notebook of November 1893 would decide.
4. Tighter disc bound (1/1000 of the disc speed): arithmetic consistent (191 m/s rim speed). Plausible; not independently recomputed from the interferometer geometry (I did not read Phil. Trans. 184).
5. Gravity as ether tension: the T = gamma M / r relation is correct as a potential (6.2e11 at Earth), but its comparison with a stress threshold is a units inconsistency; "eight light years" should be 8 parsecs (3.1). The Investigator's resolution of the garbled "108S" on p.5 as 10^33 is confirmed by my page image (p.727: "10^33 dynes per square centimetre").
6. Lodge 1920: grain size 10^-30 to 10^-33 cm and energy 10^30 to 10^33 erg/cm^3; refractive index 1 + 2 gamma M / (r c^2) [Investigator]. The weak-field light-bending index n = 1 + 2GM/(rc^2) is the standard linear approximation in an isotropic-coordinate form of general relativity for the "gravitational index of refraction" that gives half the Einstein deflection [memory, medium]. Lodge's claim to produce "Einstein's deflexion" from it is, in this form, a factor-of-two question I did not resolve. I did not verify the factor (a medium-confidence point; no verdict depends on it).
7. Thomson 1904 pp.25-27: read and confirmed; my numerical check of e g / c supports the use of "e m" (3, claim 4).
8. Larmor and FitzGerald precedents: dense ether "contemplated with equanimity" (Larmor 1894), "greater than platinum" (FitzGerald 1896) [read: 1907 App. A, via Investigator]. Lodge is not alone in thinking a dense ether possible, but these are statements of possibility, not measurements.
9. Related work: zero-point energy lineage (Kragh; Rowlands 1990 quote "infinite energy density ... almost indistinguishable from the infinite elasticity of the universal ethereal medium"); spacefed.com (an interested party advocating zero-point-energy technology; its characterization of Lodge's figure as an early vacuum-energy estimate is theirs). Mathematically: the 10^33 erg/cm^3 is an energy density; the zero-point comparison is about size, not method. Polarizable-vacuum gravity (Puthoff K = exp(2GM/rc^2)) [memory, medium; Investigator, not checked on the web]: first order equals Lodge's index, but not checked further.
10. Contemporary reception (Science 26:482; Sci. Am. Suppl. 64:212; Nature 75:519 "a billion times the density of water", "nothing paradoxical, nor ... improbable") and the history (Hunt 1986; Raia 2007; Noakes 2005; Cambridge biography chapter; Birmingham and Liverpool archives; Chronicling America blocked): not read by me; kept as open leads for the physicist, contrarian and later readers.

---

4. Corrections to Step 1

1. claims.md header: "The derivation ... is not in the folder and was not read (the density figure therefore rests on this article's statement alone)." Superseded: the derivation was found and is checked above. The density figure rests on an assumed radius and w0 = u (1.2-1.3); it does not rest on the article alone, but it is also not independently determined.
2. claims.md note on the energy figure ("about 10 times rho c^2 ... source unknown"): the factor is a conversion slip present from 1907; the source of the figure is Lodge's 1907 sentence (10^33 erg/cm^3 equals "3 x 10^17 kWh") whose conversion is about 10 times too high.
3. Deep read, section 1 (Thomson's e m): it says Lodge's e m "leaves out the factor c, which is consistent with mixing electrostatic and electromagnetic units". That is not correct. Mixing units does not remove c. In Gaussian units L = e g / c; with e in electrostatic and the pole in electromagnetic units a 1/c remains. "e m" holds exactly in a consistent unit system where c is absorbed (for example, charge in electromagnetic units). Thomson's text (read) does not state the units.
4. Deep read, section 3 (reader's check of density): it gave m_e / volume of about 2e11 for a = 1e-13 and 3e10 for a = 1.9e-13 and called them "within about one to two orders of 10^12". I confirm those figures, and add that Lodge's 10^12 requires the diameter (not radius) 1e-13 cm, that is a = 0.5e-13, and that the full range with the classical radius is 10^9.5 to 10^12.
5. Deep read, section 4 (apparent tension between "excessive slowness" and "whirl speed about c"): resolved partly by the 1909 text: the slow circulation is the drift (momentum divided by rho), a different quantity from the intrinsic turbulent speed. The 1909 Appendix 2 says "close to an electron the speed of magnetic drift ... may rise to something near the speed of light; say 1/30th", falling to a millimicron per century at 1 mm. These are consistent with each other and with my formula w = H / sqrt(4 pi rho).
6. claims.md claim 6 calls Kelvin's number "visible cosmos". It is a sphere of radius 1 kpc (1,000 pc), the Sun's galactic neighbourhood, so it should be flagged as "Kelvin's 1,000-parsec sphere", although Lodge uses "visible cosmos".
7. sites.md lead 1: "w0 is at least the electron's speed" sentence: the lead says "Lodge says a smaller w0 would make rho larger". Correct, but the Investigator did not note that w0 larger would give smaller rho, which is the unbounded direction (1.2).
8. Deep read, p.5: now resolved by page image: the threshold is "10^33 dynes per square centimetre" (journal p.727), footnote: "the necessary concentration absurdly great".

---

5. Summary table

6. For the human reader: what to take from this

• The best way to learn from this project: reproduce the one-page integral in 1.1 and watch rho cancel out. The result is the classical electromagnetic mass; the "ether density" appears only when you add the choice w0 = u.
• The 10^12 number is not a measurement. It moves by a factor of 1,000 with the choice of electron radius, and by any factor with the assumption about w0.
• The best-supported parts of the article are the unglamorous ones: Thomson's field angular momentum e g / c, the cross-product rule, the Kelvin arithmetic, and the 1893 bound (rho > 173 g/cm^3).
• Web checks: none needed for verdicts. The medium-confidence [memory] items (electron size bounds, the factor-of-two light bending, the dark-energy figure) are marked as such and no verdict depends on them; I did not run web calls in this review.