Central claims: Kennard

Source set: one document, E. H. Kennard, "On Unipolar Induction: Another Experiment and its Significance as Evidence for the Existence of the Aether" (1917; signed August 1916). Pages 1-9 and, after QE, pages 10-11 (section 7 equations (1) and (2), section 8) were read from page images. All twelve pages were also read in the extracted text and the cleaner Internet Archive OCR by the reviewers and arbiter. There is no source gap for the paper itself; the earlier papers Kennard cites (1912 Phil. Mag.; 1912-13 Phys. Zeits. and Phys. Rev.) are not in the folder, apart from the 1912 and 1917 texts saved by the Investigator.

1. A cylindrical condenser rotated inside a magnetised coaxial solenoid becomes charged as Lorentz's theory predicts, even though a metallic screen rotates with it.
2. The measured electrometer/galvanometer ratio with condenser and solenoid rotating together (one group of four, mean 0.242; runs 0.236 to 0.258) is 5% above the calculated 0.231, and a half-speed group gave 0.228 (1.4% below), using a computed charge of -6.22 V0 (stated accuracy +/-0.5% for the calculation, not for the comparison). NOTE: the -6.22 V0 charge is not independently reproduced. Kennard's printed solution gives his outer-face charge (-7.58 against -7.62 V0) only by holding the inner cylinder at about -0.53 V0, whereas his own stated rule (earthed rotating points sit at Nn, proportional to r^2) gives about +0.054 V0; re-solved with that rule on every rotating conductor, the charge is about -4.0 V0 and the Lorentz ratio about 0.15, so the quantitative agreement is open (QE deep read D1).
3. Rotating the solenoid alone produces no charge on the condenser (reversing average +1.2 mm electrometer, about 1.3 standard errors from zero, against the predicted 14.9 mm and a both-rotating mean of 16.1 mm; the rows have no speed record, galvanometer values +0.5, -0.3, -0.6, +0.2 mm averaging 0), confirming Barnett. If S has the same sense in both blocks of the table, the moving-line theory would predict about -15 mm, the opposite sign to the small residual.
4. The condenser-alone case, not completed because the solenoid short-circuited, is inferred by superposition (both minus solenoid alone), with a partial check of 14 mm electrometer deflection only (no galvanometer ratio).
5. The "moving-line" (relative-rotation) theory is disproved: its electromotive intensity (1/c)[V x B] is not divergence-free (div(v x B) = 2 omega B), so it cannot satisfy Laplace's equation in free aether, and the open-circuit experiments of Barnett, Fehrle and Kennard contradict it. Kennard writes that the theory "seems therefore to be definitely disposed of"; the same argument is already in his 1912 paper.
6. Induction therefore depends partly on absolute rotation; for the radial force on electrons in a rotating condenser, "apparently the only remaining alternative" is motion relative to a stationary Lorentz aether (the passage misprints "square of the distance" for centripetal acceleration, which is proportional to r, and the list of alternatives is incomplete).

Caveats recorded in the deep reads: Kennard calls the results "somewhat rough"; four sets were rejected and about as many abandoned for disturbances; the 13 A and 25 A ratios (0.25, 0.27) are called "not comparable with the previous values" without explanation; two printed arithmetic misprints (0.117 for 0.114; 0.140 for 1.40); the galvanometer swing time is 15 s on the page image. Raw single readings swing by tens of mm before drift removal. Kennard's characterisations of others are his own as a party to the dispute: Barnett's 1912 objection "valid, but seems decidedly weak" (remanence over 25 per cent), his own 1912 inner screen "might conceivably act so as to cut off the effect", and Fehrle's work "marred by several results which must certainly be wrong". Hertz's theory appears only in a p.2 parenthesis (predicts no charge for the rotating condenser).