A Contact Condition for the Fifth Dimension

Everything in the earlier chapters is history and cited work. This chapter is different: a short calculation of mine, checked by hand, that links the Kaluza-Klein form of the metric to the contact geometry this series is built on.

Setup

Take Klein's five-dimensional metric with the fifth coordinate y on a circle. Its off-diagonal part defines a one-form on the five-dimensional manifold:

η = dy + κ A, A = A_μ dx^μ on the 4D base

Differentiating, dη = κ dA = κ F, where F is the electromagnetic field strength pulled back from the base. Then

η ∧ dη ∧ dη = (dy + κA) ∧ κ² F ∧ F = κ² dy ∧ F ∧ F

The second term, κ³ A ∧ F ∧ F, vanishes because it would be a five-form built only from the four-dimensional base. So the top form is κ² dy ∧ F ∧ F.

[RESULT, HAND-DERIVED]η is a contact form on the 5-manifold exactly where F ∧ F ≠ 0. In vector terms F ∧ F is proportional to E·B, so the condition is that the electric and magnetic fields are not orthogonal at that point. The Reeb field is ∂/∂y, the fifth-dimension translation; Kaluza's charge is its conserved momentum.

What this does and does not claim

It does: give a precise condition under which the Kaluza-Klein circle bundle carries a contact structure whose Reeb flow is the hidden circle. It does not: say the physical universe has a contact structure, or that the series' operators C/K/F/U act on it. A pure electric field or pure magnetic field has F ∧ F = 0, so the simplest textbook configurations are not contact.

Grabowska and Grabowski (arXiv:2207.04484) work in contact Hamiltonian mechanics and mention Kaluza–Klein in one sentence of analogy; they do not derive the condition above. Lerman (arXiv:math/0002041) treats contact reduction by circle actions that preserve the contact form, with a contact manifold sitting inside its symplectization. I looked for a source contrasting Kaluza–Klein reduction with contact reduction and did not find one, so the contrast I had drafted here is dropped.

What the literature already has. The mathematical core is standard and not new. The Boothby–Wang construction gives a contact form on a principal circle bundle from a connection whose curvature is symplectic on the base. Hamilton's classification of Boothby–Wang 5-manifolds (arXiv:1001.1953) states it for U(1)-bundles over symplectic 4-manifolds, where “curvature symplectic” means F is nondegenerate, which on a four-dimensional base is the same as F ∧ F ≠ 0. Ding and Geiges (arXiv:1107.4948) treat contact structures on principal circle bundles in general. On the physics side, Goto (arXiv:2310.17925) ties F ∧ F ≠ 0 to symplectic and contact structures for Maxwell fields on three-dimensional slices of Lorentzian four-manifolds with parallel E and B, and Eager, Gary and Roberts (arXiv:1011.5231) write Sasaki–Einstein five-manifolds as U(1) bundles and Kaluza–Klein reduce on them. A delegated search found no source that states the explicit Kaluza–Klein reading, η = dy + κA with contact exactly where E·B ≠ 0, or the Darboux example in Kaluza–Klein language. That search was limited (about eight web searches; abstracts and excerpts, not full papers; Boyer–Galicki and Geiges' textbook not checked), so it is not a claim of novelty. The second pass (Boyer–Galicki–Matzeu, arXiv:math/9811098, Theorem 2.8; nLab on Boothby–Wang) states the same curvature condition in the Kähler and Sasakian setting. For the Lorentzian reading, the closest hit is a 3-dimensional contact condition ω∧dω ≠ 0 for electromagnetic helicity (arXiv:1402.1793), which is not the 5-dimensional E·B condition. I could not reach Geiges' textbook or Boyer–Galicki's book, and I found no source calling Boothby–Wang a Kaluza–Klein reduction. Treat the contact condition as a known fact in new dress.

[NOT DONE]Not checked in Lean. The literature check found the core to be standard (previous paragraph); the Kaluza–Klein/E·B phrasing was not found stated, but absence of hits is not proof of novelty. The Book 4 jet-space chapter's space, J¹(ℝ, ℝ²), is a different object (a rank-2 Pfaffian system, not a contact manifold; Book 4 Chapter 5 now says so); a true contact 5-manifold in that family is J¹(ℝ², ℝ). Nothing here rescues or confirms the Book 4 construction.