# A primer on foliations, singular points, and Deninger’s dynamical picture of zeta zeros

*Written 13 November 2026 (Pacific time) for the Zeta Zeros site. Grant is still learning this material, so the tone is plain. Every claim is labelled.*

**How to read the labels.**

- **[known]** is published mathematics, with the source named. Short keys such as [D-2008] or [CC] are expanded in [`sources.md`](sources.md).
- **[project]** is something we computed or drew ourselves. A [project] picture is an illustration of a known fact, not a new theorem.
- **[speculative]** is a proposed picture that has not been constructed, or a reading of one. Deninger himself calls the existence of a foliated dynamical system for schemes over ℤ an “optimistic guess.”

**Nothing in this note is progress on the Riemann hypothesis.** The Riemann hypothesis is a precise statement about the zeros of ζ(*s*). A dictionary that *would*, if a certain space existed, recast that statement as a positivity or Hodge-type condition, is not a proof, and is not evidence.

This note has three parts. Part I is ordinary differential topology: what a foliation is, and what a singular (zero) point of a foliation looks like. Part II is Deninger’s proposed dictionary between arithmetic and flows on foliated spaces. Part III says carefully what has actually been built, and what is open.

---

## Part I. Foliations

### 1. The idea, in one paragraph

**[known]** [CC], [CLN], [Hae]. Imagine a manifold — a space that looks locally like Euclidean space of some dimension *n* — cut into a family of thinner pieces, all of the same dimension *p*, that fit together smoothly and never cross. Each piece is called a *leaf*. A *foliation* of dimension *p* (equivalently, of *codimension* *n* − *p*) is precisely such a partition, with a compatibility condition in charts: around every point one can find coordinates (*x*₁, …, *x<sub>p</sub>*, *y*₁, …, *y<sub>n−p</sub>*) in which the leaves look like the level sets “*y* is constant” — a stack of parallel *p*-dimensional plaques. Changing charts is allowed to shuffle the plaques, but it must preserve the stack: plaques go to plaques.

The leaves need not be compact, and they need not be embedded as closed subsets. A single leaf may wind densely through the ambient manifold. That already happens on a torus, as we will see.

A useful picture: the coordinate *y* is a *transverse* coordinate, and the *x*’s run along the leaf. Holonomy — how plaques are identified when you travel around a loop in a leaf — is the germ of the transition maps in the *y* coordinates. We will not need the formal theory of holonomy here; Candel–Conlon [CC] and Camacho–Lins Neto [CLN] develop it carefully, and Haefliger’s 1958 paper [Hae] is the source of the modern classifying language.

### 2. Three standard examples

#### 2.1 Foliations by level sets

**[known]** If *f* is a smooth real-valued function on a manifold *M* whose derivative *df* never vanishes, the level sets *f*⁻¹(*c*) are a foliation of *M* of codimension one. The leaves are the ordinary level surfaces of *f*. (This is the regular level-set theorem, plus the observation that the plaques in a chart can be taken as the slices of constant *f*.)

If *df* is allowed to vanish, the picture changes. The zeros of *df* (or of a vector field, below) are *singular points* of the would-be foliation: the partition into regular level sets no longer covers the whole manifold, and near a critical point the level sets can change shape. A standard first picture is *f*(*x*, *y*) = *x*² + *y*² on ℝ²: away from the origin the levels are circles, and the origin is a singular fibre.

#### 2.2 The Kronecker (linear) foliation of the torus

**[known]** [CC, Ch. 1]. Let *T*² = ℝ² / ℤ² be the two-torus, drawn as the unit square with opposite sides glued. Fix a slope *α* ∈ ℝ, and consider the lines in the plane of slope *α*, then descend them to the torus. This is a foliation of *T*² by one-dimensional leaves.

- If *α* = *p*/*q* is rational (in lowest terms), every leaf is a closed circle, winding *q* times one way and *p* times the other before closing.
- If *α* is irrational, *every* leaf is dense in *T*²: a single line comes arbitrarily close to every point, infinitely often, and never closes. This is Kronecker’s classical density theorem for the linear flow on the torus.

The same picture in dynamical language: the vector field ∂/∂*x* + *α* ∂/∂*y* on *T*² has no zeros, and its orbits are the leaves. The flow is *φ<sup>t</sup>*(*x*, *y*) = (*x* + *t*, *y* + *α t*) mod ℤ².

This example matters later. A foliation can have a dense leaf; Deninger uses exactly that feature as the dynamical counterpart of an arithmetic scheme that is *flat* over Spec ℤ (for instance Spec ℤ itself), as opposed to a scheme over a finite field, whose counterpart is a fibration. See Part II.

#### 2.3 The Reeb foliation of the solid torus, and of *S*³

**[known]** [Ree], [CC], [CLN]. Start with the solid torus *D*² × *S*¹, thought of as a filled doughnut. There is a famous foliation of this solid torus, the *Reeb foliation* (Reeb, 1952), with the following shape:

- The boundary torus *S*¹ × *S*¹ is itself a single compact leaf.
- Every interior leaf is diffeomorphic to a plane ℝ². In a meridional disk (a slice *D*² × {θ}), the traces of the interior leaves are curves that pile up against the boundary circle, never quite reaching it. Travelling in the long direction of the solid torus, each such leaf winds and accumulates on the boundary torus.

Gluing two such Reeb solid tori along their common boundary torus — the standard genus-one Heegaard splitting of the three-sphere — produces the *Reeb foliation of S³*. It has a single compact leaf (the glued torus) and all other leaves are planes, accumulating on that torus.

The Reeb foliation is a basic example of a *codimension-one foliation of a three-manifold that is not taut*: the compact leaf is “one-sided” in the sense that every transverse loop meeting it can be pushed off. We mention tautness only to place the example in the standard landscape; we will not use it below.

### 3. Singular foliations, and zeros of a vector field

A *regular* foliation, as above, has a leaf through every point, all of the same dimension. Many naturally occurring “foliations” are allowed to have *singular points*, where the leaf dimension drops, or where a defining vector field or holomorphic form vanishes.

The most elementary source of singular points is a vector field.

#### 3.1 Isolated zeros in the plane: four linear types

**[known]** Standard local theory of planar vector fields (any graduate ODE text; the holomorphic case is in [CLN] and [Bru]). Let *Y* be a smooth vector field on a surface, and let *p* be an isolated zero: *Y*(*p*) = 0, and *Y* ≠ 0 nearby. In a neighbourhood of *p* the orbits of *Y* (except *p* itself) are the leaves of a singular foliation, with a singularity at *p*.

If the linearisation *DY<sub>p</sub>* is non-degenerate (no zero eigenvalue), the local picture is, up to topological conjugacy, one of four linear types. Write *λ*₁, *λ*₂ for the eigenvalues of *DY<sub>p</sub>*.

| Type | Eigenvalues | Local picture | Poincaré–Hopf index |
| --- | --- | --- | --- |
| **Saddle** | real, opposite signs | two incoming separatrices, two outgoing | −1 |
| **Node** | real, same sign | all orbits enter (or all leave) *p*, without spinning | +1 |
| **Focus** (spiral) | complex conjugate, nonzero real part | orbits spiral into (or out of) *p* | +1 |
| **Centre** | purely imaginary, nonzero | orbits are closed curves around *p* | +1 |

The four pictures are drawn on the site page. A *node* and a *focus* are distinguished by whether the eigenvalues are real or not; both have index +1. A *centre* is the conservative (purely rotational) case, also of index +1. Only the saddle has index −1.

#### 3.2 The Poincaré–Hopf index

**[known]** [PH]. The *index* of an isolated zero *p* of a vector field *Y* on an oriented surface is the winding number, around a small circle about *p*, of the normalised vector *Y*/|*Y*|. Equivalently, it is the degree of the Gauss map from that circle to the unit circle.

The **Poincaré–Hopf theorem**: if *Y* is a vector field with only isolated zeros on a compact oriented manifold *M*, then

  ∑<sub>*p* zero</sub> index(*Y*, *p*)  =  χ(*M*),

the Euler characteristic of *M*. On the two-sphere, χ(*S*²) = 2, so no vector field on *S*² can have a single saddle and nothing else; the indices must add to 2. On the torus, χ(*T*²) = 0, and the Kronecker field of §2.2 has *no* zeros at all, which is compatible.

This is the first place in the primer where a “zero point” of a foliation is a precise, computable object: a zero of a vector field, with an integer invariant, constrained by topology.

#### 3.3 Holomorphic foliations of ℂ², Seidenberg reduction, and the Camacho–Sad index

**[known]** [Sei], [CS], [Bru], [CLN]. In complex dimension two the natural object is a *holomorphic foliation*, given locally by a holomorphic vector field, or equivalently by a holomorphic 1-form *ω* = *A dx* + *B dy*, with leaves the integral curves of *ω* = 0. A *singularity* is a point at which *A* and *B* both vanish (or the vector field vanishes).

Two classical facts, stated here only at a high level.

**Seidenberg’s reduction of singularities** [Sei] (1968). Starting from an isolated singularity of a holomorphic foliation of a complex surface, a finite sequence of blow-ups produces a foliation whose only singularities are *reduced*: in local coordinates, the linearisation has at least one nonzero eigenvalue, and if both eigenvalues are nonzero their ratio is not a positive integer. (The precise definition of “reduced” is in [Sei] and in [Bru, Ch. 1]; the point for this primer is that singularities of holomorphic foliations can be simplified, by a canonical geometric process, to a short list of local models.) After reduction one has a well-defined collection of exceptional divisors and a finite list of reduced singular points on them.

**The Camacho–Sad index** [CS] (1982). If *S* is a *separatrix* through a singularity *p* — a germ of an invariant analytic curve through *p* — Camacho and Sad attach a complex number CS(ℱ, *S*, *p*), the *Camacho–Sad index* of the foliation along *S* at *p*. If *C* is a compact invariant curve (a compact union of such separatrices and regular leaves), the sum of the Camacho–Sad indices at the singular points of *C* equals the self-intersection *C* · *C*. In particular, the formula produces at least one separatrix through every isolated singularity of a holomorphic foliation of a surface — the original existence theorem of Camacho–Sad.

We will not compute these indices in this first note. They are recorded because they are the standard local invariants of a singular holomorphic foliation, and because “a foliation with a zero point” in the holomorphic category is a rich, completely rigorous subject, independent of any arithmetic dictionary.

---

## Part II. Deninger’s picture

The rest of the primer is about a *proposed* correspondence, due to Christopher Deninger, between arithmetic schemes (and their zeta functions) and flows on foliated spaces. The correspondence is laid out in the 1998 ICM address [D-ICM], the 2001 Jahresbericht survey [D2], and the 2008 survey [D-2008] (preprint arXiv:0709.2801, the source we read in full). The cohomological side of the picture goes back to the 1992 / 1994 motives paper [D1].

**[known], as a proposal.** Deninger is explicit, in the first paragraph of [D-2008], that the cohomology theory he wants for schemes over Spec ℤ “has not been constructed.” What *has* been constructed is a class of dynamical systems whose reduced leafwise cohomology has several of the formal properties that such a cohomology would need, and a handful of genuine examples (function fields, solenoids, a circle) in which pieces of the dictionary can be checked.

We present the dictionary as Deninger states it. We do not add correspondences of our own.

### 4. The formal dictionary

**[known]** [D-ICM], [D-2008, Dictionary 1 and Dictionary 4]. Let 𝒳 be a regular algebraic scheme of dimension *d* over Spec ℤ (for *d* = 0 this is a finite set of closed points; for *d* = 1, examples are Spec ℤ and Spec 𝔬<sub>*K*</sub>). Deninger proposes to compare 𝒳 with a triple (*X*, ℱ, *φ<sup>t</sup>*), where:

- *X* is a space of dimension 2*d* + 1 (in the most optimistic form, a manifold; more accurately, in light of the function-field examples, a *generalized solenoid*, which can still carry a foliation and a flow, [CC, 11.3], [D-2008, Remark 2]);
- ℱ is a foliation of *X* of *codimension one*;
- *φ<sup>t</sup>* is a flow (a smooth ℝ-action) which maps leaves of ℱ to leaves, and whose orbits are typically *transverse* to the leaves.

The proposed matching is:

| Arithmetic side | Dynamical side |
| --- | --- |
| scheme 𝒳 / Spec ℤ, of dimension *d* | triple (*X*, ℱ, *φ<sup>t</sup>*), dim *X* = 2*d* + 1, ℱ of codimension one, flow mapping leaves to leaves |
| a closed point *x* of 𝒳 | a closed orbit *γ* of the flow |
| the residue cardinality *Nx* | *Nγ* = exp *ℓ*(*γ*), where *ℓ*(*γ*) is the length of *γ* |
| the Hasse–Weil zeta *ζ*<sub>𝒳</sub>(*s*) = ∏<sub>*x*</sub> (1 − *Nx*<sup>−*s*</sup>)<sup>−1</sup> | a Ruelle zeta *ζ*<sub>*X*</sub>(*s*) = ∏<sub>*γ*</sub> (1 − *Nγ*<sup>−*s*</sup>)<sup>±1</sup> (when the product makes sense) |
| 𝒳 over a finite field 𝔽<sub>*p*</sub> | the foliation ℱ is a *fibration*, coming from a suspension of a discrete Frobenius action |
| the number of residue characteristics of 𝒳 | the rank of the period group *Λ* of (*X*, ℱ, *φ<sup>t</sup>*) |
| ℱ is a fibration | if and only if rank *Λ* = 1 |
| infinite places / 𝒳(ℂ) / *F*<sub>∞</sub> | fixed points of the flow (the leaf through a fixed point is flow-invariant) |
| the product formula ∏<sub>𝔭</sub> ‖*f*‖<sub>𝔭</sub> = 1 | Kopei’s product formula on meromorphic-on-leaves functions [Kop] |
| the explicit formula of analytic number theory | a transversal index theorem for the ℝ-action and the leafwise Laplacian |
| −log \|*d*<sub>*K*/ℚ</sub>\| | Connes’ Euler characteristic *χ*<sub>ℱ</sub>(*X*, *μ*) of a Riemann-surface lamination, [D2, §4] |

Two remarks, both Deninger’s.

First, when 𝒳 is a scheme over 𝔽<sub>*p*</sub>, there *is* a Frobenius, and one can suspend a discrete dynamical system to a flow. If *M* carries a diffeomorphism (or self-cover) *φ* playing the role of Frobenius, the suspension

  *X* = *M* ×<sub>*p*^ℤ</sub> ℝ₊<sup>∗</sup>,

with *p<sup>ν</sup>* · (*m*, *u*) = (*φ*<sup>−*ν*</sup>(*m*), *p<sup>ν</sup> u*) and *φ<sup>t</sup>*[*m*, *u*] = [*m*, *e<sup>t</sup> u*], is a space of dimension dim *M* + 1 with a flow, and a codimension-one foliation whose leaves are the images of *M*. Closed orbits *γ* of the flow are in bijection with finite *φ*-orbits 𝔬 on *M*, and *ℓ*(*γ*) = |𝔬| log *p*. This is [D-2008, §2], and it is the one case of the dictionary that can be written down without speculation. For an elliptic curve over a finite field, Deninger carries the construction out with *X* a *generalized solenoid*, not a manifold [D3].

Second, when 𝒳 is flat over Spec ℤ there is no Frobenius, and the proposed counterpart is a flow that is not a suspension of a discrete map — typically a foliation with a *dense leaf*, corresponding to the fact that Spec ℤ has more than one residue characteristic. The period group *Λ* ⊂ ℝ is the image of the cohomology class of the closed 1-form *ω<sub>φ</sub>* that vanishes on the leaves and equals 1 on the flow field; ℱ is a fibration if and only if rank *Λ* = 1 [D-2008, Construction 3].

### 5. Leafwise cohomology, and zeros as eigenvalues

**[known]** [D-2008, §3], [AK1]. On a foliated space one can differentiate only in the leaf direction. The resulting “de Rham complex along the leaves” has a *reduced leafwise cohomology*

  Ḧ<sup>•</sup>_ℱ(*X*)  :=  Ker *d*<sub>ℱ</sub>  /  (closure of Im *d*<sub>ℱ</sub>).

In general these are infinite-dimensional Fréchet spaces. Álvarez López and Kordyukov proved that, for a *Riemannian* foliation (one that admits a bundle-like metric), this reduced cohomology is the kernel of the leafwise Laplacian *Δ*<sub>ℱ</sub>: a Hodge theorem in the leafwise direction [AK1]. The result uses the Riemannian (bundle-like) hypothesis in an essential way. Deninger and Singhof gave a counterexample to smooth leafwise Hodge decomposition for general foliations [DS, the 2001 *Ann. Inst. Fourier* paper].

The flow *φ<sup>t</sup>*, because it maps leaves to leaves, acts on Ḧ<sup>•</sup>_ℱ(*X*). The *infinitesimal generator* of that action is an operator

  *θ*  =  lim<sub>*t*→0</sub> (1/*t*) (*φ<sup>t</sup>*<sup>∗</sup> − id)

on the leafwise cohomology. Deninger’s proposal is that the non-trivial zeros of the zeta function of 𝒳 should correspond to the *eigenvalues of θ* on the degree-one (or middle-degree) reduced leafwise cohomology.

On *L*² leafwise cohomology the same picture can be phrased in terms of a self-adjoint leafwise Laplacian, [D-2008, §3].

### 6. The dynamical Lefschetz / transversal trace formula, and the explicit formula

**[known]** [D-2008, Theorem 5], [AK2], [DS]. A simple form of the explicit formula for ζ, as a distributional identity on ℝ<sub>>0</sub>, is

  1 − ∑<sub>*ρ*</sub> *e*<sup>*tρ*</sup> + *e*<sup>*t*</sup>  =  ∑<sub>*p*</sub> log *p*  ∑<sub>*k*≥1</sub> *δ*<sub>*k* log *p*</sub>  +  (1 − *e*<sup>−2*t*</sup>)<sup>−1</sup>,

where *ρ* runs over the zeros of ζ in the critical strip, and the last term comes from the archimedean place. (This is [D-2008, (2)]; it is a standard form of the explicit formula, not a new identity.)

On the dynamical side, Álvarez López–Kordyukov and Deninger–Singhof prove a transversal index theorem. In the setting of a compact 3-manifold *X*, with a codimension-one foliation that has a dense leaf, and a flow acting *conformally* with factor *e*<sup>*α t*</sup> on a leafwise metric (a strong hypothesis), and with non-degenerate closed orbits, one has an equality of distributions on ℝ<sub>>0</sub>,

  1 − ∑<sub>*ρ*</sub> *e*<sup>*tρ*</sup> + *e*<sup>*α t*</sup>  =  ∑<sub>*γ*</sub> *ℓ*(*γ*)  ∑<sub>*k*≥1</sub> ± *δ*<sub>*k ℓ*(*γ*)</sub>,

where now *ρ* runs over the spectrum of *θ* on Ḧ¹_{ℱ,(2)}(*X*), and — under those hypotheses — every such *ρ* satisfies Re *ρ* = *α*/2 [D-2008, Theorem 5].

The two identities have the same shape: a spectral side built from eigenvalues, and a geometric side built from closed orbits (or primes). That is the comparison Deninger draws.

Two limitations, both stated in [D-2008]:

- On *manifolds*, the conformal hypothesis of Theorem 5 forces *α* = 0, so the eigenvalues are purely imaginary. To have *α* = 1, which is the value that would match Re *ρ* = 1/2, one must leave the manifold category and work with solenoids; Deninger constructs such examples from elliptic curves over finite fields [D3], with *α* = 1.
- The identity above has no archimedean term, because the flow was assumed to have no fixed points. Deninger records ongoing work of Álvarez López and Kordyukov aimed at allowing fixed points (and possibly replacing reduced *L*² leafwise cohomology by “adiabatic cohomology”).

A formal identity of Ruelle type is also proposed:

  *ζ<sub>R</sub>*(*s*)  =  ∏<sub>*i*</sub> det<sub>∞</sub>(*s* · id − *θ*  |  Ḧ<sup>*i*</sup>_ℱ(*X*))<sup>(−1)<sup>*i*+1</sup></sup>,

relating the Ruelle zeta built from closed-orbit lengths to a zeta-regularized determinant of *θ* on leafwise cohomology [D-2008, (8)]. This is proved, under the assumption that the regularized determinants exist, for a simple class of *isometric* flows, in [D4] (and it is checked by hand for the circle, below).

### 7. What the Riemann hypothesis would become, if the space existed

**[known, as a proposed translation; not a theorem].** In the dynamical identity of §6, all eigenvalues of *θ* on the relevant leafwise cohomology satisfy Re *ρ* = *α*/2, once the strong conformal / isometric hypotheses are granted. Matching the explicit formula for ζ would require *α* = 1, hence Re *ρ* = 1/2. In the 1998 ICM address [D-ICM], Deninger discusses this as a positivity or Hodge-type statement on the (still unconstructed) arithmetic cohomology: a Riemann–Hodge condition that would force the eigenvalues onto the critical line.

**[speculative].** That translation *presupposes* the existence of a foliated dynamical system for Spec ℤ (or for a number field) whose leafwise cohomology realises the cohomology hoped for in [D1]. Deninger calls this an “optimistic guess” [D-2008, §1]. The guess has not been confirmed, the space has not been constructed, and **this is not progress on the Riemann hypothesis.**

### 8. Connes’ picture is related, and is not the same

**[known]** [Con]; see also the dictionary entry in [D-2008] for Connes’ Euler characteristic. Alain Connes’ 1999 *Selecta* paper constructs a spectral realization of the zeros of ζ as an *absorption spectrum* of an operator on an adelic space, in the language of noncommutative geometry, and recasts the Riemann hypothesis as a form of Weil positivity. The explicit formula appears, again, as a trace formula.

The two programmes share a family of ideas — a dynamical or spectral reading of the explicit formula, a role for positivity — but they are not the same construction. Connes works with an adelic quotient and a noncommutative flow of weights; Deninger works with leafwise cohomology of a foliated space. This primer is about Deninger’s picture. We mention Connes only to keep the two from being conflated.

---

## Part III. What has been built, and what is open

### 9. What has actually been built

The following items are theorems, or fully written constructions, in the papers cited. They are **[known]**.

1. **The formal dictionary and the cohomological wish-list.** The shape of a cohomology theory for schemes over ℤ whose regularized determinants would be *L*-functions was set out in [D1]. The foliated-dynamical dictionary was set out in [D-ICM], [D2], [D-2008]. This is a precise proposal, not a construction of the arithmetic cohomology.

2. **Function fields, and elliptic curves over finite fields.** For 𝒳 / 𝔽<sub>*p*</sub>, the suspension *X* = *M* ×<sub>*p*^ℤ</sub> ℝ₊<sup>∗</sup> (or a generalized solenoid, when *φ* is a self-cover of degree ≥ 2) is a genuine dynamical system, and closed orbits correspond to closed points with *ℓ*(*γ*) = |𝔬| log *p*. For an *elliptic curve over a finite field*, Deninger carries this out with *X* a generalized solenoid, not a manifold [D3], [D-2008, §1 and Remark 2]. In these examples one can have conformal factor *α* = 1.

3. **Leafwise Hodge theory, for Riemannian foliations.** Álvarez López–Kordyukov [AK1]: reduced leafwise cohomology equals the kernel of the leafwise Laplacian, when the foliation is Riemannian. This is the analytic engine of the later trace formulas. It fails, in general, without the Riemannian hypothesis (Deninger–Singhof counterexample, *Ann. Inst. Fourier* 51, 2001).

4. **A transversal index theorem, under strong hypotheses.** [AK2], [DS], [D-2008, Theorem 5]: a distributional identity of explicit-formula type, with Re *ρ* = *α*/2. On manifolds the hypotheses force *α* = 0. Solenoid examples can have *α* = 1.

5. **Kopei’s product formula.** [Kop]: for suitable functions meromorphic along the leaves of a 3-dimensional foliated flow, with divisors supported on closed orbits, one has ∏_γ ‖*f*‖_γ = 1, the dynamical counterpart of the product formula for a number field. A compact-leaf contribution plays the role of the infinite places.

6. **A Lichtenbaum analogue, for a simple class of isometric flows.** [D4], [D-2008, Theorem 8]: if the flow acts isometrically and the zeta-regularized determinants of *θ* exist, the order of vanishing and the leading coefficient of the Ruelle zeta at *s* = 0 are given by a formula of Lichtenbaum type, identified with analytic / Reidemeister torsion via the Cheeger–Müller theorem. This is a theorem about those dynamical systems, not about arithmetic zeta functions.

7. **The circle, as a complete toy.** [D-2008, “The simplest example”]. Let *q* > 1 and let *X* be the circle ℝ / (log *q*) ℤ, foliated by *points*, with ℝ acting by translation. The only “closed orbit” is the circle itself, of length log *q*, and the Ruelle zeta is

  *ζ<sub>R</sub>*(*s*)  =  (1 − *q*<sup>−*s*</sup>)<sup>−1</sup>.

The operator *θ* = *d/dx* on Ḧ⁰_ℱ(*X*) = *C*<sup>∞</sup>(*X*) has eigenvalues 2*π i ν* / log *q*, *ν* ∈ ℤ, and Lerch’s formula gives det<sub>∞</sub>(*s* · id − *θ*) = 1 − *q*<sup>−*s*</sup>. The Lichtenbaum-style identities of [D4] can be checked by hand: ord<sub>*s*=0</sub> *ζ<sub>R</sub>* = −1, and the leading coefficient is 1 / log *q*. This is the one case in which the whole formal machine is elementary.

8. **Leichtnam’s Lefschetz formula for laminated spaces with a *p*-adic transversal.** [Lei]. A Lefschetz trace formula in a setting that mixes real leaves with a *p*-adic transverse space. Under Leichtnam’s hypotheses, the eigenvalues of *θ* on a degree-one group *H*¹_τ have real part 1/2. Leichtnam states that the existence of the corresponding system for ℚ is still unknown, except in cases such as elliptic curves over finite fields.

We found no paper in the 1998–2008 literature under a standard name “Kim–Deninger.” The verified constructions around laminated or *p*-adic-transversal spaces in this programme are Leichtnam’s and the solenoid constructions of [D3].

### 10. What is open

**[known], as open problems stated by the authors.**

- The conjectured arithmetic cohomology of [D1], for schemes over ℤ, has not been constructed [D-2008, §1].
- In particular, there is no known foliated dynamical system (*X*, ℱ, *φ<sup>t</sup>*) for Spec ℤ, or for the ring of integers of a number field, whose leafwise cohomology realises that cohomology. This is the “optimistic guess,” and it remains a guess.
- Even on the dynamical side, a transversal index theorem that allows fixed points of the flow (the proposed counterparts of infinite places) and produces the archimedean term of the explicit formula was, at the time of [D-2008], ongoing work, not a finished theorem in that paper.
- The identification *ζ<sub>R</sub>*(*s*) = ∏ det<sub>∞</sub>(*s* − *θ* | Ḧ<sup>*i*</sup>) is formal, or is proved under existence assumptions on the regularized determinants, for a restricted class of isometric flows. It is not known in the generality the dictionary suggests.
- Leichtnam’s *p*-adic-transversal formalism does not yet include a system for ℚ.

**[speculative].** Any claim that this picture “explains the zeros of ζ,” or that a positivity condition on a leafwise cohomology “is” the Riemann hypothesis, is speculative in the strict sense: the object one would apply the condition to has not been shown to exist. We will not make that claim.

---

## 11. What this note does not do

- It does not invent correspondences beyond Deninger’s published dictionary.
- It does not treat Connes’ adelic construction as a foliated-space construction, or vice versa.
- It does not claim any progress, evidence, or “approach that nearly proves” the Riemann hypothesis.
- It does not compute Camacho–Sad indices, Seidenberg reductions, or a numerical check of a trace formula. Those are natural next computational steps, recorded in `OVERNIGHT-20261114.md`.

## 12. A short reading path

1. For foliations: Candel–Conlon [CC], Chapters 1–2, and the Reeb and linear examples there; Camacho–Lins Neto [CLN] for singularities.
2. For holomorphic singularities: Brunella [Bru], Chapter 1 (Seidenberg, reduced singularities, Camacho–Sad), with the original papers [Sei] and [CS] for the theorems.
3. For Deninger’s picture, in this order: the short ICM address [D-ICM]; the 2008 survey [D-2008] (arXiv:0709.2801), which we followed for Part II; then [D3] for the solenoid / function-field construction, [D4] for the Lichtenbaum analogue, [Kop] and [Lei] for the two most complete side-theorems.
4. For the Hodge and trace side: [AK1], [AK2], [DS].

The companion page on the site, `/foliations`, has four illustrations: an animation of the Kronecker foliation, phase portraits of the four linear singular points with their indices, a diagram of the Reeb foliation, and a labelled drawing of Deninger’s dictionary.
