Inventiones mathematicae (2023) 234:129–169

https://doi.org/10.1007/s00222-023-01198-1

Pólya’s conjecture for Euclidean balls

Nikolay Filonov1,2 · Michael Levitin3 · Iosif Polterovich4 · David A. Sher5

Received: 17 December 2022 / Accepted: 8 May 2023 / Published online: 5 June 2023
© The Author(s) 2023

Abstract
The celebrated Pólya’s conjecture (1954) in spectral geometry states that the eigen-
value counting functions of the Dirichlet and Neumann Laplacian on a bounded Eu-
clidean domain can be estimated from above and below, respectively, by the leading
term of Weyl’s asymptotics. Pólya’s conjecture is known to be true for domains which
tile Euclidean space, and, in addition, for some special domains in higher dimensions.
In this paper, we prove Pólya’s conjecture for the disk, making it the first non-tiling
planar domain for which the conjecture is verified. We also confirm Pólya’s conjec-
ture for arbitrary planar sectors, and, in the Dirichlet case, for balls of any dimension.
Along the way, we develop the known links between the spectral problems in the disk
and certain lattice counting problems. A key novel ingredient is the observation, made
in recent work of the last named author, that the corresponding eigenvalue and lattice
counting functions are related not only asymptotically, but in fact satisfy certain uni-
form bounds. Our proofs are purely analytic, except for a rigorous computer-assisted
argument needed to cover the short interval of values of the spectral parameter in the
case of the Neumann problem in the disk.

Mathematics Subject Classification (2020) Primary 35P15 · Secondary 35P20 ·

33C10 · 11P21

Contents

1 Weyl’s law and Pólya’s conjecture . . . . . . . . . . . . . . . . . . . . . . . 130

2 Dirichlet and Neumann eigenvalues of the ball and lattice counting problems 135
3 Proof of Theorem 2.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
4 From the weighted lattice point count towards Pólya’s conjecture . . . . . . 144
5 Proof of Theorem 4.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
6 Proof of Theorem 4.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
7 Proof of Theorem 4.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
8 Closing the gap: the proof of Theorem 4.4 . . . . . . . . . . . . . . . . . . . 161

A Mathematica script used for a computer-assisted part of the paper is available for download at
https://michaellevitin.net/polya.html.
Extended author information available on the last page of the article
130 N. Filonov et al.

9 Proof of Theorem 1.12 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Authors and Affiliations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168

1 Weyl’s law and Pólya’s conjecture

Let  ⊂ Rd be a bounded domain. Consider the Dirichlet eigenvalue problem for the
Laplacian

 d
∂2
− := −
j =1
∂xj2

in :

−u = λu in ,
(1.1)
u=0 on ∂.

It is well known that the spectrum of (1.1) is discrete and consists of isolated eigen-
values of finite multiplicity accumulating to +∞,

0 < λ1 () ≤ λ2 () ≤ · · · ≤ λn () ≤ . . . ,

which we enumerate with account of multiplicities.

Similarly, assuming additionally that ∂ is Lipschitz, consider the Neumann
eigenvalue problem

−u = μu in ,
(1.2)
∂n u = 0 on ∂,

where ∂n u = ∇u, n|∂ denotes the normal derivative of u with respect to the exte-
rior unit normal n on the boundary. The spectrum of (1.2) again consists of isolated
eigenvalues of finite multiplicity accumulating to +∞,

0 = μ1 () ≤ μ2 () ≤ · · · ≤ μn () ≤ . . . ,

enumerated with account of multiplicities.

Let, for λ ∈ R,
   
ND (λ) := # n : λn () ≤ λ2 and NN (λ) := # n : μn () ≤ λ2
Pólya’s conjecture for Euclidean balls 131

denote the counting functions1 of the Dirichlet and Neumann eigenvalue problems on
.2 It follows from the variational principles for (1.1) and (1.2) that

ND (λ) ≤ NN (λ)

for any λ ≥ 0.3

Under the assumptions stated above, the leading term asymptotics of the counting
functions is given by Weyl’s law [41],

N (λ) = Cd ||d λd + R(λ), (1.3)

 d  either N (λ) or N (λ), | · |d denotes the d-dimensional

where N (λ) denotes D N

volume, R(λ) = o λ as λ → +∞, and

1
Cd := d  
(4π)  d2 + 1
2

is the so-called Weyl constant. We refer to [38] for a historical review, as well as
numerous generalisations and improvements.
H. Weyl himself conjectured [42] a sharper version of (1.3) taking into account
the boundary conditions: for  ⊂ Rd with a piecewise smooth boundary,
 
N (λ) = Cd ||d λd ± Cb,d |∂|d−1 λd−1 + o λd−1 as λ → +∞, (1.4)

where the minus sign is taken for the Dirichlet boundary conditions and the plus sign
for the Neumann ones, and
1
Cb,d := d−1
 .
d+1
2d+1 π 2  2

We note that for planar domains (1.4) takes the particularly simple form

Area() 2 Length(∂)
N (λ) = λ ± λ + o (λ) . (1.5)
4π 4π
The two-term Weyl’s law (1.4) remains open in full generality. It has been proved
by V. Ivrii [20] under the condition that the set of periodic billiard trajectories in 
has measure zero. While this condition is conjectured to be satisfied for all Euclidean
domains, it has been verified only for a few classes, such as convex analytic domains
and polygons, see [38] and references therein. Specifically for a disk, it was proved
by N. Kuznetsov and B. Fedosov in [24].

1 Strictly speaking, we are counting the number of eigenvalues less than or equal to a given λ2 , but such
normalisation will be convenient to us throughout.
2 One can also define the counting functions using strict inequalities; this does not affect any of the results
below.
3 This in fact can be improved to N D (λ) + 1 ≤ N N (λ), see [10] and [4].
 
132 N. Filonov et al.

Assuming that the two-term Weyl’s asymptotics (1.4) holds for a domain  ⊂ Rd ,
we immediately obtain that for λ above some sufficiently large but unspecified value
1 we have

ND (λ) ≤ Cd ||d λd ≤ NN (λ). (1.6)

We refer also to [34] for results of the same kind in the Riemannian setting.
In 1954, G. Pólya [36] conjectured that the inequalities (1.6) hold for all λ ≥ 0.4
He later proved this conjecture in [37] for tiling domains : that is, domains such
that Rd can be covered, up to a set of measure zero, by a disjoint union of copies of
. In fact, in the Neumann case, some additional assumptions were imposed in [37]
that have been removed in [21]. It has been also shown that Pólya’s conjecture in the
Dirichlet case holds for a Cartesian product  = 1 × 2 ⊂ Rd1 +d2 if it holds for
1 ⊂ Rd1 with d1 ≥ 2, and 2 ⊂ Rd2 is bounded, see [26, Theorem 2.8]. For general
domains, somewhat weakened versions of (1.6) are known to hold as a consequence
of the so-called Berezin–Li–Yau inequalities: we have
d/2
d d +2 N
ND (λ) ≤ Cd ||d λd ≤ N (λ)
d +2 2
for all λ ≥ 0, see [31], [23], and [26]. We refer also to [32], [25], [9], and [8] for some
recent results on Pólya’s conjecture and further interesting links to other problems in
spectral geometry.

Remark 1.1 Pólya’s conjecture (1.6) can be equivalently restated as the inequalities
for the eigenvalues (instead of the counting functions),
2 2
μn+1 () ≤ (Cd ||d )− d n d ≤ λn () (1.7)

for all n ≥ 1. It is known that inequalities (1.7) hold for any domain in any dimension
for n = 1, 2. In particular, for n = 1 this follows from the celebrated Faber–Krahn and
Szegő–Weinberger inequalities, and for n = 2 in the Dirichlet case from the Krahn–
Szego inequality, see [17]. For n = 2 in the Neumann case, we refer to [11], [1].
These are the only eigenvalues for which it is known in full generality. We refer also
to [7] for further results on the validity of the Dirichlet Pólya’s conjecture for low
eigenvalues in higher dimensions. 

Remarkably, since balls do not tile the space, Pólya’s conjecture has so far re-
mained open for Euclidean balls, including planar disks.5 Although all the eigenval-
ues of the Dirichlet and Neumann Laplacians on the unit disk are explicitly known in
terms of zeros of the Bessel functions or their derivatives, see §2 below, in each case
the spectrum is given by a two-parametric family, and rearranging it into a single
monotone sequence appears to be an unfeasible task.

4 In fact, Pólya’s original conjecture was only for planar domains, and in a slightly different form.
5 As stated in [27, p. 638]: “Remarkably this conjecture still remains open even for such a simple domain
as the disc, where the eigenvalues of the Dirichlet Laplacians could be calculated via the roots of Bessel
functions.” See also [6, p. 1366] and [28, p. 66].
Pólya’s conjecture for Euclidean balls 133

d/2
Let Bd ⊂ Rd be the d-dimensional unit ball. Then |Bd |d = π . Therefore the
2 +1
d

leading Weyl’s term in (1.3) for Bd becomes

1
Wd (λ) := Cd |Bd |d λd = wd λd , wd =  d 2 , (1.8)
2d  2 + 1

in particular

λ2 2λ3
W2 (λ) = and W3 (λ) = .
4 9π
The main results of this paper address the validity of Pólya’s conjecture for disks
and balls. Namely, we prove the following results.

Theorem 1.2 The Dirichlet Pólya’s conjecture for the unit ball holds in any dimen-
sion d ≥ 2, that is we have

NBDd (λ) < Wd (λ)

for all λ > 0.

Our results in the Neumann case are restricted to the case d = 2. Higher-
dimensional Neumann problems are harder, and we intend to treat them in a sub-
sequent paper.
We first state

Lemma 1.3 The Neumann Pólya’s conjecture for D = B2 is valid for all λ ∈ [0, 0 ],
where
√
0 := 2 3.

Proof Taking the span of {1, x, y} as a test space in the Rayleigh quotient for the
Neumann Laplacian on D gives μ3 (D) ≤ 4. Therefore,

1, λ ∈ [0, 2), λ2
NDN (λ) ≥ ≥ for λ ∈ [0, 0 ]. 
3, λ ≥ 2, 4

We then prove

Theorem 1.4 The Neumann Pólya’s conjecture for the unit disk holds for all

6π
λ ≥ 1 := . (1.9)
3π − 8

We note that 0 > 3 and 1 < 14, so we already have the validity of the Neumann
Pólya’s conjecture for the disk for all λ outside the interval (3, 14).
134 N. Filonov et al.

Theorem 1.5 The Neumann Pólya’s conjecture for the unit disk holds for all λ ∈
[3, 14].

The proof of Theorem 1.5 is rigorous but computer-assisted. More specifically, it

is based on a realisation of an algorithm which satisfies two fundamental principles.
Principle 1. The algorithm should complete in a finite number of steps.
Principle 2. The algorithm should operate only with integer or rational numbers,
thus avoiding any use of floating-point arithmetic and any rounding errors.
The combination of Lemma 1.3 and Theorems 1.4 and 1.5 ensures that the Neu-
mann Pólya conjecture for the disk is valid for all λ > 0, that is we have

λ2
Corollary 1.6 NDN (λ) > 4 for all λ > 0.

Remark 1.7 Since Pólya’s conjecture is scale-invariant, its validity for a unit ball im-
mediately implies that it is valid for any ball of the same dimension. 

We additionally have the following generalisation of Pólya’s result for tiling do-
mains: we show that Pólya’s conjecture holds not only for domains which tile Eu-
clidean space, but also for domains which tile another domain for which it is known
to be true.

Theorem 1.8 Let  ⊂ Rd be a domain for which either the Dirichlet or the Neu-
mann Pólya’s conjecture holds, and let  be a domain which tiles . Then the same
Pólya’s conjecture also holds for  .

Proof Assume that  can be tiled by ≥ 2 congruent copies of  , so that ||d =

| |d . We have, by bracketing and since the eigenvalues of all the congruent copies
coincide with those of  ,

ND (λ) ≤ ND (λ) < NN (λ) ≤ NN (λ).

Assuming now (1.6) for all λ ≥ 0, we get

ND (λ) ≤ Cd ||d λd = Cd | |d λd ≤ NN (λ),

and the result follows by cancelling . 

Remark 1.9 If the inequalities in Pólya’s conjecture (1.6) for  are strict, they are
also strict for  . 

Theorem 1.8 immediately implies the following

Corollary 1.10 Let  ⊂ Sd−1 be a spherical domain which tiles Sd−1 . Then the
Dirichlet Pólya’s conjecture holds for the spherical cone in Rd with the base  and
the vertex at the origin.
Pólya’s conjecture for Euclidean balls 135

In the planar case we get

Corollary 1.11 Pólya’s conjecture holds for any circular sector Sα with an aperture
α = 2π , where ∈ {2, 3, . . . }.

We refer also to [8] for an alternative proof of Corollary 1.11 for sufficiently large
(but unspecified) .
We can in fact extend the result of Corollary 1.11 to arbitrary sectors.

Theorem 1.12 Pólya’s conjecture holds for any circular sector Sα with an aperture
α ∈ (0, 2π], that is

αλ2
NSDα (λ) < < NSNα (λ)
8π
for all λ > 0.

Remark 1.13 The result of Theorem 1.12 in the case α = 2π (the disk with the radial
slit) follows immediately from Theorems 1.2 (d = 2) and 1.4 by Dirichlet–Neumann
bracketing. 

Plan of the paper

In the next section we describe two lattice counting problems (2.6) and (2.7), variants
of which were originally introduced by N. Kuznetsov and B. Fedosov in [24], and
which are closely linked to the Dirichlet and Neumann eigenvalue counting problems
in the ball. The key novel tool is Theorem 2.3, originally obtained in part in [39],
which gives a uniform bound between the eigenvalue and the lattice counting func-
tions, as opposed to asymptotic relations that were previously known. We provide an
independent proof of this result in §3. In §4 we state the results on the lattice counting
functions which are sufficient for proving Pólya’s conjecture for balls. The bulk of
the paper, §§5–8, is devoted to the proofs of these results. Theorem 1.12 is proved in
§9.

2 Dirichlet and Neumann eigenvalues of the ball and lattice counting

problems

Throughout this paper, with ν ≥ 0, and k ∈ N, let Jν (z) be the Bessel functions of
order ν, let jν,k be the kth positive zero of Jν , and let jν,k be the kth positive zero of
its derivative Jν , with the exception of J0 for which j0,1 = 0.
It is well known that the eigenvalues of the Dirichlet Laplacian in the unit ball
are given by the squares of the zeros of the cylindrical Bessel functions. Namely,
considering the Dirichlet Laplacian in Bd , we have the simple eigenvalues
 2
λd,0,k = jd/2−1,k , k ∈ N,
136 N. Filonov et al.

that is of multiplicity

κd,0 := 1,

and the eigenvalues

 2
λd,m,k = jm+d/2−1,k , m, k ∈ N,

of multiplicity
m+d −1 m+d −3
κd,m := − .
d −1 d −1

Remark 2.1 We note that the numbers κd,m , d ≥ 2, m ∈ N0 = N ∪ {0}, coincide with
the multiplicity of the eigenvalue m(m + d − 2) of the Laplace–Beltrami operator on
the unit sphere Sd−1 , or alternatively with the dimension of the space of homogeneous
harmonic polynomials of degree m in Rd . In the planar case we have

κ2,m = 2 for m ∈ N.

We therefore have
∞
 
NBDd (λ) = κd,m # k ∈ N : jm+d/2−1,k ≤ λ . (2.1)
m=0

Remark 2.2 The sum in (2.1) is in fact finite: we have6

λ−d/2+1
 
NBDd (λ) = κd,m # k ∈ N : jm+d/2−1,k ≤ λ , (2.2)
m=0

This is due to the fact that jν,1 > ν [35, Eq. 10.21.3]. Note that in (2.2) and further
on, any sum in which the lower limit exceeds the upper limit is assumed to be zero,
which immediately gives NBDd (λ) = 0 for λ < d2 − 1. 

In the planar case the expression (2.1) simplifies to


λ−d/2+1
  
NDD (λ) = # k ∈ N : j0,k ≤ λ + 2 # (m, k) ∈ N2 : jm,k ≤ λ .
m=1

Similarly, the eigenvalues of the Neumann Laplacian in the unit disk D are given
by the squares of the zeros of the derivatives of Bessel functions. We have the simple
eigenvalues
 2
μ0,k = j0,k , k ∈ N,

6 Here and further on, x = max{k ∈ Z : k ≤ x} denotes the integer part of x ∈ R, and x = min{k ∈ Z :
k ≥ x} denotes its ceiling.
Pólya’s conjecture for Euclidean balls 137

Fig. 1 The Dirichlet eigenvalue counting function NDD (λ) (blue), the Neumann eigenvalue counting func-
2
tion NDN (λ) (red), and the leading Weyl’s term Wd (λ) = λ4 (black) in dimension d = 2. The plot is pro-
duced using the floating-point evaluation of zeros of the Bessel functions and their derivatives. If we were
to assume (contrary to the philosophy of this paper) the validity of floating-point arithmetic, this plot would
have presented a numerically assisted (as opposed to computer-assisted) “proof” of Pólya’s conjecture for
the disk for λ  15

and the double eigenvalues

 2
μm,k = jm,k , m, k ∈ N.

We therefore have


λ−d/2+1
  
NDN (λ) = # k ∈ N : j0,k ≤ λ + 2 # (m, k) ∈ N2 : jm,k ≤ λ , (2.3)
m=1

where the sum is again finite since jν,1 ≥ ν [35, Eq. 10.21.3].
For illustrative purposes only, we show the graphs of the Dirichlet and Neumann
eigenvalue counting functions for the disk in Fig. 1.
We will be comparing the counting functions NBDd (λ) and NDN (λ) with some
weighted lattice counting functions. Let

1  
h(x) := 1 − x 2 − x arccos x , x ∈ [0, 1],
π
and let P be a planar region under the graph of h(x),

P := {(x, y) : x ∈ [0, 1], y ∈ [0, h(x)]} .

Let, for λ > 0,

z 1  2 z
Gλ (z) := λh = λ − z2 − z arccos , z ∈ [0, λ], (2.4)
λ π λ
138 N. Filonov et al.

Fig. 2 The region Pλ , and the sets of shifted lattice points QD N

d (λ) (blue disks) and Qd (λ) (red diamonds),
shown here for d = 2 and λ = 23

and let Pλ be a dilation of P with coefficient λ with respect to the origin,

Pλ = {(z, y) : 0 ≤ z ≤ λ, 0 ≤ y ≤ Gλ (z)} , (2.5)

that is, the region under the graph of Gλ (z).

Let
 
d 1
Qd (λ) := (m, k) +
D
− 1, − ∈ Pλ : (m, k) ∈ N0 × N
2 4

and
 
3
2 (λ) :=
QN (m, k) + 0, − ∈ Pλ : (m, k) ∈ N0 × N
4

be the sets of shifted integer lattice points which lie in Pλ , see Fig. 2. The defini-
tions of the two sets for d = 2 differ by a vertical shift. The reason for choosing this
particular notation will become evident later.
We now introduce the weighted lattice point counting functions

PdD (λ) := κd,m (2.6)
 
m+ d2 −1,k− 14 ∈QD
d (λ)

and

P2N (λ) := κ2,m . (2.7)
 
m,k− 34 ∈QN
2 (λ)
Pólya’s conjecture for Euclidean balls 139

It is immediately seen from the definitions (2.5)–(2.7) that with ℵ ∈ {D, N} we have
λ−d/2+1
  
d
Pdℵ (λ) = κd,m Gλ m + − 1 + s ℵ , (2.8)
2
m=0

where
1 3
s D := , s N := . (2.9)
4 4
It is well known that as λ → +∞, the asymptotics of the lattice point counting
function PdD (λ) is intricately linked to the asymptotics of the eigenvalue count-
ing function NBDd (λ). This was first shown in the planar case in [24] and later re-
discovered in [2], see also [12]. Namely, in some appropriate sense,

NBDd (λ) ∼ PdD (λ) as λ → +∞.

This observation, together with asymptotic bounds on the difference between the two
functions, has been used to great effect to estimate the remainder in Weyl’s law for
the unit ball. In particular, for the Dirichlet problem in the disk the two-term Weyl
asymptotics (1.5) holds with an improved remainder estimate
 
O λ131/208 (log λ)18627/8320 ,
 
see [15] (the remainder estimate O λ2/3 was already obtained in [24], [2]). Sim-
ilar improved remainder estimates are also known in the Dirichlet case for higher-
dimensional balls [13] and in the planar Neumann case [14].
As has been recently found in [39] in the Dirichlet case, there is a further simple
non-asymptotic relation between the lattice point and the eigenvalue counting func-
tions, which lies at the cornerstone of our proofs of Theorems 1.2 and 1.4.

Theorem 2.3 For any d ≥ 2 and any λ ≥ 0, we have

NBDd (λ) ≤ PdD (λ).

We also have, for any λ ≥ 0,

P2N (λ) ≤ NDN (λ).

3 Proof of Theorem 2.3

We start by introducing some additional notation. Set, for ν ≥ 0, λ ≥ 0, and ℵ ∈

{D, N},

Gλ (ν) + s ℵ , if λ ≥ ν,
Aℵν (λ) := (3.1)
sℵ, if 0 ≤ λ < ν,
140 N. Filonov et al.

where Gλ is defined by (2.4) and s ℵ is defined by (2.9). Some typical graphs of the
functions Aℵν (λ) are shown in Fig. 3.
The crucial step in the proof of Theorem 2.3 comes from the following bounds on
the number of zeros of Bessel functions and their derivatives below a given number.

Proposition 3.1 Let ν ≥ 0 and λ ≥ 0. Then

  
# k ∈ N : jν,k ≤ λ ≤ AD
ν (λ) (3.2)

and
  
# k ∈ N : jν,k ≤ λ ≥ AN
ν (λ) . (3.3)

Remark  For λ ∈ [0, ν], the inequalities (3.2) and (3.3) become the trivial identities
 3.2
0 = s ℵ = 0. 
 
Remark 3.3 For ν = 0, the inequality (3.2) is equivalent to j0,k ≥ π k − 14 , which
was proved in [18]. 

Proof of Proposition 3.1 We recall the representations of the Bessel functions of the
first and second kind, Jν and Yν , and their derivatives in terms of the so-called mod-
ulus functions Mν and Nν and the phase functions θν and φν ,

Jν (x) = Mν (x) cos θν (x), Yν (x) = Mν (x) sin θν (x),

Jν (x) = Nν (x) cos φν (x), Yν (x) = Nν (x) sin φν (x)

(see [35, Eqs. 10.18.4–5]). We will be using various properties of the phase functions
below; for a review of these properties see [19]. We will be only considering the cases
ν ≥ 0 and x ≥ 0 for which the moduli Mν (x) and Nν (x) are both positive.
Let us concentrate first on (3.2). We have Jν (x0 ) = 0 if and only if cos θν (x0 ) = 0,
and so if and only if
1 1
θν (x0 ) + ∈ Z. (3.4)
π 2
Note that the phase function θν (x) satisfies θν (x) → − π2 as x → +0 [35, Eq.
10.18.3], and that it is monotone increasing for x ∈ (0, +∞) [19, Theorem 1], there-
fore (3.4) can be replaced by

BνD (x0 ) ∈ N,

where
1 1
BνD (x) := θν (x) + , (3.5)
π 2
and therefore
  
# k ∈ N : jν,k ≤ λ = BνD (λ) . (3.6)
Pólya’s conjecture for Euclidean balls 141

We have
λ ν 1
Gλ (ν) = − +O as λ → +∞. (3.7)
π 2 λ

Using the asymptotics7 [19, Eq. (21)], [35, Eq. 10.18.18],

π 1 1
θν (λ) = λ − ν+ +O as λ → +∞,
2 2 λ
and (3.5), (3.7), we obtain

BνD (λ) − AD
ν (λ) → 0 as λ → +∞.

Further,
√
dBνD (λ) 1 λ2 − ν 2 dAD
ν (λ)
= θν (λ) > =
dλ π πλ dλ
for λ ≥ ν by [19, Eq. (56)]. As we additionally have

dBνD (λ) dAD

ν (λ)
>0=
dλ dλ
for λ ∈ (0, ν], the function BνD (λ) − AD
ν (λ) is monotone increasing on (0, +∞) and
tends to zero at infinity. Thus, we have proved that

BνD (λ) < AD

ν (λ) for λ ∈ [0, +∞). (3.8)

Combining (3.6) and (3.8) proves (3.2).

We now prove (3.3). In the same manner we have Jν (x0 ) = 0 if and only if
1 1
φν (x0 ) + ∈ Z. (3.9)
π 2
We note that the phase function φν satisfies φν (x) → π2 as x → +0 [35, Eq. 10.18.3].
Also, φν (x) is monotone increasing for x ∈ (ν, +∞) and monotone decreasing for
x ∈ (0, ν) [19, Theorem 1], with φν (ν) > − π2 [19, formula (60)]. Thus, the condition
(3.9) can be replaced by

BνN (x0 ) ∈ N,

where
1 1
BνN (x) := φν (x) + , (3.10)
π 2
and therefore
  
# k ∈ N : jν,k ≤ λ = BνN (λ) . (3.11)

7 See also [16] for all the coefficients of the full asymptotic expansion and some useful remarks.
142 N. Filonov et al.

Fig. 3 An illustration of inequalities (3.8) and (3.12). The plots of BνD (λ) and BνN (λ) are drawn using the
recipe from [19]. We remark that BνN (λ) has a minimum at λ = ν

Using the asymptotics [19, Eq. (22)], [35, Eq. 10.18.21],

π 1 1
φν (λ) = λ − ν− +O as λ → +∞,
2 2 λ

and (3.10), (3.7), we get

BνN (λ) − AN
ν (λ) → 0 as λ → +∞.

Also,
√
dBνN (λ) 1 λ2 − ν 2 dAN
ν (λ)
= φν (λ) < =
dλ π πλ dλ
for λ ≥ ν by [19, formula following Eq. (58)]. As we additionally have

dBνN (λ) dAN

ν (λ)
<0=
dλ dλ

for λ ∈ (0, ν], the function BνN (λ) − AN

ν (λ) is monotone decreasing on (0, +∞) and
tends to zero at infinity. Thus, we have proved that

BνN (λ) > AN

ν (λ) for λ ∈ [0, +∞). (3.12)

Combining (3.11) and (3.12) proves (3.3). 

We illustrate inequalities (3.8) and (3.12) in Fig. 3.

The bound (3.2) can be also proved without relying on the properties of the Bessel
phase function. Instead, one uses the known asymptotics of the Bessel zeros and
the Sturm comparison theorem. We present this alternative argument below for an
Pólya’s conjecture for Euclidean balls 143

interested reader. The bound (3.3) can be proved in the same manner; we omit the
details.
Let ν ≥ 0, and let us consider the function
 
sin πAD ν (x) .

Let aν,k be its kth zero in [ν, +∞), ordered increasingly. Obviously,
  
# k ∈ N : aν,k ≤ λ = AD ν (λ) . (3.13)

We will prove

Lemma 3.4 aν,k ≤ jν,k for all k ∈ N.

Together with (3.13), Lemma 3.4 immediately implies (3.2).

√
Proof of Lemma 3.4 The function Uν (x) := xJν (x) satisfies the differential equa-
tion
ν 2 − 1/4
Uν (x) + 1 − Uν (x) = 0. (3.14)
x2
Consider the function
√  
x
Vν (x) := 2 sin πAD
ν (x) + b
(x − ν 2 )1/4
 
with some b ∈ 0, π4 . Then

   1/4  
ν2 x2 − ν2
Vν (x) = − √  5/4 sin ν (x) + b
πAD + √ cos πAD
ν (x) + b ,
2 x x2 − ν2 x

6ν 2 x 2 − ν 4 (x 2 − ν 2 )3/4  
Vν (x) = − sin πA D
ν (x) + b .
4x 3/2 (x 2 − ν 2 )9/4 x 3/2
Therefore,

ν2 ν 2 (6x 2 − ν 2 )
Vν (x) + 1 − − 2 2 Vν (x) = 0 for x ∈ (ν, +∞). (3.15)
x 2 4x (x − ν 2 )2
Denote by vν,k = vν,k (b) the kth zero of the function Vν . By the definitions of Vν and
AD
ν,

2 − ν 2 − ν arccos ν + π + b = πk.
vν,k
vν,k 4
As
   ν π  
2 − ν2 = v −1 −1
vν,k ν,k + O vν,k and arccos = + O vν,k as k → ∞,
vν,k 2
144 N. Filonov et al.

we have
ν 1  
vν,k (b) = π k + − − b + O k −1 as k → ∞.
2 4
On the other hand the asymptotics
ν 1  
jν,k = π k + − + O k −1 as k → ∞,
2 4
is well known, see for example [35, Eq. 10.21.19].
Suppose that b > 0. Then there exists K ∈ N such that

jν,k > vν,k (b) for k ≥ K. (3.16)

The coefficient in front of Uν in (3.14) is greater than the coefficient in front of Vν

in (3.15):

ν 2 − 1/4 ν2 ν2 ν 2 (6x 2 − ν 2 )
1− >1− 2 >1− 2 − 2 2 for all x ∈ [ν, +∞).
x 2 x x 4x (x − ν 2 )2
By the Sturm comparison theorem there is a zero of Uν between vν,k (b) and
vν,k+1 (b). So, if jν,k0 ≤ vν,k0 (b) for some number k0 , then jν,k0 +1 ≤ vν,k0 +1 (b), and
by induction jν,k ≤ vν,k (b) for all k ≥ k0 which contradicts (3.16). Therefore, (3.16)
holds for all natural k.
Finally, each function vν,k (b) is continuous in b. Thus,

jν,k ≥ vν,k (0) = aν,k . 

Returning now to the proof of Theorem 2.3, we rewrite (2.8) as

λ−d/2+1
  
Pdℵ (λ) = κd,m Aℵm+d/2−1 (λ) . (3.17)
m=0

Theorem 2.3 now immediately follows from Proposition 3.1 with account of (3.17),
(2.2), and (2.3).

4 From the weighted lattice point count towards Pólya’s conjecture

By Theorem 2.3, the Dirichlet Pólya’s conjecture for Bd would follow immediately
if we can prove that

PdD (λ) < Wd (λ) (4.1)

for all λ ∈ (0, +∞), where Wd (λ) is defined by (1.8).

Similarly, the Neumann Pólya’s conjecture for D would follow immediately if we
can prove that

P2N (λ) > W2 (λ) (4.2)

Pólya’s conjecture for Euclidean balls 145

Fig. 4 A numerical experiment: the computed P2N (λ)/W2 (λ) − 1 as a function of λ

for all λ ∈ (0 , +∞); we note that we have already dealt with λ ≤ 0 by Lemma
1.3.
We establish (4.1) in the following cases, which will be dealt with separately.

Theorem 4.1 The inequality

λ2
P2D (λ) < W2 (λ) = (4.3)
4
holds for all λ > 0.

Theorem 4.1 will be proved in §5. Together with Theorem 2.3, it implies Theorem
1.2 in the planar case.

Theorem 4.2 The inequalities (4.1) hold for all d ≥ 3 and λ > 0.

Theorem 4.2 will be proved in §7. Together with Theorem 2.3, it implies Theorem
1.2 for higher-dimensional balls.
In the Neumann case, the situation is more delicate, as we cannot expect (4.2) to
hold for all values of λ ∈ (0, +∞) since P2N (λ) is identically zero for λ < π4 , see
Fig. 4.
We prove the following results.

Theorem 4.3 The inequality

λ2
P2N (λ) > W2 (λ) =
4
holds for all λ ≥ 1 , where 1 is given by (1.9).

Theorem 4.3 will be proved in §6. Together with Theorem 2.3, it implies Theorem
1.4.
We are further able to eliminate the remaining gap in the Neumann case.
146 N. Filonov et al.

Theorem 4.4 The inequality (4.2) holds for any λ ∈ (0 , 1 ).

The proof of this result, presented in §8, is computer-assisted. Theorem 4.4 implies
Theorem 1.5.

In all cases, we deal with estimating a (weighted) count of (shifted) lattice points
under the graph of a particular function Gλ . Such problems have been extensively
studied in number theory, going back to the Gauss circle problem. Important contri-
butions in the general case can be traced through the works of van der Corput [40] and
Krätzel [22] to some very recent results of Laugesen and Liu [29, 33]. In particular,
[30, Proposition 15] is directly applicable (with account of the fact that Laugesen and
Liu do not count the points on the vertical axis and do not double-count the points
inside) to our shifted lattice point count P2D (λ), yielding the bound
 √ 
λ2 2 1 3 λ2
P2 (λ) ≤
D
+ + − λ≈ + 0.7093λ.
4 3 π 2π 4

Unfortunately, since the coefficient in front of λ in this formula is positive, this bound
is weaker than our required bound (4.3). We need therefore to obtain sharper lattice
point count bounds than those available generally, and to do so we additionally use
some properties of the derivative of the function Gλ in addition to the properties of
the function itself, see Theorems 5.1 and 6.1, and also Remarks 5.2 and 6.3 for an
informal explanation.
For future use, we summarise below some elementary properties of the function
Gλ .
The first lemma is checked by a direct calculation.
 
Lemma 4.5 The function Gλ : [0, λ] → 0, πλ defined by (2.4) is a strictly monotone
decreasing convex C 1 function with
λ
Gλ (0) = , Gλ (λ) = 0,
π
1 z 1
Gλ (z) = − arccos , Gλ (0) = − , Gλ (λ) = 0,
π λ 2
1
Gλ (z) = √ .
π λ2 − z 2
 λ
We can therefore define the inverse function G−1λ : 0, π → [0, λ] which is also
monotone decreasing and convex. Sometimes, it will be also convenient for us to
consider Gλ on the interval [0, λ] by extending it by zero to (λ, λ]: the resulting
function, which we for simplicity denote by the same symbol, remains monotone
decreasing, convex, and C 1 .

Lemma 4.6 Let β ≥ 0. Then

 
 λ  β+1
λβ+2
2
z Gλ (z) dz = √
β  .
0 4 π (β + 2)  β+42
Pólya’s conjecture for Euclidean balls 147

In particular,
 λ λ2
Gλ (z) dz = . (4.4)
0 8

Proof In fact, the identity can be checked using computer algebra software, but we
include a proof for the sake of completeness. After a change of variables z = λ cos τ ,
we obtain
 λ 
λβ+2 π/2
z Gλ (z) dz =
β
(cos τ )β (sin τ − τ cos τ ) sin τ dτ. (4.5)
0 π 0

By [35, Eqs. 5.12.1–2],

   
 π/2  ρ+1
 σ +1
2 2
(cos τ )ρ (sin τ )σ dτ =  
ρ+σ +2
0 2 2

for any ρ, σ ≥ 0. Therefore,

 √
 π/2 π β+1
2
(cos τ )β (sin τ )2 dτ =   ,
β+4
0 4 2

 π/2  π/2
π/2 τ (cos τ )β+2  1
τ (cos τ ) β+1
sin τ dτ = − + (cos τ )β+2 dτ
0 β + 2 0 β +2 0
√  
π  β+3 2
=  ,
2 (β + 2)  β+4 2

and so
√  
 π/2 π β+1
2
(cos τ )β (sin τ − τ cos τ ) sin τ dτ =  .
β+4
0 4(β + 2) 2

Substituting this into (4.5) we get the result. 

Corollary 4.7 Let d ∈ N, d ≥ 2. Then

 λ
2
zd−2 Gλ (z) dz = Wd (λ) = wd λd .
(d − 2)! 0

Proof Applying Lemma 4.6 with β = d − 2 we get

 
 λ  d−1
λd
2 2
z d−2
Gλ (z) dz = √  .
(d − 2)! 0 2d π (d − 2)!  d+2
2
148 N. Filonov et al.

The duplication formula [35, Eq. 5.5.5]

√ √ d −1 d
π(d − 2)! = π(d − 1) = 2d−2  
2 2
implies
 
d−1
 λd 2 λd λd
√   =     =   2 = Wd (λ). 
2d π(d − 2)!  d+2
2 2 d−1 d  d  d+2
2 2 2 d  d+2
2

An important role inour study in the Neumann case will be played by the inverse
function value G−1
λ 4 (defined for all λ ≥ 4 ). We will use the following bounds.
1 π

Lemma 4.8 We have

1
<λ−1 G−1
λ (4.6)
4
 
for all λ ≥ 2. Additionally, for any σ ∈ 0, π2 we have

1
G−1
λ ≥ λ cos σ (4.7)
4
whenever
π
λ ≥ r1 (σ ) := . (4.8)
4(sin σ − σ cos σ )

Proof Since Gλ is monotone decreasing, the claim (4.6) is equivalent to

1
Gλ (λ − 1) < . (4.9)
4
√
We have G2 (1) − 1
4 = π
3
− 7
12 < 0, and additionally

d 1 1√ 1
(Gλ (λ − 1)) = 2λ − 1 − arccos 1 − <0
dλ π λ λ
as
1√ 2λ − 1 1
cos 2λ − 1 > 1 − >1− ,
λ 2λ2 λ
thus implying (4.9) for λ ≥ 2.
Similarly, the claim (4.7) is equivalent to
λ(sin σ − σ cos σ ) 1
Gλ (λ cos σ ) = ≥ ,
π 4
given (4.8). 
Pólya’s conjecture for Euclidean balls 149

5 Proof of Theorem 4.1

We first state the following

Theorem 5.1 Let b > 0, and let g be a non-negative decreasing convex function on
[0, b] such that g(b) = 0 and
1
|g(z) − g(w)| ≤ |z − w| (5.1)
2
for all z, w ∈ [0, b]. Then
  b 
   b
1 1
g(0) + +2 g(m) + ≤2 g(z) dz. (5.2)
4 4 0
m=1

The equality is possible only if g is identically zero on [0, b].

Remark 5.2 We explain here, very informally, the ideas behind the proof of Theorem
5.1. The area under the graph of the function g on the interval [m, m + 1] is approxi-
mately equal to the area under the straight line passing through the points (m, g(m))
and (m + 1, g(m + 1)), so
 m+1
1
g(z)dz ≈ (g(m) + g(m + 1)) .
m 2
Summing up these equalities over m we obtain
 b b

2 g(z)dz ≈ g(0) + 2 g(m).
0 m=1
 
If a number x is chosen randomly then x ≈ x − 12 on average. Thus, g(m) + 14 ≈
g(m) − 14 , and these extra contributions of − 14 ensure the sign of the inequality in
(5.2). In order to prove Theorem 5.1 rigorously we divide the graph by horizontal
lines y = n, with n = 0, 1, . . . , g(0), and we consider what happens in the intervals
where n + 1 ≥ g(z) ≥ n. The values of g(m) + 1/4 there are either n or n + 1.
The point m is “bad” if g(m) ≥ n + 34 and thus g(m) + 1/4 = n + 1: these “bad”
points contribute more to the sum than we expect “on average”. The convexity of the
function g and condition (5.1) ensure that the number of such “bad” points is not
greater than half of the total number of integer points in an interval, and this yields
the required estimate in the interval we are considering. 

In order to prove Theorem 5.1 we require the following

Lemma 5.3 Let i, j ∈ Z, i < j . Let g be a decreasing convex function on [i, j + 1]

satisfying (5.1) for all z, w ∈ [i, j + 1]. Assume additionally that

n + 1 ≥ g(i + 1) ≥ · · · ≥ g(j ) ≥ n ≥ g(j + 1) (5.3)

150 N. Filonov et al.

for some n ∈ Z. Then

  j −1 
     j
1 1 1 1 1
g(i) + + g(m) + + g(j ) + ≤ g(z) dz. (5.4)
2 4 4 2 4 i
m=i+1

Proof of Lemma 5.3 The validity of the claim does not change if we add a constant
integer number to the function g. So, without loss of generality we can assume n = 0,
so that (5.3) becomes

1 ≥ g(i + 1) ≥ · · · ≥ g(j ) ≥ 0 ≥ g(j + 1). (5.5)

Additionally, (5.1) implies that g(i) ≤ g(i + 1) + 12 ≤ 32 .

Set
 
3
K = # m ∈ {i, . . . , j } : ≤ g(m) ,
4

and consider four cases.

Case K = 0. The left-hand side of (5.4) is zero, and the right-hand side is non-
negative by (5.5).
Case K = 1. Here
5 3
> g(i) ≥ > g(i + 1) ≥ · · · ≥ g(j ) ≥ 0,
4 4
 
and the left-hand side of (5.4) is equal to 12 . The assumption (5.1) yields g i + 12 ≥
1
2, therefore by non-negativity and convexity of g on [i, j ],
 j  i+1 1 1
g(z) dz ≥ g(z) dz ≥ g i + ≥ .
i i 2 2

Case K = 2. Here
3 3
≥ g(i) ≥ g(i + 1) ≥ > g(i + 2) ≥ · · · ≥ g(j ) ≥ 0 ≥ g(j + 1),
2 4

and the left-hand side of (5.4) equals 32 . By (5.1) we have j ≥ i + 2, and therefore
by non-negativity and convexity of g,
 j  i+2 3
g(z) dz ≥ g(z) dz ≥ 2g(i + 1) ≥ .
i i 2
Case K ≥ 3. Here
3 3
≥ g(i) ≥ g(i + 1) ≥ · · · ≥ g(i + K − 1) ≥ > g(i + K) ≥ · · · ≥ g(j )
2 4
≥ 0 ≥ g(j + 1).
Pólya’s conjecture for Euclidean balls 151

The left-hand side of (5.4) is equal to K − 12 . By convexity of g,

(K − 1)g(i + 1) + (K − 2)g(i + 2K − 2) ≥ (2K − 3)g(i + K − 1),

and therefore
2K − 5
g(i + 2K − 2) ≥ > 0.
4(K − 2)

Thus, j ≥ i + 2K − 2. Next,
 
j i+2K−2 3K − 3 1
g(z) dz ≥ g(z) dz ≥ (2K − 2)g(i + K − 1) ≥ ≥K −
i i 2 2

as K ≥ 3. 

Remark 5.4 One can easily see from the proof that the equality in (5.4) is attained in
the following three cases only:
• g(z) ≡ n on [i, j ] (if K = 0);
• j = i + 1 and g(z) = n + i−z 2 +
3
on [i, i + 1] (if K = 1);
4 
• j = i + 2 and g(z) = n + s(i + 1 − z) + 34 on [i, i + 2] with s ∈ 3 1
8, 2 (if K = 2).


We can now proceed to the proof of Theorem 5.1 proper.

Proof of Theorem 5.1 Let

N = g(0) .

If N = 0, then applying Lemma 5.3 with i = 0, j = b, and g(z) extended by zero
for z ∈ (b, b + 1], gives
  b 
   b
1 1
g(0) + +2 g(m) + ≤2 g(z) dz,
4 4 0
m=1

and therefore (5.2).

Assume now N ≥ 1. For k = 0, 1, . . . , N , let

Lk := max {m ∈ {0, . . . , b} : g(m) ≥ k} , (5.6)

see Fig. 5. Therefore, we have

0 ≤ LN < LN −1 < · · · < L0 = b ,

where the strict inequalities Lk < Lk−1 , k = 1, . . . , N , follow from (5.1).

We will consider two cases depending on whether LN = 0.
152 N. Filonov et al.

Fig. 5 The numbers Lk , see (5.6) and also Remark 5.5

Case LN = 0. We write

  b 
 
1 1
g(0) + +2 g(m) +
4 4
m=1
⎛ ⎞

N −1   k −1
L    
=2 ⎝ 1 g(Lk+1 ) + 1 + g(m) +
1
+
1
g(Lk ) +
1 ⎠
.
2 4 4 2 4
k=0 m=Lk+1 +1
(5.7)

For each k ∈ {0, . . . , N − 1} we have

k + 1 > g (Lk+1 + 1) · · · ≥ g (Lk ) ≥ k ≥ g (Lk + 1) ,

and applying Lemma 5.3 with i = Lk+1 , j = Lk , and n = k yields

  k −1
L      Lk
1 1 1 1 1
g(Lk+1 ) + + g(m) + + g(Lk ) + ≤ g(z) dz.
2 4 4 2 4 Lk+1
m=Lk+1 +1
(5.8)
Substituting (5.8) into (5.7) gives

b 
   b  b
1
g(0) + 2 g(m) + ≤2 g(z) dz ≤ 2 g(z) dz, (5.9)
4 0 0
m=1

as required, where in the last inequality we used non-negativity of g.

Pólya’s conjecture for Euclidean balls 153

Case LN > 0. We write

  b 
 
1 1
g(0) + +2 g(m) +
4 4
m=1
⎛ ⎞
 N −1 
 L   
1 1 1 1 1
= 2 ⎝ g(0) + + g(m) + + g(LN ) + ⎠
2 4 4 2 4
m=1
⎛ ⎞

N −1   k −1
L    
+2 ⎝ 1 g(Lk+1 ) + 1 + g(m) +
1
+
1
g(Lk ) +
1 ⎠
.
2 4 4 2 4
k=0 m=Lk+1 +1
(5.10)

We have

N + 1 > g(0) ≥ · · · ≥ g (LN − 1) ≥ g (LN ) ≥ N > g (LN + 1) ,

therefore applying Lemma 5.3 with i = 0, j = LN , and n = N , we get

⎛ ⎞
  L
N −1      LN
⎝ 1 1 1 1 1 ⎠
2 g(0) + + g(m) + + g(LN ) + ≤2 g(z) dz.
2 4 4 2 4 0
m=1
(5.11)
Substituting (5.11) and (5.8) into (5.10) gives (5.9).
Finally, assume that we have the equality in (5.2). Due to Remark 5.4, the func-
tion g is linear on each interval [Lk+1 , Lk ], and either dist(g(Lk ), Z) ≥ 14 or g ≡ 0
on [Lk+1 , Lk ]. The situation when dist(g(Lk ), Z) ≥ 14 and g ≡ 0 on [Lk , Lk−1 ] is
impossible due to continuity of g. If dist(g(Lk ), Z) ≥ 14 for all k, then in particular
g(L0 ) = g(B) ≥ 14 , and the last inequality in (5.9) is strict. Therefore, the equality
in (5.2) requires g ≡ 0 on the whole interval [0, b]. 

Remark 5.5 If g is strictly monotone on [0, b], then the inverse function g −1 is well-
defined on [0, g(0)], and the definitions (5.6) may be equivalently rewritten as
 
Lk = g −1 (k) , k = 0, . . . , g(0) .

We finally use Theorem 5.1 to prove Theorem 4.1.

Proof of Theorem 4.1 We apply (5.2) with b = λ and g = Gλ (which we can do since
Lemma 4.5 ensures that (5.1) holds in this case), and use (4.4), giving the bound
(4.3) and therefore confirming the validity of the Dirichlet Pólya’s conjecture for the
disk. 
154 N. Filonov et al.

6 Proof of Theorem 4.3

We start by stating

Theorem 6.1 Let b > 0, and let g be a non-negative decreasing convex function on
[0, b] such that g(0) ≥ 14 , g(b) = 0, and (5.1) holds for all z, w ∈ [0, b]. Let
 
1
M0 = Mg,0 := 1 + max m ∈ {0, . . . , b} : g(m) ≥ ,
4
and assume that M0 ≤ b. Then
b 
   b
3 b − 3M0
g(m) + ≥ g(z) dz − . (6.1)
4 0 8
m=0

Remark 6.2 If g is strictly monotone on [0, b], then the inverse function g −1 is well-
defined on [0, g(0)], and
 
−1 1
M0 = g + 1, (6.2)
4
cf. Remark 5.5. 

Remark 6.3 Once more, we first outline a very informal  plan of proving Theorem
6.1. As we have argued in Remark 5.2, we have g(m) + 34 ≈ g(m) + 14 , which
should in principle ensure the correct inequality sign in (6.1). “Bad” points are now
the points with n ≤ g(m) ≤ n + 14 . So, we divide the graph of g by the horizontal
 
lines at y = n + 14 , where n = 0, 1, . . . , g(0) + 34 . Again, this guarantees that the
number of “bad” points in each resulting interval is less than half the total number of
points there. This still leaves an unresolved issue of points m lying under the tail of
the graph of g, where 0 ≤ g(m) < 14 . Such points make no contribution to the left-
hand side of (6.1), but the tail does contribute to the integral: consider, for example,
a toy case of a function g(z) = 10−z80 on the interval [0, 10]. To account for that, we
subtract an additional term in the right-hand side of (6.1). 

Before proceeding to the proof of Theorem 6.1, we require

Lemma 6.4 Let i, j ∈ Z, i < j . Let g be a decreasing convex function on [i, j ]. As-
sume that
1 3
n+ > g(i) ≥ · · · ≥ g(j − 1) ≥ n − > g(j )
4 4
for some n ∈ Z. Then
  j −1 
     j
1 3 3 1 3 j −i 1
g(i) + + g(m) + + g(j ) + ≥ g(z) dz + − .
2 4 4 2 4 i 4 2
m=i+1
(6.3)
Pólya’s conjecture for Euclidean balls 155

Proof of Lemma 6.4 The left-hand side of (6.3) is equal to (j − i)n − 12 . By convexity
of g,
 j j −i 1
g(z) dz ≤ (g(j ) + g(i)) ≤ (j − i) n − . 
i 2 4

Proof of Theorem 6.1 As g(0) ≥ 14 , we have

 
3
N := g(0) + ≥ 1. (6.4)
4

For k = 1, . . . , N − 1, denote
 
1
Mk = Mg,k := 1 + max m ∈ {0, . . . , b} : g(m) ≥ k + .
4

We also denote MN = Mg,N := 0. The assumption (5.1) yields Mk > Mk+1 for all
k = 0, . . . , N − 1. Therefore, by Lemma 6.4,
  b  
1 3 3
g(0) + + g(m) +
2 4 4
m=1
⎛ ⎞

N −1   Mn −1    
= ⎝ 1 g (Mn+1 ) + 3 + g(m) +
3
+
1
g (Mn ) +
3 ⎠
2 4 4 2 4
n=0 m=Mn+1 +1


N −1  
Mn Mn − Mn+1 1 M0 M0 N
≥ g(z) dz + − = g(z) dz + − .
Mn+1 4 2 0 4 2
n=0

Due to (6.4) we get

b 
   M0
3 M0
g(m) + ≥ g(z) dz + .
4 0 4
m=0

By convexity of g,
 b 1 b − M0
g(z) dz ≤ g(M0 )(b − M0 ) ≤ ,
M0 2 8

and we get (6.1). 

We can now proceed to the proof of Theorem 4.3. We start with

Proposition 6.5 Let λ ≥ 2. Then

λ2 1
P2N (λ) > + R2 (λ), (6.5)
4 4
156 N. Filonov et al.

where
1 4
R2 (λ) := 3G−1
λ −λ 1+ − 3.
4 π

Proof We have, with account of Gλ (0) = πλ ,

  λ 
 
3 3
P2N (λ) = Gλ (0) + +2 Gλ (m) +
4 4
m=1
  λ 
 
λ 3 3
=− + +2 Gλ (m) + .
π 4 4
m=0

We apply Theorem 6.1 to the sum in the right-hand side, with g = Gλ , b = λ, and
 
1
MGλ ,0 = G−1 λ +1
4
(see Remark 6.2). Given that λ ≥ 2, we take into account the bound (4.6) (which
ensures that MGλ ,0 ≤ λ), and the value of the integral from Lemma 4.6, leading to
    
−1 1
λ 2 3 Gλ 4 + 3 − λ − 4 λ
π + 3
4
P2N (λ) − ≥ .
4 4
Finally, we use

x − 1 < x ≤ x

to obtain (6.5). 

We now have

Proposition 6.6 Let λ ≥ 1 , where 1 is given by (1.9). Then R2 (λ) ≥ 0.

 
Proof of Proposition 6.6 Let σ = arccos 56 ∈ 0, π2 , and let

3π/2
λ ≥ r1 (σ ) = √ ,
11 − 5 arccos 56

see (4.8). Then by Lemma 4.8, the bound (4.7) holds. Therefore,

4 3π − 8
R2 (λ) ≥ λ 3 cos σ − 1 − −3=λ − 3,
π 2π
and is non-negative if
6π
λ≥ = 1 > 0.
3π − 8
Pólya’s conjecture for Euclidean balls 157

As 1 > r1 (σ ) (with the inequality easy to prove rigorously using the method of
verified rational approximations discussed in §8), the result follows. 

Theorem 4.3 now immediately follows from Propositions 6.6 and 6.5.

7 Proof of Theorem 4.2

We have the following “dimension reduction” formula.

Theorem 7.1 Let d ≥ 3. Then

 
λ− d2 +1
 n+d −3
PdD (λ) = P̃n+
D
(λ) , (7.1)
d −3 2 −1
d
n=0

where for r ∈ [0, λ] we denote by

   
λ−r 
1 1
P̃rD (λ) := G̃λ,r (0) + +2 G̃λ,r (j ) +
4 4
j =1
   
λ−r 
1 1
= Gλ (r) + +2 Gλ (j + r) + (7.2)
4 4
j =1

the “standard” two-dimensional weighted shifted lattice point count under the graph
of the function G̃λ,r (t) := Gλ (t + r), t ∈ [0, λ − r].

We remark that in comparison to our original definition

 
λ− d2 +1  
 d 1
PdD (λ) = κd,m Gλ m+ −1 + ,
2 4
m=0

see (2.8), where the weights κd,m are attached at each individual abscissa m, the
 
formula in the right-hand side of (7.1) attaches weights n+d−3
d−3 to the whole counts
P̃ d (λ), which we will later estimate using the previously proven Theorem 5.1.
D
n+ 2 −1
Note also that for λ < d2 − 1, the equation (7.1) becomes the trivial identity 0 = 0 by
our notational convention for sums, see Remark 2.2.

Proof of Theorem 7.1 With account of (7.2), the right-hand side of (7.1) reads
 
λ− d2 +1
 n+d −3
d −3
n=0
158 N. Filonov et al.
⎛   ⎞
  λ−n− d2 +1  
⎜ d 1  d 1 ⎟
×⎜
⎝ Gλ n + 2 − 1 + 4 + 2 Gλ j + n + − 1 + ⎟.
2 4 ⎠
j =1

   
For a fixed m ∈ {0} ∪ N, the contribution of Gλ m + d
2 − 1 + 14 in this expression
appears with the factor

m+d −3  j +d −3m−1 j +d −3  j +d −3 m m−1

+2 = +
d −3 d −3 d −3 d −3
j =0 j =0 j =0


m+d−3
i 
m−d−4
i
= +
d −3 d −3
i=d−3 i=d−3

m+d −2 m+d −3
= +
d −2 d −2
m+d −1 m+d −3
= − = κd,m ,
d −1 d −1

where we have used the standard identity [35, Eq. 26.3.7]


r
i r +1
= , r ≥ l, (7.3)
l l+1
i=l

and another identity [35, Eq. 26.3.5],

k k+1 k
= − , k ≥ l.
l l+1 l+1
   
Thus, the contributions of Gλ m + d
2 − 1 + 14 in both sides of (7.1) coincide. 

Before proceeding to the proof of Theorem 4.2 we will introduce some additional
notation and state some auxiliary facts which will be required later. Let, for x ≥ 0,

'
n
n (x) := (x + j ) = (x + 1) · · · (x + n) for n ∈ N, 0 (x) := 1.
j =1

The function n (x) is closely related to Pochhammer’s symbol, or the rising factorial
(x)n := x · · · (x + n − 1) (for which numerous other notation is also used) in the sense
that n (x) = (x + 1)n . We also have

i+j (i + 1) · · · (i + j ) j (i) i (j )
= = = .
i j! j! i!
Pólya’s conjecture for Euclidean balls 159

Let us introduce, for d ≥ 3, a piecewise-constant function

0, if t < d2 − 1,
fd (t) := d−2 (m)     (7.4)
(d−2)! = m+d−2
d−2 , if t − d2 + 1 = m ≥ 0,

and let
 z
Fd (z) := fd (t) dt.
0

In what follows we will require an upper polynomial bound on Fd (z). If z < d2 − 1,

 
then Fd (z) = 0. Let z ≥ d2 − 1 and m = z − d2 + 1 ≥ 0. Then
 z m−1   
 k+d/2 z
Fd (z) = fd (t) dt = fd (t) dt + fd (t) dt
0 k=0 k+d/2−1 m+d/2−1
m−1 
 k+d −2 d m+d −2
= + z−m− +1
d −2 2 d −2
k=0

m+d −2 d m+d −2
= + z−m− +1
d −1 2 d −2
d−2 (m) (d − 1)(d − 2)
= (d − 1)z − (d − 2)m − ,
(d − 1)! 2
(7.5)
where we used (7.3) to evaluate the sum of binomial coefficients.
To establish a bound on Fd (z), we apply the AM-GM inequality

l (β1 · · · β )1/ l ≤ β1 + · · · + βl , l ∈ N, β1 , . . . , βl ≥ 0,

with l = d − 1 and

(m + j )d−1 (m) if j ≤ d − 2,
βj =
zd−1 if j = d − 1,

yielding

(d − 1)zd−2 (m) ≤ d−2 (m) ((m + 1) + · · · + (m + d − 2)) + zd−1

(d − 2)(d − 1)
= (d − 2)m + d−2 (m) + zd−1 .
2
Collecting together the terms with d−2 (m) and substituting the resulting bound into
the right-hand side of (7.5), we deduce that

zd−1
Fd (z) ≤ F̃d (z) := for all z ≥ 0. (7.6)
(d − 1)!
We will require an auxiliary
160 N. Filonov et al.

Lemma 7.2 Let f be a locally integrable function on [0, +∞), and let F̃ ∈
C 1 [0, +∞) with F̃ (0) = 0 be such that
 z
F (z) := f (t) dt ≤ F̃ (z) for all z ≥ 0.
0

Let b > 0, and let g ∈ C 1 [0, b] be a decreasing function such that g(b) = 0. Then
 b  b
f (z)g(z) dz ≤ F̃ (z)g(z) dz. (7.7)
0 0

Proof Integrating by parts, we get

 b  b
F̃ (z)g(z) dz − f (z)g(z) dz
0 0
 b 
= F̃ (z) − F (z) g(z) dz
0
  b  b 

= F̃ (z) − F (z) g(z) − F̃ (z) − F (z) g (z) dz ≥ 0,
0 0
 
since F̃ (0) = F (0) = g(b) = 0 and F̃ (z) − F (z) g (z) ≤ 0, thus proving (7.7). 

We now proceed to the proof of Theorem 4.2 proper.

Proof of Theorem 4.2 First of all, note that for λ < d2 − 1 there is nothing to prove
as in this case PdD (λ) = 0. For λ ≥ d2 − 1, we apply Theorem 5.1 with g = G̃λ,r ,
r = n + d2 − 1, and b = λ − r = λ − n − d2 + 1 to the right-hand side of (7.2), giving
 λ−n− d2 +1  λ
d
P̃n+
D
d
−1
(λ) < 2 Gλ t + n + − 1 dt = 2 Gλ (z) dz
2 0 2 n+ d2 −1
 
for n = 0, . . . , λ − d2 + 1 . We now substitute the results into (7.1), yielding, with
account of (7.3), the bound
 
λ− d2 +1 
 n+d −3 λ
PdD (λ) < 2 Gλ (z) dz
d −3 n+ d2 −1
n=0
⎛  ⎞
 z− d2 +1
λ ⎜  n+d −3 ⎟ (7.8)
=2 ⎜ ⎟ Gλ (z) dz
⎝ d −3 ⎠
2 −1
d
n=0

  
λ z− d
+1 +d −2
=2 2 Gλ (z) dz.
2 −1
d d −2
Pólya’s conjecture for Euclidean balls 161

Using notation (7.4), we rewrite (7.8) as

 λ
PdD (λ) < 2 fd (z)Gλ (z) dz. (7.9)
0

We now apply Lemma 7.2 to the right-hand side of (7.9), taking f = fd , g = Gλ , and
zd−1
F̃ (z) = F̃d (z) = (d−1)! by (7.6), which gives
 λ  λ
2
PdD (λ) < 2 F̃d (z)Gλ (z) dz = zd−2 Gλ (z) dz.
0 (d − 2)! 0

Finally, applying Corollary 4.7 gives

PdD (λ) < Wd (λ)

as required. 

8 Closing the gap: the proof of Theorem 4.4

We describe the algorithm (based on the two Principles stated in §1) of verifying the
statement

λ2
P2N (λ) > for all λ ∈ [3, 14], (8.1)
4
with the lattice point counts given explicitly by
λ
  
3
P2N (λ) = κ2,m Gλ (m) + .
4
m=0

The first Principle is realised with the help of the following simple lemma, which
ensures that our algorithm described below requires only a finite number of steps.

Lemma 8.1 If the inequality (8.1) holds for a particular λ0 , that is, we have

λ20
e(λ0 ) := P2N (λ0 ) − > 0,
4

this inequality also holds for all λ ∈ (λ0 , λ0 + δ(λ0 )) = λ0 , λ20 + 4e(λ0 ) , where


δ(λ0 ) := λ20 + 4e(λ0 ) − λ0 .

Proof The result immediately follows from the facts that P2N (λ) is non-decreasing in
2 λ20
λ and that  = λ0 + δ(λ0 ) is the positive root of the equation 4 = 4 + e(λ0 ). 
162 N. Filonov et al.

To implement the second Principle, we work with rational numbers only. Let, for
x ∈ R, x ≤ x ≤ x, where x, x ∈ Q are some lower and upper rational approximations
of x. The function Gλ (z) does not as a rule take rational values even for rational λ
and z, so to overcome this we work instead with
λ
  
3
PN
2 (λ) = κ2,m Gλ (m) + , λ ∈ Q,
4
m=0

where
1  2 z
Gλ (z) = λ − z2 − z arccos , z ∈ Q ∩ [0, λ].
π λ
Of course, Gλ (z) ≤ Gλ (z) and therefore P N 2 (λ) ≤ P2 (λ), for λ, z ∈ Q. Obvi-
N

ously, taking integer parts of rational numbers, as well as other arithmetic operations
on them is exact and does not introduce any numerical errors. We √ now describe how
to construct verified rational approximations of the square roots · and arccosines
arccos(·). For the former, any guess (say, obtained from numerics) can be directly
verified by taking squares and comparing rationals, which is rigorous. To verify our
approximations of arccosines (taken  at rational
 points) one may proceed as follows.
Define the functions cos, cos : Q ∩ 0, π2 → Q as

cos x := T12 [cos](x), cos x := T14 [cos](x),

where Tt [cos](x) is the Taylor polynomial of cos x at x = 0 of degree t, so that

cos x < cos x < cos x, see, for example, [3, Problem 15].8 Then
   
cos β < x < cos β implies β := arccos x < β = arccos x < arccos x =: β,

and the verification is again reduced to elementary operations on rationals. In the

same manner,
1 1
π = 3 arccos and π = 3 arccos
2 2
provide verified rational approximations for π .
To finish describing our process, we need also to rationalise the square root ap-
pearing in the definition of δ(λ): we effectively replace e(λ) by a smaller number

λ2
e(λ) := P N
2 (λ) − , (8.2)
4
and also replace δ(λ) by a smaller number

δ(λ) := λ2 + 4e(λ) − λ, (8.3)

where a verification is again by taking squares.

8 We thank the referee for pointing out this reference.

Pólya’s conjecture for Euclidean balls 163

Fig. 6 The basic algorithm

The algorithm for proving (8.1)
λ ← 0
stepnumber ← 0
while λ ≤ 1 do
stepnumber ← stepnumber + 1
p ← PN 2 (λ)
2
e ← p − λ4
if e > 0 then
√
λ ← λ2 + 4e
else
print “Proof failed ”; exit
end if
end while
print “Success in”, stepnumber, “steps ”

Remark 8.2 In practice, we use the following process to find lower and upper rational
approximations of a number x ∈ R (which may be a square root, or an arccosine).
Throughout, we fix a relatively small number ε (say, ε = 10−3 ) as an accuracy pa-
rameter. We find numerically some approximation x0 of x (which may be above or
below x) with some better accuracy. Then, we define x as the rational number in the
interval [x0 − 3ε, x0 − ε] with the smallest possible denominator, and x as the ratio-
nal number in the interval [x0 + ε, x0 + 3ε] with the smallest possible denominator,
using a modification of a fast algorithm for traversing the Stern–Brocot tree [5]. As
we always verify the resulting approximations using the procedures described above,
we do not in fact depend on the quality of an original numerical “guess” x0 as long
as |x0 − x| < ε. 

Thus, our main algorithms work as follows. In order to prove (8.1) for λ ∈
[0 , 1 ], we move upwards: set λ = 0 , compute the margin e(λ) from (8.2), set
λnew = λ + δ (λ) using (8.3), and continue the process. If the margins are positive on
each step, the process will stop successfully if after a finite number of steps we reach
λ > 1 , see Fig. 6.
The algorithm works extremely fast (when implemented in Mathematica, see
the footnote on the title page), thus proving Theorem 4.4: in principle, with enough
patience the whole implementation can be done by hand. We summarise its outcomes
in Table 1.

9 Proof of Theorem 1.12

Let Sα be a circular sector of aperture 0 < α ≤ 2π . The eigenvalues of the Dirichlet

and Neumann Laplacians on Sα are easily found by separation of variables. They are
164 N. Filonov et al.

Table 1 Detailed output of the

computer-assisted algorithm Step λ e(λ) δ(λ)

1 3 < 0 3 6
4 13
2 45 1355 223
13 676 221
3 76 868 584
17 289 493
4 164 3368 995
29 841 783
5 187 11687 29
27 2916 27
6 8 3 43
60
7 523 57671 227
60 14400 260
8 374 6098 719
39 1521 897
9 239 10591 339
23 2116 368
10 181 4103 11
16 1024 16
11 12 6 24
25
12 324 2506 241
25 625 400
13 217 7183 271
16 1024 272
14 495 > 
34 1

all simple, and are given by

 2
λm,k = j mπ
α ,k
m ∈ N, k ∈ N,

and
 2
μm,k = j mπ ,k , m ∈ N ∪ {0}, k ∈ N,
α

respectively. Therefore, the corresponding eigenvalue counting functions are

   
αλ αλ
  π    π 
NSDα (λ) = # k ∈ N : j mπ
α ,k ≤ λ and NSNα (λ) = # k ∈ N : j mπ ,k ≤ λ ,
α
m=1 m=0

where the both sums are finite since jν,1 > jν,1 ≥ ν.
Assume for the moment that the sector Sα contains a half-disk, that is α ∈ [π, 2π].
By Proposition 3.1 with account of (3.1), (2.4) and (2.9), we have
 
αλ
π   mπ  
 1
NSDα (λ) ≤ Gλ + . (9.1)
α 4
m=1

Set
πt
gλ,α (t) := Gλ , t ∈ [0, b] ,
α
Pólya’s conjecture for Euclidean balls 165

where
αλ
b := .
π
Then gλ,α is a monotone decreasing convex function on [0, b] with g(b) = 0; more-
over,
 π ( )
π πt 1
gλ,α (t) = Gλ ∈ − ,0 ⊂ − ,0
α α 2α 2
due to our assumption α ≥ π . With this notation, the right-hand side of (9.1) becomes
b 
 
1
gλ,α (m) + ,
4
m=1

and we can estimate it from above directly by Theorem 5.1 with g = gλ,α , giving
b 
   αλ/π 
1 α λ αλ2
gλ,α (m) + < gλ,α (t) dt = Gλ (z) dz = .
4 0 π 0 8π
m=1

Substituting this into (9.1) proves the Dirichlet Pólya’s conjecture for sectors con-
taining a half-disk.
We now turn to the Neumann problem in Sα , still assuming that α ∈ [π, 2π]. Fol-
lowing the same argument as in Lemma
 1.3, we conclude that the Neumann Pólya’s
conjecture for Sα holds for λ ≤ 2 6π
α . We therefore may assume that
*
6π √
λ>2 ≥2 3 (9.2)
α
from now on.
Using once more Proposition 3.1 yields
 
αλ
π   mπ   b  
 3 3
NSNα (λ) ≥ Gλ + = gλ,α (m) + , (9.3)
α 4 4
m=0 m=0

with the same function gλ,α and parameter b as above.

We are now going to use Theorem 6.1 with g = gλ,α to estimate the right-hand
side of (9.3) from below. Note that gλ,α (0) = πλ > 14 by (9.2). Also, by (6.2)
   
−1 1 α −1 1
M0 = Mgλ,α ,0 = gλ,α +1= Gλ + 1,
4 π 4

and in order to apply Theorem 6.1 we need to ensure that M0 ≤ b = αλ

π . But this is
true since, with account of α ≥ π ,
α −1 1 αλ α 1
M0 − b ≤ G +1− ≤ G−1 +1−λ <0
π λ 4 π π λ
4
166 N. Filonov et al.

by (4.6). Then, (6.1) implies

b 
 
3 αλ2 b − 3M0
gλ,α (m) + ≥ − . (9.4)
4 8π 8
m=0

We now √ reason as in the proof of Theorem 4.3: if we can show that b − 3M0 < 0 for
all λ ≥ 2 3, this would prove, via the combination of (9.3) and (9.4), that NSNα (λ) >
αλ2
8π . We have
 
αλ α −1 1
b − 3M0 = −3 Gλ −3
π π 4
α 1 α 1
≤ λ − 3G−1
λ < λ − 2G−1
λ .
π 4 π 4
We now apply the second statement of Lemma 4.8 with σ = π3 which guarantees that
b − 3M0 < 0 for
π  3π
λ > r1 = √ .
3 6 3 − 2π
√
As √3π < 2 3, this finishes the proof of Theorem 1.12 for sectors Sα of aperture
6 3−2π
α ∈ [π, 2π].
To complete the proof of Theorem 1.12 we now need to consider the case α ∈
(0, π). Set
 
2π
:= ≥ 2.
α
Then α̃ = α ∈ [π, 2π], Pólya’s conjecture holds for Sα̃ , and Sα tiles Sα̃ . By Theorem
1.8, Pólya’s conjecture holds for Sα .
Acknowledgements The authors would like to thank L. Friedlander, R. Laugesen, Z. Rudnick, and I.
Wigman for useful discussions. We are also very grateful to the anonymous referee for numerous helpful
suggestions. Research of NF was supported by the grant No. 22-11-00092 of the Russian Science Foun-
dation. Research of ML was partially supported by the EPSRC grants EP/W006898/1 and EP/V051881/1,
and by the University of Reading RETF Open Fund. Research of IP was partially supported by NSERC
and FRQNT. Research of DS was partially supported by an FSRG from DePaul University.

Declarations
Competing Interests The authors declare no competing interests.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as
you give appropriate credit to the original author(s) and the source, provide a link to the Creative Com-
mons licence, and indicate if changes were made. The images or other third party material in this article
are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly
from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/
4.0/.
Pólya’s conjecture for Euclidean balls 167

References
1. Bucur, D., Henrot, A.: Maximization of the second non-trivial Neumann eigenvalue. Acta Math.
222(2), 337–361 (2019). https://doi.org/10.4310/ACTA.2019.v222.n2.a2
2. Colin de Verdière, Y.: On the remainder in the Weyl formula for the Euclidean disk. Sémin. Théor.
Spectr. Géom. 29, 1–13 (2010–2011). https://doi.org/10.5802/tsg.283
3. Dörrie, H.: 100 Great Problems of Elementary Mathematics: Their History and Solution. Dover, New
York (1965)
4. Filonov, N.: On an inequality between Dirichlet and Neumann eigenvalues for the Laplace opera-
tor. Algebra Anal. 16(2), 172–176 (2004). https://doi.org/10.1090/S1061-0022-05-00857-5 (Russian).
English translation St. Petersb. Math. J. 16(2), 413–416 (2005)
5. Forišek, M.: Approximating rational numbers by fractions. In: Crescenzi, P., Prencipe, G., Pucci, G.
(eds.) Fun with Algorithms. FUN 2007. Lecture Notes in Computer Science, vol. 4475, pp. 156–165.
Springer, Berlin (2007). https://doi.org/10.1007/978-3-540-72914-3_15
6. Frank, R.L., Loss, M., Weidl, T.: Pólya’s conjecture in the presence of a constant magnetic field. J.
Eur. Math. Soc. 11(6), 1365–1383 (2009). https://doi.org/10.4171/JEMS/184
7. Freitas, P.: A remark on Pólya’s conjecture at low frequencies. Arch. Math. 112, 305–311 (2019).
https://doi.org/10.1007/s00013-018-1258-x
8. Freitas, P., Salavessa, I.: Families of non-tiling domains satisfying Pólya’s conjecture. arXiv:2204.
08902
9. Freitas, P., Lagacé, J., Payette, J.: Optimal unions of scaled copies of domains and Pólya’s conjecture.
Ark. Mat. 59, 11–51 (2021). https://doi.org/10.4310/ARKIV.2021.v59.n1.a2
10. Friedlander, L.: Some inequalities between Dirichlet and Neumann eigenvalues. Arch. Ration. Mech.
Anal. 116(2), 153–160 (1991). https://doi.org/10.1007/BF00375590
11. Girouard, A., Nadirashvili, N., Polterovich, I.: Maximization of the second positive Neumann eigen-
value for planar domains. J. Differ. Geom. 83(3), 637–661 (2009). https://doi.org/10.4310/jdg/
1264601037
12. Gravel, C.: Géométrie spectrale sur le disque: loi de Weyl et ensembles nodaux. MSc thesis, Université
de Montréal (2007). arXiv:1208.5275
13. Guo, J.: A note on the Weyl formula for balls in Rd . Proc. Am. Math. Soc. 149(4), 1663–1675 (2021).
https://doi.org/10.1090/proc/15343
14. Guo, J., Wang, W., Wang, Z.: An improved remainder estimate in the Weyl formula for the planar
disk. J. Fourier Anal. Appl. 25, 1553–1579 (2019). https://doi.org/10.1007/s00041-018-9637-z
15. Guo, J., Müller, W., Wang, W., Wang, Z.: The Weyl formula for the planar annuli. J. Funct. Anal.
281(4), 109063 (2021). https://doi.org/10.1016/j.jfa.2021.109063
16. Heitman, Z., Bremer, Ja., Rokhlin, V., Vioreanu, B.: On the asymptotics of Bessel functions in the
Fresnel regime. Appl. Comput. Harmon. Anal. 39, 347–356 (2015). https://doi.org/10.1016/j.acha.
2014.12.002
17. Henrot, A.: Extremum Problems for Eigenvalues of Elliptic Operators. Frontiers in Mathematics.
Birkhäuser, Basel (2006). https://doi.org/10.1007/3-7643-7706-2
18. Hethcote, H.W.: Bounds for zeros of some special functions. Proc. Am. Math. Soc. 25(1), 72–74
(1970). https://doi.org/10.2307/2036528
19. Horsley, D.E.: Bessel phase functions: calculation and application. Numer. Math. 136, 679–702
(2017). https://doi.org/10.1007/s00211-016-0853-7
20. Ivrii, V.Ya.: Second term of the spectral asymptotic expansion of the Laplace–Beltrami opera-
tor on manifolds with boundary. Funct. Anal. Appl. 14, 98–106 (1980). https://doi.org/10.1007/
BF01086550
21. Kellner, R.: On a theorem of Polya. Am. Math. Mon. 73(8), 856–858 (1966). https://doi.org/10.2307/
2314181
22. Krätzel, E.: Analytische Funktionen in der Zahlentheorie. Teubner–Texte zur Mathematik, vol. 139.
B. G. Teubner, Stuttgart (2000). https://doi.org/10.1007/978-3-322-80021-3
23. Kröger, P.: Upper bounds for the Neumann eigenvalues on a bounded domains in Euclidean space. J.
Funct. Anal. 106(2), 353–357 (1992). https://doi.org/10.1016/0022-1236(92)90052-K
24. Kuznecov, N.V., Fedosov, B.V.: An asymptotic formula for eigenvalues of a circular membrane. Differ.
Uravn. 1, 1682–1685 (1965). Full text (in Russian) on Mathnet.ru
25. Kwaśnicki, M., Laugesen, R.S., Siudeja, B.A.: Pólya’s conjecture fails for the fractional Laplacian. J.
Spectr. Theory 9(1), 127–135 (2019). https://doi.org/10.4171/JST/242
26. Laptev, A.: Dirichlet and Neumann eigenvalue problems on domains in Euclidean spaces. J. Funct.
Anal. 151(2), 531–545 (1997). https://doi.org/10.1006/jfan.1997.3155
168 N. Filonov et al.

27. Laptev, A.: Spectral inequalities for partial differential equations and their applications. In: Fifth Inter-
national Congress of Chinese Mathematicians, Part 2. AMS/IP Stud. Adv. Math., vol. 51, pp. 629–643.
Am. Math. Soc., Providence (2012)
28. Laugesen, R.S.: Spectral Theory of Partial Differential Equations — Lecture Notes. arXiv:1203.2344
29. Laugesen, R.S., Liu, S.: Optimal stretching for lattice points and eigenvalues. Ark. Mat. 56(1),
111–145 (2018). https://doi.org/10.4310/ARKIV.2018.v56.n1.a8
30. Laugesen, R.S., Liu, S.: Shifted lattices and asymptotically optimal ellipses. J. Anal. 26, 71–102
(2018). https://doi.org/10.1007/s41478-017-0070-5
31. Li, P., Yau, S.-T.: On the Schrödinger equation and the eigenvalue problem. Commun. Math. Phys.
88, 309–318 (1983). https://doi.org/10.1007/BF01213210
32. Lin, F.: Extremum problems of Laplacian eigenvalues and generalized Polya conjecture. Chin. Ann.
Math., Ser. B 38(2), 497–512 (2017). https://doi.org/10.1007/s11401-017-1080-y
33. Liu, S.: Asymptotically optimal shapes for counting lattice points and eigenvalues. Ph.D. dissertation,
University of Illinois, Urbana–Champaign (2017). Full text at the University of Illinois website
34. Melrose, R.: Weyl’s conjecture for manifolds with concave boundary. In: Geometry of the Laplace
Operator. Proc. Symp. Pure Math., vol. 36, pp. 257–274 (1980). https://doi.org/10.1090/pspum/036
35. NIST Digital Library of Mathematical Functions, dlmf.nist.gov. Release 1.1.9 of 2023-03-15. F.W.J.
Olver, A.B. Olde Daalhuis, D.W. Lozier, B.I. Schneider, R.F. Boisvert, C.W. Clark, B.R. Miller, B.V.
Saunders, H.S. Cohl, and M.A. McClain, eds.
36. Pólya, G.: Mathematics and plausible reasoning. In: Two Volumes. Princeton University Press, Prince-
ton (1954). https://doi.org/10.1515/9780691218304
37. Pólya, G.: On the eigenvalues of vibrating membranes. Proc. Lond. Math. Soc. (3) 11, 419–433
(1961). https://doi.org/10.1112/plms/s3-11.1.419
38. Safarov, Yu., Vassiliev, D.: The Asymptotic Distribution of Eigenvalues of Partial Differential Opera-
tors. Translations of Mathematical Monographs, vol. 155. Am. Math. Soc., Providence (1997). https://
doi.org/10.1090/mmono/155
39. Sher, D.A.: Joint asymptotic expansions for Bessel functions. Pure Appl. Anal. (in press). arXiv:2203.
06329
40. van der Corput, J.G.: Zahlentheoretische Abschätzungen mit Anwendung auf Gitterpunktprobleme.
Math. Z. 17, 250–259 (1923). https://doi.org/10.1007/BF01504346
41. Weyl, H.: Über die asymptotische Verteilung der Eigenwerte. Nachrichten der Königlichen
Gesellschaft der Wissenschaften zu Göttingen (1911). Full text available at GDZ Document Server
42. Weyl, H.: Das asymptotische Verteilungsgesetz linearen partiellen Differentialgleichungen. Math.
Ann. 71, 441–479 (1912). https://doi.org/10.1007/BF01456804

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations.

Authors and Affiliations

Nikolay Filonov1,2 · Michael Levitin3 · Iosif Polterovich4 · David A. Sher5

 M. Levitin
M.Levitin@reading.ac.uk; url: https://www.michaellevitin.net
N. Filonov
filonov@pdmi.ras.ru
I. Polterovich
iossif@dms.umontreal.ca; url: https://www.dms.umontreal.ca/~iossif
D.A. Sher
dsher@depaul.edu

1 St. Petersburg Department of Steklov Institute of Mathematics of RAS, Fontanka 27, 191023,
St. Petersburg, Russia
Pólya’s conjecture for Euclidean balls 169

2 St. Petersburg State University, University emb. 7/9, 199034, St. Petersburg, Russia
3 Department of Mathematics and Statistics, University of Reading, Pepper Lane, Whiteknights,
Reading RG6 6AX, UK
4 Département de mathématiques et de statistique, Université de Montréal, CP 6128 succ
Centre-Ville, Montréal QC H3C 3J7, Canada
5 Department of Mathematical Sciences, DePaul University, 2320 N. Kenmore Ave, 60614,
Chicago, IL, USA