Skip to content

Erdős Problem #5

Editorial depth: Reviewed explanatory article — evidence qualified.
Snapshot classification: open (not automatically the live result).
Subject areas: number theory, primes.

Read this status correctly

The status above is reproduced from GCL's immutable external-catalogue snapshot. It is not a new claim of current mathematical resolution, a source-semantic review, an independent proof check, or GCL certification. A published proof and a GCL replay are separate events.

The question in plain language

The gap between consecutive primes is irregular: sometimes two primes occur unusually close together, and sometimes a long stretch contains none. Let \(p_n\) be the prime of index \(n\) and write

\[ g_n=\frac{p_{n+1}-p_n}{\log n}\qquad(n\ge2). \]

Erdős Problem #5 asks whether, for each prescribed real number \(C\ge0\), there are infinitely many increasing prime indices along which \(g_n\) approaches \(C\). Equivalently, does every nonnegative real number occur as a finite limit point of the normalized prime-gap sequence? This problem is open in the reviewed 2026 sources.

A small example

The consecutive primes \(5,7,11\) have successive gaps 2 and 4. With indices \(n=3\) and \(n=4\), the corresponding normalized values are \(2/\log3\) and \(4/\log4\). A few such values demonstrate the definition but tell us nothing about limit points requiring arbitrarily large indices.

There is a useful normalization equivalence. The prime number theorem implies

\[ \frac{\log p_n}{\log n}\longrightarrow1. \]

Hence dividing the prime gap by \(\log p_n\) instead of by \(\log n\) yields exactly the same finite subsequential limits. This elementary equivalence depends on the prime number theorem; it does not prove the limit-point conjecture.

What is known

Classical results establish extremely large normalized gaps (Westzynthius) and zero as a limit point (Goldston–Pintz–Yıldırım). Later results identify substantial parts of the spectrum. In particular Jori Merikoski, Limit points of normalized prime gaps, Journal of the London Mathematical Society 102 (2020), 99–124, proves that the set \(S\) of finite limit points has

\[ \mathrm{meas}(S\cap[0,T])\ge \frac{T}{3} \qquad(T\ge0), \]

and that gaps between finite limit points are bounded by an absolute constant. This is a theorem about at least one third of each interval by measure, not a theorem that every nonnegative \(C\) lies in \(S\).

GCL's current state

GCL source audit #1274 checked the current question and the published Merikoski theorem's abstract. It specifies the precise logarithm normalization and a bounded source-fidelity task. The old admitted GCL formal-provider snapshot did not contain a file for #5; a newer upstream statement cannot silently replace that pinned source. This is editorial analysis, not certification of the primary analytic proof.

What remains

Show every prescribed finite \(C\ge0\) is attained as a limit along an unbounded sequence of consecutive-prime indices, or find a missing value. Positive measure, a small initial interval and bounded gaps between limit points do not imply that all possible values occur.

Sources and claim boundaries

Catalogue and evidence provenance

  • Pinned catalogue source: teorth/erdosproblems at 536408716724.
  • GCL issue: #1274 — protected issue/assignment identity; its labels and results are live, not reproduced automatically here.
  • GCL triage: FORMAL_SOURCE_REFRESH_REQUIRED.
  • Formal-statement coverage: No statement was confirmed in GCL's pinned formal-provider snapshot. Presence does not mean the statement or proof has been checked.
  • External mathematical status: consult current problem page and its dated bibliography.

Back to catalogue · Compendium introduction