Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.
Updated
Statement
Lemma 5 (printed p. 383). For each , however large, there is with the following property. Given integers , (of any size and sign) and , some arithmetic progression with common difference has (a) length , (b) first term in the interval of length , and (c) every term divisible by some prime .
The authors' remarks (p. 383): the lemma "is merely an extension of the Westzynthius--Erdös--Rankin result ([17], [1], [12]) that sometimes exceeds 'any constant' times (to obtain this last, set , )"; the condition means the terms have "rather small factors"; and "the crucial point involves getting the first term to fall between and ."
Source. D. Hensley and I. Richards, Primes in intervals, Acta Arith. 25 (1973/74), 375--391; Lemma 5 and its proof on printed pp. 383--384 (PDF pp. 5--6 of the retained scan), read on the page images.
Read depth. Claims checked: the statement and the remarks were read clause by clause on the page image. The proof (pp. 383--384) was read for its structure; nothing is independently reviewed.
Proof pointer
Combine Lemma 4 (given the sieving bound of Lemma 3, for every some makes every term of divisible by a prime , by the Chinese remainder theorem) with Lemma 3 (): since is fixed, and hold together once and are large. For (b), the first term may be moved by any multiple of , which the prime number theorem for makes about , much smaller than , so can be placed in any interval of length .
Dependencies
Lemma 3 (Mertens's theorem and a two-range hard sieve), Lemma 4 (the Chinese remainder theorem), the prime number theorem for .
Bears on
- Problem 1204: the lemma is what the Polymath paper cites ("It follows from Lemma 5 of [45] that one can take ", p. 78) for the second-order bound (150) on the diameter of the narrowest admissible -tuple, the problem's .