Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.
Updated
Source. Lemma 2, p. 357, of Yong-Gao Chen, On integers of the form , Proceedings of the American Mathematical Society 129(2), 355--361 (electronically published 28 August 2000), https://doi.org/10.1090/s0002-9939-00-05916-5, the edition named on the source card. A weak form with an ineffective constant is proved in Section 4, part (II), p. 360.
Read depth. Claims checked: the statement was read clause by clause on the printed page; the proof (p. 357) and the weak form (p. 360) were read for structure only. Nothing here is independently reviewed.
Statement
Lemma 2 (p. 357). "Let be distinct odd primes and . Then the number of positive odd integers such that there exist a positive integer and distinct primes with
is less than , where depends only on and ."
The constant is effective (p. 355).
Weak form (Section 4 (II), p. 360). For distinct primes and with , , : those with number fewer than , and those with satisfy , where every depends only on . The bound comes from the Mahler--Ridout theorem (Lemma 3, p. 359), and the paper calls the constants of Section 4 noneffective (p. 355). The paper says this weak form suffices for its purpose.
Proof pointer
Proof on p. 357. For fixed , Yu's bound (Lemma 1, p. 357) applied to gives , equation (2); comparing sizes then gives , equation (3), so , and counting the choices of and the gives the bound, summed over the choices of primes. The weak form replaces Lemma 1 by the Mahler--Ridout theorem (Lemma 3, p. 359).
Dependencies
Lemma 1 (p. 357), a special case of the corollary of Theorem 1 of K. Yu, Linear forms in -adic logarithms, III, Compositio Math. 91 (1994); for the weak form, Lemma 3 (p. 359), from Mahler (1957) and Ridout (1957).
Bears on
- Problem 1113: the paper does not mention the problem. The lemma counts coefficients relative to a prime set fixed in advance; it says nothing about the covering sets of any single coefficient, so it neither produces a Sierpiński number without a finite covering set nor rules one out.