home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!destroyer!sol.ctr.columbia.edu!usc!chaph.usc.edu!news
- From: rmurphy@aludra.usc.edu (Bob Murphy)
- Newsgroups: sci.math
- Subject: Re: u(v^n)w prime puzzle
- Date: 19 Aug 1992 13:56:16 -0700
- Organization: University of Southern California, Los Angeles, CA
- Lines: 50
- Message-ID: <l95dbgINN8t6@aludra.usc.edu>
- References: <l93702INNaq9@aludra.usc.edu> <1992Aug19.153755.14825@wri.com>
- NNTP-Posting-Host: aludra.usc.edu
-
- In article <1992Aug19.153755.14825@wri.com> roach@bikini.wri.com (Kelly Roach) writes:
- >In article <l93702INNaq9@aludra.usc.edu> rmurphy@aludra.usc.edu (Bob
- >Murphy) writes:
- BM>> If there were such values of u, v, and w, then one could construct
- BM>> arbitrarily large prime numbers. Since I have often heard
- BM>> reference to "the largest known prime number", I would guess that
- ^^^^^
- BM>> there are no such values. Now if only I had a proof.
-
- KR> Not quite. The largest known prime number is 2^756839-1.
- KR>But emphasize the word "known" here.
-
- If you look carefully at my posting I said 'known'. I am well aware
- that there is NO largest prime number. I simply took the fact that a
- Mersenne number is the largest KNOWN prime and used it to GUESS that
- no such u,v,and w existed. I specifically said that I don't have a
- proof.
-
- KR> There ARE infinitely many prime numbers and they do
- KR>get arbitrarily large. A classic proof, due to Euclid is to
- KR>take
- KR>
- KR> P = p1*p2*p3*...*pN+1
- KR>
- KR>where p1,p2,p3,...,pN are the primes you already know.
- KR>The number P is not divisible by any of p1,p2,p3,...,pN.
- KR>All the primes dividing P are new primes.
-
- Actually, Euclid's proof is slightly different than above.
-
- Euclid's proof:
-
- Suppose that p1 < p2 < ... < pn are all the primes.
- Let P = p1*p2*...*pn + 1 and let p be a prime dividing P;
- then p cannot be any of p1,p2,...,pn, otherwise p would divide
- the difference P - p1*p2*...*pn = 1, which is impossible.
- So this prime p is still another prime.
- This is a contradiction.
- Hence there are an infinite number of primes.
-
-
- Bob Murphy (rmurphy@aludra.usc.edu)
-
-
- p.s. For those interested, Euclid's proof can be found as
- Proposition 20 in Book IX of The Elements. However, for a
- more readable proof, I would suggest The Book Of Prime Number Records
- by Paulo Ribenboim. This book gives, as the author puts it,
- 9 and 1/2 proofs on the infinitude of primes.
-
-