## Access

You are not currently logged in.

Access your personal account or get JSTOR access through your library or other institution:

If you need an accessible version of this item please contact JSTOR User Support

# On the Number of False Witnesses for a Composite Number

Paul Erdös and Carl Pomerance
Mathematics of Computation
Vol. 46, No. 173 (Jan., 1986), pp. 259-279
DOI: 10.2307/2008231
Stable URL: http://www.jstor.org/stable/2008231
Page Count: 21
If you need an accessible version of this item please contact JSTOR User Support
Preview not available

## Abstract

If $a$ is not a multiple of $n$ and $a^{n - 1} \not\equiv 1 \operatorname{mod} n$, then $n$ must be composite and $a$ is called a "witness" for $n$. Let $F(n)$ denote the number of "false witnesses" for $n$, that is, the number of $a \operatorname{mod} n$ with $a^{n - 1} \equiv 1 \operatorname{mod} n$. Considered here is the normal and average size of $F(n)$ for $n$ composite. Also considered is the situation for the more stringent Euler and strong pseudoprime tests.

• 259
• 260
• 261
• 262
• 263
• 264
• 265
• 266
• 267
• 268
• 269
• 270
• 271
• 272
• 273
• 274
• 275
• 276
• 277
• 278
• 279