Abstract
It is argued that the Griffiths singularity which has been proven to exist in dilute Ising magnets is an artifact of the thermodynamic limit procedure. For finite macroscopically large samples, the behavior will be regular in most experimental sampies while minute sample-dependent corrections, whose observability is uncertain, may appear in an exceedingly small fraction of the real samples. The Griffiths singularity may at most appear by averaging these effects over an extremely large set of samples.

This publication has 13 references indexed in Scilit: