Full Idea
There is a case for calling a generalisation a law when its only confirmation is the multiplication of instances, if they don't conflict with other criteria. In fact any supported generalisation could count as a law if there is no counter-evidence.
Gist of Idea
We could call any generalisation a law, if it had reasonable support and no counter-evidence
Source
Harré,R./Madden,E.H. (Causal Powers [1975], 2.II)
Book Reference
Harré,R/Madden,E.H.: 'Causal Powers: A Theory of Natural Necessity' [Blackwell 1975], p.33
A Reaction
This is the beginning of the modern doubts about laws of nature, fully articulated in Mumford 2004. It seems to me inescapable that laws drop out if our ontology is based on powerful particulars. They are just patterns of outcome.