Full Idea
The problem with falsification is that it fails to work with logically particular claims such as 'some swans are purple'. Examining a million swans and finding no purple ones does not falsify the claim, as there might still be a purple swan out there.
Gist of Idea
A proposition such as 'some swans are purple' cannot be falsified, only verified
Source
J Baggini / PS Fosl (The Philosopher's Toolkit [2003], §3.29)
Book Reference
Baggini,J and Fosl,P.S.: 'The Philosopher's Toolkit' [Blackwells 2003], p.131
A Reaction
Isn't it beautiful how unease about a theory (Popper's) slowly crystallises into an incredibly simple and devastating point? Maybe 'some swans are purple' isn't science unless there is a good reason to propose it?