Full Idea
Theories that take colours to be properties of the surfaces of objects have difficulty accounting for a host of phenomena including coloured light emitted by radiant sources and so-called film colours (the colour of the sky, for instance).
Gist of Idea
Colours aren't surface properties, because of radiant sources and the colour of the sky
Source
John Heil (From an Ontological Point of View [2003], 17.4)
Book Reference
Heil,John: 'From an Ontological Point of View' [OUP 2005], p.201
A Reaction
Personally I never thought that colours might be actual properties of surfaces, but it is nice to have spelled out a couple of instances that make it very implausible. Neon and sodium lights I take to be examples of the first case.