Light as a Ray

Light is really a spreading electromagnetic wave, but its wavelength is about half a micron, far smaller than mirrors, lenses, and openings we deal with day to day. When the things light interacts with are huge compared to that wavelength, the wave behavior washes out and light acts like it travels in straight lines. That's the ray model: a bookkeeping trick that lets you trace light's path with a ruler instead of solving wave equations.

A point source actually radiates rays in every direction at once, infinitely many of them. We only draw a handful, usually straight lines from the source through key points like the edges of an object or the corners of a mirror, because those are the rays that decide where edges, images, and shadows land. Everything in between fills in smoothly. This is exactly why a shadow from a point source has a crisp edge: rays either graze past the object's edge and keep going, or get blocked entirely. There's no middle ground, so the boundary between lit and dark is sharp, a true edge with no fuzzy region.

This straight-line picture is the foundation for everything else in geometric optics: reflection off mirrors and refraction into lenses are both just rules for how rays change direction at a surface. But the model has a breaking point. Once light passes through a gap near its own wavelength, or once a source is wide rather than a single point, the simple sharp-ray picture needs upgrading, setting up diffraction and extended sources as the next layer of the story.