The Horizon Does Not Rise to Eye Level

"The horizon is always at eye level, no matter how high you go. On a globe it would dip below your line of sight."

~$10. Buy 2 ft of 1-inch clear vinyl pipe and two 90-degree elbows; cut into three pieces and assemble a U-tube; fill partly with water (food coloring helps) - that's a homemade clinometer. Climb to the highest place you can find (hill, building, lighthouse). Sight along the two water lines toward the lowest land you can see. Take three photos: sight line centered, at the bottom of the frame, and at the top of the frame (so nobody can claim the photo is skewed).

The horizon sits BELOW your eye level - the sight line dips. The dip angle is ~0.032° × √(h in metres): 15 m (5th floor) → 0.12°; 45.7 m (150 ft hotel) → 0.22°; 100 m → 0.32°. A 'considerable angle downwards' is NOT what the globe model predicts - a small, measurable one, exactly as observed.

The claim goes back to Samuel Rowbotham's 'Earth Not a Globe' (1881), experiment 15 - but his 'experiment' used an unspecified 'clinometer', recorded no elevation, weather or tide, and is not repeatable. The dip is centuries older than the claim: Al-Biruni (11th c.) used horizon dip from a mountain to compute Earth's radius, and British navigation manuals have published dip tables for hundreds of years.

If the horizon were always at eye level, the surface would be flat. Instead the dip is exactly what a sphere of radius 6,371 km predicts - and you can measure it yourself for $10. Modern surveyors confirm it (e.g. a Maine surveyor's measurement from 21.986 m elevation).