>>366
That the squares of sine and cosine sum to 1. Or rather that R[sin,cos] is but R[x,y]/(1-x^2-y^2).