P.G. Ciarlet on the words "linear triangle" and "biquadratic rectangle"

Since the readers of this article are perfectly aware that a triangle may be isoceles, but certainly not genuinely "linear", or that a rectangle may be square, but certainly not genuinely "biquadratic", I have blithely committed these serious abus de langage, which clearly convey more information than the names that I had originally chosen.

P.G. Ciarlet, Handbook of Numerical Analysis, Part 1