The C# Programming Yellow Book

by Rob Miles

would cause an error as well, since the compiler knows that a double is wider than a float.