Thus, in order to implement the polynomial, it is suﬃcient to store its coeﬃcients
a0, a1, a2, . . . , an.
The naive way to do this is in an (n + 1)-dimensional vector. This approach, however, is not suﬃciently general and ﬂexible. Indeed, in a vector object, the coeﬃcients would actually be stored in an array, which requires that all of them are of the same size. This is good enough for polynomials of one variable, in which the coeﬃcients are scalars, but not for polynomials of two or more variables, in which the coeﬃcients are themselves polynomials that may have variable sizes.