An implementation of a method for numerical multiple integration based
on a sequence of imbedded lattice rules is given. Besides yielding an
approximation to the integral, this implementation also provides an e
rror estimate which does not require much extra computation. The resul
ts of some numerical experiments conclude the paper.