Show that the central limit theorem cannot be extended to convergence in probability, that is, if X1, X2, ... are independent, identically distributed random variables with finite variance, then the sum, suitably normalized, converges in distribution to the standard normal distribution but not in probability.