THE ESTIMATION OF DISTRIBUTED LAGS: A COMMENT'