We show that the average number of characters examined to search for r random patterns of length m in a text of length n over a uniformly distributed alphabet of size a cannot be less than Omega(n log(sigma)(rm)/m). When ...