hin partitions do their bounds divide.' There cannot be a surer proof of a low origin or of an innate meanness of di