A note on the use of temporal frequency distributions in studies of prehistoric demography