From analyses of 22,087 vertebral counts, embracing 12 consecutive year-classes and five contiguous areas along the west coast of Vancouver island, it is concluded that more than one population of herring is present, that each area, with the possible exception of one, has an essentially separate run, and that mixture or interchange between areas is generally limited. A significant interaction between mean counts for year-classes and areas might be interpreted as indicating mixing between areas in certain of the year-classes, thus qualifying the general conclusion. Comparison with the conclusions drawn from tagging shows essential agreement. However, the latter method of study demonstrates that mixture between areas does take place and is considerably more extensive than might be anticipated by the term "limited". Furthermore, it reveals that mixture of fish from different areas takes place within the school, i.e., within the sampling unit, and therefore cannot be demonstrated by analyses of mean vertebral counts.