Is the graph measuring the number of distinct names that end with 'n', or the number of recently born boys who have names that begin with 'n'? If it's the form

Is the graph measuring the number of distinct names that end with 'n', or the number of recently born boys who have names that begin with 'n'? If it's the former, then unpopular names with many distinct spellings may be vastly overrepresented in the graph.