The impossibility of separating age, period and cohort effects by Mr Andrew Bell