The German social statistician Sombart rightly asserts that in many cases we obtain a sharper and better impression by limiting our investigations to the conditions prevailing in a single year-class than by taking several ageclasses together, which are always in some respects heterogeneous.