centage of agglomerated population would be much larger than that of the latter.
It is thus evident that agglomeration alone does not furnish a true means of distinguishing urban population from rural population according to the meaning attached to those terms in common parlance. An urban population must indeed be agglomerated; but it must also exceed a certain number of inhabitants or it will remain practically rural in character. A German Dorf or village containing, say, 300 peasants, is just as properly called rural as an American township containing 300 farmers living perhaps half a mile apart. Thus it is that modern statisticians have agreed upon a numerical boundary line between city and country. Many objections to this conclusion can of course be raised. It is rightly said to ignore the individuality of communities. Here is a suburban village in close contact with a manufacturing city and possessing most of the characteristics of city life; if its population does not reach the arbitrarily chosen number selected as the qualification for entrance to the urban group, it must be put in the rural class along with the veriest farming community. Here again is the townsman with a country residence; he is in the country but not of it. Nevertheless he must be counted in the rural population. The young manufacturing place, a harbor, a thousand-year old village, a residence town, all places of the same size must be thrown into one class if we follow numerical distinctions. Then again, as population becomes more dense one naturally expects to find the average size of the town or village increasing; the town of 5,000 now plays the role in the formation of social judgments formerly taken by the village of 1,000. Hence the geographical statistician Supan favors a sliding line of division, varying in different countries with the density of population. In his Ortsstatistik he has included all places with over 1,000 inhabitants in thinly