and 2,000 inhabitants, and 209 of between 2,000 and 4,000 inhabitants each. Of the former class the majority were doubtless largely if not altogether rural communities, as were many of the latter; and considerable rural population is often included in New England towns with a still greater aggregate number of inhabitants. The impossibility of drawing as hard and fast a line between the rural and urban or semi-urban population in New England as may be done in the other portions of the country is of course due to the fact that while geographically the towns in New England correspond to the towns, townships, election, militia, or magisterial districts, hundreds, wards, precincts, beats, etc., into which the counties in the other parts of the country are divided, it is in New England very unusual to incorporate a village or borough within a town. When separate government is desired by a portion of a New England town, the more common practice is to set off the area asking for it as a new town. It is not possible to determine with mathematical precision the precise increase during the decade of the 3,715 places having each over 1,000 inhabitants in 1890. In a number of instances the territorial limits of cities and towns were not the same in 1890 as in 1880. Usually, of course, when changes have occurred there have been extensions of corporate boundaries. Among the smaller towns and villages there are many whose population in 1880 was not separately returned. In some instances the places did not exist in 1880, but more frequently their not being mentioned in the census was due to failure of the enumerators to separate their inhabitants from the persons residing in other portions of their census districts. Both the circumstances last mentioned would operate to make the apparent increase in the population of the cities and towns greater than it actually was. On the other hand, many of the larger cities are surrounded by more or less extensive belts of territory outside of their corporate limits, the increase of the population of which belts is due entirely to the growth of the cities around which they lie. On the whole, therefore, it is believed that to compare the population as returned by the eleventh census of cities, towns, and villages of 1,000 inhabitants or upward in 1890 with the population of the same places as returned in 1880 will afford a practically accurate measure of the rate of growth of the city, town, and village population of the country as a whole. In particular States, however, one or the other of the above causes of error may so predominate as to exert an appreciably disturbing influence on the accuracy of the comparison. In local comparisons, therefore, proper allowance has been made whenever necessary for the operation of these causes. The increase of the rural and urban population, as above defined, during the decade has been:
