in the United States in Fifty Years. 133 



By thus extending our estimate to all the " principal towns" 

 mentioned in the census, we find that the number is increased from 

 thirty-one towns to two hundred and fifty, and that the proportion 

 of town population is augmented from about a thirteenth to near a 

 seventh, with a yet greater disparity among the States than was 

 shown as to the towns of more than 10,000 inhabitants. But this 

 sta.te of facts is, in part, fallacious. It involves an important error, 

 resulting from the application of the term " towns," in New England, 

 to those subdivisions of a country, which are generally called 

 " townships" or " parishes ;" and whose whole population in New 

 England, though the greater part is essentially rural, has, by reason 

 of this inconvenient provincialism, been returned by the census as 

 town population. For the want of adequate means of separating 

 the inhabitants of the town or village from those of the township, 

 (\Vhich, moreover, would, from the irregular dispersion of the 

 buildings, be not always easy even to those on the spot,) the census 

 has been implicitly followed as to these " principal towns" in New 

 England ; though, from the proportion of their inhabitants who are 

 agricultural, it seems probable that more than half their population 

 should be deducted from the town population here estimated. 



In New York, where the same provincialism extensively prevails, 

 the census has erred in an opposite way, by noticing in the northern 

 part of the State none but incorporated cities ; and thus busy and 

 compactly built towns, here called " villages," of 5,000 inhabitants 

 and upwards, have been omitted in one-half the State, while, in the 

 other, much smaller towns, and. even townships, have been occa- 

 sionally noticed ; though in neither district has it descended to 

 towns of but 2,000 inhabitants. To supply these omissions, the es- 

 timate made of the town population of New York, in " Holley's State 

 Register," for 1843, has been adopted. 



Similar omissions of small towns may also have occurred in other 

 States, which we have not the same means of correcting. They, 

 altogether, cannot equal the omissions in New York. 



But were these errors corrected, the three more southern New 

 England States would still have the largest proportion of town pop- 

 ulation of any of the States. The circumstances which determine 

 this proportion, in a State, are the density of its population, the ex- 

 tent of its commerce, and that of its manufactures. It is mainly owing 

 to the first cause, that all the New England and the Middle States 

 have a greater town population than the other divisions. It is from 

 their extensive commerce, that Maryland and Louisiana exceed the 



12 



