MISTAIvES OF THE CENSUS. 



demands of local markets, in the form of milk, vegetables, 

 small fruits, poultry and eggs, &c. ; but this change is not 

 fairly indicated in the statistics of the census of 1870. 



The returns of that census wholly omit the production of 

 every description, on more than nine thousand farms of this 

 State, and hence no deductions of any value as to the present 

 condition of our agriculture, or of its condition in 1870 as 

 compared with 1850 or 1860 can be made. The census of 

 1850, for example, states the number of farms as 34,069 ; that 

 of 1860, as 35,601 ; while that of 1870 gives only 26,500,— 

 a difference of 9,101 farms since 1860. Now, apart from the 

 fact that the same causes were operating to increase the num- 

 ber from 1860 to 1870, as from 1850 to 1860, we know that 

 the selectmen and assessors of taxes in each town are far more 

 likely to be correct than the United States marshals, whose 

 jurisdiction extended over many towns, embracing a large area 

 of country. According to the Statistics of Industry of 1865, 

 made up from official returns of the selectmen of each town to 

 the Secretary of the Commonwealth, the number of farms in 

 the State at that date was 46,904, which would leave the 

 number of ftirms that were overlooked in o-atherinjr the cen- 

 sus of 1870 still greater than that stated, or more than 

 twenty thousand instead of nine thousand. 



But we have a still better means of comparison, so far as 

 a few items are concerned; for the assessors in May, 1870, 

 returned the number of cows taxed in the State as 161,185, 

 and in May, 1871, as 162,782 ; Avhile the census of 1870, taken 

 at the same time, returns only 114,771 , — a discrepancy of very 

 nearly 50,000, which can be accounted for in no other way than 

 on the supposition that a large number of farms were entirely 

 overlooked. Again, the assessors, in 1870, return the number 

 of horses as 107,198, and in 1871, as 112,782 ; while the United 

 States census of 1870 returns only 41,039. Now if it be said 

 in explanation, that the number given in the census includes 

 only horses kept on farms, it does not help the matter any ; 

 for the census states the number of horses not on farms as 

 only 45,227,— making the total number in the State, 86,266 

 only, still leaving a discrepancy of 26,516, on the large num- 

 ber of farms whose statistics are not included in the census, 

 having been entirely overlooked by the marshals. Moreover, 



