566 Wisconsin Academy of Sciences,, AriSj and Letters. 



Chaptee YI. 



some cokrections ix the united states and state census 



eepoets eeom 1850 to 1870. 



It was discovered upon comparing the results obtained in this 

 study with those of the United States and State Census returns 

 that there were considerable variations. The following tables 

 show the differences for the years 1850, 1860, and 1870. In 

 column one the population is given which was obtained from the 

 records already referred to. Columns 2 and 3 show these varia- 

 tions from the United States Census returns. The plus sign 

 before the number indicates that the returns for the towns ex- 

 ceed and the minus sign that they are less than those found in 

 the records. The blanks indicate that there is no variation. 



In getting the results from the records a table similar 

 to Table 11 was used. The number of persons on each page of 

 the Census report from each separate state or country was 

 counted and the result placed in a small square opposite the 

 name of the state or country. When all the names on the page 

 were classified according to nativity, the column was added and 

 the result placed below. Then to make sure that no names were 

 omitted, the names on the page were counted and placed above 

 the column. This result had to correspond to the footing if the 

 Avork was correct. When all the material for a town was exam- 

 ined the number from each state or country was added and the 

 same placed in a column to the right. To still further avoid a 

 mistake and to be positive that the results were correct, the hori- 

 zontal and vertical colmnns of sums were added, and if these 

 results were equal the work was considered correct. 



