Ratio INIf.tiiod of Determixixg Cost of Producing Hogs 31 



1864 7.3 1874 8.9 



1865 16.3 1875 11.8 



1866 16.2 1876 15.3 



1867 7.2 1877 11.5 



Ten-year average 10.6 Ten-year average 11.7 



1878-1887. 1888-1897. 



1878 9.7 1888 12.3 



1S79 10.3 1889 12.5 



1880 12.3 1890 9.9 



1881 12.1 1891 7.4 



1882 10.9 1892 11.8 



1883 11.3 1893 16.5 



1884 10.5 1894 11.6 



1885 9.7 1895 10.8 



1886 11.1 1896 10.7 



1887 12.4 1897 14.2 



Ten-year average 11.0 Ten-year average 11.8 



1898-1907. 1908-1917. 



1898 11.6 1908 8.4 



1899 12.0 1909 11.3 



1900 13.2 1910 15.2 



1901 11.8 1911 11.2 



1902 11.6 1912 10.9 



1903 13.0 1913 13.2 



1904 10.2 1914 11.7 



1905 10.4 1915 9.6 



1906 13.4 1916 11.5 



1907 '. 11.4 1917 9.7 



Ten-year average 12.2 Ten-year average 11.3 



To refine the method to meet market conditions, we need to 

 know the ratio between corn and hogs at different seasons of tlie 

 year. There are seasonal periods of over-supply and scarcity of 

 both corn and hogs. In November, for instance, the 190T-1916 

 price of corn was 67.2 cents and the price of hogs .$7.23, or a ratio 

 of 10.6 bushels to one hundred pounds of hog flesh, while in March 

 of the ten-3'ear period the average price of corn was 61.7 cents 

 and the price of liogs $7.66, or a ratio of 12.4 bushels of corn for 

 one hundred pounds of hog flesh. In like manner, there is a fairly 

 normal ratio for each month of the year and for each week of the 

 year. All this is on the assumption that hogs are simply con- 

 densed corn. It docs not take into account the fact that hogs have 

 been made out of corn at varying values during a period of about 

 a year preceding time of marketing. Obviously, then, we must 

 have a composite corn value. In matters of this sort, statisticians 

 know that it is absolutely impossible to weight matters so as to 

 represent actual conditions, but at tlie same time the)' know that 

 absolute accuracy is not at all essential, that in fact a difference in 

 wcigliting will ordinarily make very little difference in results. 



