CONTROL OF HOG CHOLERA. 11 
ties and may therefore be regarded as reasonably exact. The figures 
in the columns showing the total deaths represent the total losses 
of hogs from cholera. A large proportion of these occurred either 
in herds where no treatment was applied, because the disease was 
not reported, or took place in herds before inspectors could apply 
the treatment. Striking evidence of this is obtained by comparing 
the total deaths from cnolera in 1914 and 1915 with the deaths among 
treated hogs during the same period. For example, the total deaths 
in 1911 numbered 59,863, of which only 13,985 hogs had been treated. 
In 1915, though there was a total loss of 30,965 hogs, only 9,148 of 
these had been treated. The prompt treatment of all hogs on in- 
fected farms undoubtedly would have reduced the losses still further. 
Inasmuch as the experimental work was carried on in only 3 
of the 14 counties in 1913, that year may be regarded as a fair rep- 
resentative of conditions in the 14 counties, taken as a whole, when 
the work was begun. By comparing the death rate in 1913 with 
that in 1914 and 1915, we find that not only did the actual losses 
from cholera steadily diminish, but that the ratio of deaths per 
thousand decreased to an even greater extent, due to increased hog 
production. In 1914 there were 119,262 fewer hogs lost than in 
1913, and in 1915, 28,898 fewer than in 1914. The death rates for 
the same years were : 1913, 170 per thousand ; 1914, 53 per thousand ; 
and 1915, 23 per thousand. 
During this same period we find that the number of hogs raised 
was constantly increasing, the number being 1,052,408 in 1913, 
1,121,229 in 1914, and 1,334,644 in 1915, an increase of 282,236 hogs 
in two years. 
Again, the question naturally arises as to whether this remarkable 
improvement with respect to losses and to the conditions of the indus- 
try is to be referred primarily to the natural subsidence of the 
disease or to the experimental control work. To decide this we 
must have recourse to comparison with neighboring counties. It 
has been possible to obtain what are believed to be reasonably accu- 
rate estimates of hogs raised and hogs lost during 1913, 1914, and 
1915 in several counties surrounding each of seven of the experi- 
mental counties, namely, Dallas County, Iowa, Pettis County, Mo., 
Montgomery County. Ind., Clay County, Iowa, Gage and Johnson 
Counties, Nebr., Marshall County, Kans., and Renville County, 
Minn. The experimental work was begun in the first three of these 
counties in 1913 and in the remaining four in 1914. Therefore in 
making our comparisons it is necessary to divide the experimental 
coimties into two groups, one representing counties in which work 
was done for three years, and the other those in which only two 
years' work was carried out. Table 3 contains the data relative to 
the first group and Table 4 those of the second group. 
