CORN AND HOG CORRELATIONS 
47 
It will be recalled that summer weight was determined 60 per cent 
by the preceding winter price of hogs (positive) combined with the 
price of corn (negative) at the same time, and that this was a higher 
percentage determination than by any other combination not involv- 
ing these two factors. This result was interpreted as meaning that 
the amount of breeding is determined primarily by the memory of 
profits derived from hog raising during the preceding year. This 
conclusion is supported to some extent by the correlations of winter 
slaughter with price of hogs ( + 0.11) and corn price ( — 0.52) of two 
years before. Winter price of hogs must therefore be taken into the 
central system of relations. 
jThis price itself is determined to the greatest extent by a preceding 
summer weight (indicating breeding) and preceding prices of corn. 
Thus a system of hog-corn relations can be built around the relations 
between breeding (as indicated by summer weight), winter price of 
hogs, and corn price. In fact, a fairly satisfactory central system 
Fig. 27.— A diagram illustrating the system of interactions between corn crop (C), corn price (P), the 
summer price of hogs (S), the winter price of hogs ( W), and the amount of hog breeding (B) in successive 
years, which has been found to be most successful in explaining the observed correlations. The negative 
paths of influence are represented by plain arrows, the positive, by crosshatched arrows. The most 
important paths Cc=— .85, d= +.65, ra= — .65, p P c = — .80) are represented by heavy lines, the least 
important ones (e=— .15, /=+.10, g=+.15) by broken lines, and the paths of intermediate importance 
(a=-.45, b= +.35, fc =+.50, i=-A0, j= -.40, fc =+.45, Z =+.25) by light lines 
can be constructed around these three variables. It is improved by 
taking into account also the summer price of hogs. 
There are many possible paths of influence among these four 
variables. A great many have been tested. The following system 
(fig. 27) has given the most satisfactory results and no system differ- 
ing from it to an important extent has been at all successful. The 
corn crops {(J) and prices (P), summer and winter hog prices (S, W), 
and the hypothetical factor, breeding (B) , are shown for five succes- 
sive years. The paths representing the influence of variables on 
each other are indicated by connecting lines, with arrows indicating 
the direction of influence. The positive effects are indicated by 
crossed lines, the negative by uncrossed lines. 
For the reasons previously discussed, the corn crop is represented 
as a wholly independent variable, unaffected by the hog situation 
or even by the corn crops of the preceding year. Corn price is rep- 
resented as affected only by the crop. 
Summer hog price is represented as influenced by four factors — the 
breeding of the two preceding years (negative effects) and the corn 
