1158 
Journal of Agricultural Research 
Vol. V. No. as 
litter sizes have suggested that segregations of fecundity factors may 
take place. 
(4) Numerous nongenetic factors limit the full expression of the inborn 
possibilities of fertility. 
(5) Certain few somatic characters may be correlated either in a 
physiological or genetic manner with the different degrees of fecundity, 
but the bulk of characters usually assumed to be so related are probably 
entirely independent 
of it. 
(6) Herdbook data 
on the fertility of swine 
present sources of er¬ 
ror, but the percentage 
of error is low enough 
to permit the statistics 
to be suggestive. 
(7) Numerous influ¬ 
ences exist which lower 
the size of litter, which 
sources of error may 
operate in a manner 
compensatory to those 
just mentioned. 
(8) It is questionable 
whether the size of lit¬ 
ter represents the he¬ 
reditary factors trans¬ 
mitted, but the somatic 
character was perforce 
accepted at face value 
in these studies. 
(9) There is no re¬ 
duction in variability 
in the litter sizes of the dams as compared with the grandparents or 
progeny, as would result if there were homozygous differences for fertility 
in the grandparents. Hence, the fertility deviations are either non- 
germinal or else the degree of heterozygosis is so great in the grandparents 
that no increased variability in the F 2 generation is possible. The latter 
explanation is probably the correct one. 
(10) The frequency curves for the 3,540 litters studied make it appear 
that there are at least three centers of deviation in swine fertility. These 
centers possibly correspond to genetic factors involved in the inheritance 
of fecundity. 
Fig. 4.—Diagram of the combined litter frequencies for the three 
generations of swine analyzed into its component curves. 
