48 BULLETIN 1121, U. S. DEPARTMENT OF AGRICULTURE. 
most obvious in the fixation of such characteristics as color, number 
of toes, and tendency toward the production of particular types of 
abnormalities. There has also, however, been a significant differen- 
tiation in the averages made in all elements of vigor. These ele- 
ments of vigor have proved to be inherited independently of each 
other. Each family has come to be characterized by a particu- 
lar combination of traits, usually involving strength in some respects 
with weakness in others. 
Crosses between different inbred families have resulted in a marked 
improvement over both parental stocks in every respect, due allow- 
ance having been made for the effects of size of litter on the other 
characters. This improvement appears to its full extent in the 
progeny of the first cross in the case of adult weight (about 12 per 
cent) and resistance to tuberculosis (about 20 per cent). The mor- 
tality between birth and weaning is found to depend about three- 
fourths on the breeding of the young and one-fourth on that of the 
dam. There is thus a marked improvement in the first cross (about 
11 per cent) in spite of the inbred dam, but there is some additional 
advance in the progeny of a crossbred dam with an unrelated male. 
In the rate of gain between birth and weaning, the breeding of the 
dam and of the young are about equally important. An improve- 
ment of about 16 per cent was obtained in this respect. Birth weight 
depends largely on the dam—about three-fourths—and only one- 
fourth on the breeding of the young. There is thus only slight im- 
provement before the second generation in which it amounted to 
some 9 per cent. The mortality at birth is almost wholly a mater- 
nal affair. Crossbreeding of the dam adds about 7 per cent to the 
chances of the young. The heredity of the young also counts for 
nothing in frequency or size of litter. The sire is somewhat more 
responsible than the dam in the former case; the dam seems to be 
wholly responsible in the latter. Frequency of litter was increased 
over 30 per cent and size of litter over 10 Petes cent when both sire 
and dam were crossbred. 
The number of young raised per year by an average mating depends 
on four of the above elements of vigor—the mortality at birth, that 
between birth and weaning, and the frequency and size of litters. 
The relatively small improvement in crossbred matings in each 
separate respect as given above, is compounded into an advance of 
over 80 per cent in the combination, which goes well beyond the 
superiority of the random-bred control stock over the inbreds. 
Analysis of the various crosses, indicates that the results are all 
the direct or indirect consequence of the Mendelian mechanism of 
heredity. The fundamental effect of inbreeding is the automatic 
increase in homozygosis in all respects. An average decline in 
vigor is the consequence of the observed fact that recessive factors, 
more extensively brought into expression by an increase in homo- 
