10 



BULLETIN 1121, U. S. DEPARTMENT OF AGRICULTTTKE. 



nate represents the percentage of deaths, to the other side, survivals. 

 The change in the percentage of deaths due to a given shift in the 

 conditions is equal to the area between two ordinates at a given dis- 

 tance apart. This area is of course greater at the middle of the 

 curve (50 per cent deaths) than toward either limit. On this hypoth- 

 esis a table of probability integrals can be used for comparing per- 

 centage differences at different points of the range between per cent 

 and 100 per cent. 



The numbers in Table 2 give the departure of each experiment 

 from the total inbred stock, adjusted on the above basis to the record 

 of the latter for the whole four years 1916-1919, i. e., 77.7 per cent 

 born alive, 72.1 per cent raised of those born alive, and 56.3 per cent 

 raised. This correction, it will be noticed, makes the records of AC 

 and CC, with which this discussion started, approximately equal. 

 So far as most of the conclusions are concerned, it makes little differ- 

 ence whether the correction is made or not. It seems important, 

 however, to show the order of effect. 



ALLOWANCE FOR HEREDITY. 



Another consideration, which has doubtless occurred to the reader 

 m comparing the records of the inbreds and crossbreds, is the exact 

 heredity of the latter. There are fairly large differences among the 

 inbred families themselves. If only animals from the better families 

 were used in making crosses, the latter would naturally be superior, 

 apart from the effect of the system of mating. 



Table 3. — The inhred ancestry of the males and females used in the various crossbreeding 



experiments, in percentages. 



[Each mating weighted by the number of litters produced through 1919.] 



Inbred family. 



CO. 



01. 



C2. 



CC. 



CA. 



AC. 



Sire. 



Dam. 



Sire. 



Dam. 



Sire. 



Dam. 



Sire. 



Dam. 



2 



16.1 

 7.7 

 5.3 

 11.1 

 16.6 



19.3 

 23.6 

 7.2 

 5.1 

 7.9 



9.7 

 16.1 

 4.2 

 6.5 

 18.3 



8.0 

 8.0 

 5.3 

 



18.7 



5.4 

 16.3 

 5.6 

 7.5 

 21.0 



5.8 

 23.7 

 2.9 

 4.2 

 17.7 



20.9 

 21.8 



8.5 

 9.8 

 17.7 



18.4 

 42.4 

 26.6 

 12.0 

 0.6 



16.9 

 18.0 

 15.7 

 27.3 

 17.4 



27.0 



13 



15.2 



32 



7.0 



35 



7.0 



39 



17.3 







7. .. 



4.8 



3.1 



3.8 



4.1 



4.1 



7.0 







4.6 



3.8 



5.5 



2.4 



2.7 

 3.8 

 6.8 

 1.2 

 6.5 

 2.1 

 0.9 

 2.7 

 1.7 

 4.1 

 4.4 



0.6 



4.7 



8.7 



3.0 



3.6 



2.4 







4.4 



2.8 



6.7 



8.3 







5.3 

 12.0 







8.0 









 10.7 



0.7 



8.0 

 15.3 



1.0 

 



6.5 

 0.6 

 3.5 

 7.9 

 



4.0 

 10.5 

 5.4 

 4.8 



2.5 

 1.0 

 1.0 

 2.6 

 5.8 

 4.2 

 



5.0 

 7.9 

 4.0 

 11.8 







0.9 



1.3 



1.3 



2.5 



2.5 



1.3 



1.6 



9.5 







0.3 





 

 

 

 

 

 

 

 

 

 











3.5 































1.2 



3.2 



9 



2.0 



17 — 



18 



0.6 

 2.9 



20 



5.3 



23 



1.7 



24 







31 



1.7 



34 



1.2 



36 



4.7 



38 



3.2 







OI 



4.3.2 



36.9 



45.2 



60.0 



44.2 



45.7 



21.3 







4.7 



26.5 







As it was, however, an effort was made to use all of the families. 

 So far as any preference was given it was to the weaker ones. The 

 inbred ancestry of the sires and dams of the litters produced in each 



