ROTATIONAL CROSSBREEDING AND HETEROSIS 375 



of breed 1, and 50 y^er cent of breed 2. The second year, wherein three breeds 

 are used, the resulting pigs will, on the average, possess: 25 per cent of 

 the chromosomes from breed 1, 25 per cent from breed 2, and 50 per cent 

 from breed ^. The third year, the pigs will possess 62.5 per cent of the 

 chromosomes from breed 1 , 12.5 per cent of breed 2, and 25 per cent of breed 

 3. The fourth year, the pigs will carry, on the average 31.25 per cent of the 

 linkage groups from breed 1, 56.25 per cent from breed 2, and 12.5 per cent 

 from breed 3. The lifth year, the pigs will possess, on the average: 15.63 per 

 cent from breed 1, 28.12 per cent from breed 2, and 56.25 per cent from breed 

 3. From that time on, they will remain in a continuous cross, in about that 

 general state of equilibrium, but the percentage of relationship to the differ- 

 ent breeds will change. 



On the basis of these calculations, we advocated rotational crossbreeding. 

 Some of our critics could not understand how we felt Justified in recommend- 

 ing rotational crossbreeding when our experiments had been carried only to 

 the three-breed cross. Calculations showed so clearly that if the three-breed 

 cross was good, then the continuous cross, by rotation, could not help being 

 successful, insofar as the system of breeding was concerned. On the basis of 

 the theory I have always contended that there was very little advantage in a 

 four-breed cross. Now, however, I am not so sure that that is correct, if we 

 are to take seriously what my commercial hybrid corn producer told me 

 regarding the merits of the double cross of corn in contrast to the single 

 cross. There may be merits in the four- or even the five-way cross that are 

 not generally revealed in short-time experiments. 



We have recommended rotational crossbreeding for commercial swine 

 production, and it seemed, on the basis of theory again, that the rotational 

 scheme of crossing had a particular aptitude for swine production, and was 

 perhaps questionable with other classes of four-footed farm animals. The 

 reason for this is that in swine it is possible for the commercial producer to 

 turn a generation every year if he so desires. I have, however, a number of 

 friends who are breeding commercial flocks of sheep after this general pattern 

 with remarkably good results. If you look at their flocks with the strictly 

 commercial viewpoint, they do not have the variance that most critics of the 

 plan have contended would result. Further than that, the experiments con- 

 ducted by the United States Department of Agriculture with beef cattle and 

 dairy cattle have shown that the same basic principles ap})ly to these classes 

 of livestock as in swine. Dairymen have perhaps been more reluctant to de- 

 part from the purebred philosophy of breeding than any other group of live- 

 stock breeders. Yet by a strange coincidence, the experiments of the United 

 States Department of Agriculture are showing a greater increased yield as 

 the result of crossing dairy cattle than the crossing of any of our other species 

 of farm animals. Their data show an increase of 25 per cent in milk and 32 

 per cent in butterfat yield. 



