106 ILLINOIS STATE DAIRYMEN'S ASSOCIATION 



are almost certain sooner or later, to abandon the pure bred 

 sire and use one of the grades because ''he looks all right." 



The moment this is done, all progress through the sire 

 stops, and no further improvement is possible except through 

 the slow process of selection which is the same a breed imprve- 

 ment. That is, the farmer has stopped half way in securing 

 what has been accomplished by others and goes at it smgle 

 handed and alone with a half improved stock, for half bloods 

 can never give rise to anything but half bloods, and the pro- 

 geny of three-quarter bloods can never be better than three- 

 quarters; they may be less, but they cannot be more. 



By the law of ancestral heredity the offspring inherits from 

 all its ancestry. Now there are 126 of them within the six 

 generations under discussion, and over 2,000 of them within 

 ten generations. It is pretty well-known now, that the pro- 

 geny inherits from these ancestors on the average in pretty 

 definite proportions, depending upon their relative nearness; 

 that is to say, it gets one-half its hereditary influence from 

 its immediate parents, one fourth from its four grand-parents, 

 one-eighth from its eight great-grand parents, and so on back- 

 wards in a decreasing series whose ratio is one- half. 



By this we see that if the immediate parents only are good 

 then the chances are only about one to two that the progeny 

 will be their equal. But if the two parents and all four grand- 

 parents are of the same high excellence then the chances are 

 raised to 75 per cent, whatever the ancestry back of this point. 

 But again, if we go one step further and have the advantage 

 of another generation (16 great-grandparents) of selected ances- 

 try, now the chances arise to 87.5 out of 100; and this fraction 

 will increase with a selected ancestry till it will in time rise to 

 100 per cent or absolute purity. 



This law is the friend of the farmer as long as he persists 

 in using the best sires, and it is the one that proves his undoing 

 when the sire is changed for a part bred animal, no matter how 

 well bred he ''looks." 



The best cow census the Station has been able to take, in- 

 dicates, as Prof. Fraser has shown, that approximately one-fourth 



