12 MAINE AGRICUIvTURAL EXPERIMENT STATION. I917. 



somatic characters of its parents, but by the composition or 

 constitution of the parental gametes. Thus the size of a bean 

 is determined not by the sise of its parent bean, but by the 

 gametic constitution of the latter. 



In the principle above stated "performance" is used in the 

 broadest and most inclusive sense. It may mean performance 

 in the show ring, at the butcher's block, in the milk pail, at 

 the shearing shed, in the trap nest, at the race track, in the 

 pulling contest, etc. The essential point is that it is not pos- 

 sible to tell with any certainty by looking at a cow, for example, 

 or its pedigree, whether the heifers from that cow will be good 

 milkers. Nor does the fact that the cow herself is a superior 

 milker ensure or prove that her heifers will be superior milkers. 

 They may be or they may not. The only way to be sure about 

 it is to t7y them. If they are good milkers then the use of that 

 cow as a breeder is by just so much improving the quality of 

 the herd. Again the fact that a bull's dam made a great record 

 at the pail does not ensure that his daughters will be superior 

 milkers. We can only know whether he possesses the ability 

 to transmit dairy productivity by getting the actual records 

 from some of his daughters. If these records are good the 

 breeding worth of the bull is presumptively high. At any rate 

 we know in that case that he is not lowering the average quality 

 of the herd. Nothing else can furnish the sure and certain 

 kind of information which the actual progeny test furnishes. 



The principle of the progeny test carries with it a certain 

 implication as to the age to which breeding stock should be 

 retained. Obviously if we are to profit from our knowledge 

 as to the breeding worth of a bull gained by the progeny test 

 we must have the bull alive and in breeding condition after we 

 have made the test. This means that we must keep him in the 

 herd longer than bulls are usually kept by Maine breeders. If 

 a herd bull is disposed of before any of his progeny have 

 reached an age where their performance as milkers, for ex- 

 ample, can be measured, then clearly this guiding principle of 

 progeny test is pla3ang no part in the breeding of the herd. 

 Without this principle in active operation the breeder is in 

 much the circumstances of a mariner without a compass. Pro- 



