SELECTION. 395 



it would be desirable to separate closely-related ani- 

 mals, and subject them to different conditions of food 

 and climate, that their development in all particulars 

 should not be the same. A tendency to a loss of fe- 

 cundity may be corrected by this method; but the 

 conditions to which the animals are subjected should 

 not differ so widely as to destroy the characteristics 

 of the family that it is desirable to retain. 



A defective performance of the function of repro- 

 duction may frequently be corrected by suitable selec- 

 tions within the family, without resorting to a change 

 of conditions or an infusion of other blood. 



As an illustration of the manner in which the 

 latent function is made active, let us take the case of 

 two animals, kept under the same conditions, that are 

 closely related, so that their dominant characteristics 

 are essentially the same. 



The male may exhibit the family defect of a ten- 

 dency to impaired fecundity ; while the female, with 

 the same general bias of the system, may be a good 

 milker. 



When bred together, the acquired quality of se- 

 creting an abundant supply of milk may supplement 

 the conditions that give rise to the family defect, and 

 restore the balance of the organization, so that the 

 function that was comparatively latent in the parents 

 may become active in their offspring. 



It has been claimed that Duke of Airdrie (12730) 

 (see Diagram 2) owes his superiority as a sire to char- 

 acters inherited in accordance with this principle. 



Duke of Gloster (11382), his sire, was not remark- 

 ably prolific, and there was a marked peculiarity in 



