42 A TREATISE ON HORSE-BREEDING. 



by no means certain; for, as I have remarked, 

 only the general and firmly-fixed characteris- 

 tics which distinguish the species are trans- 

 mitted with absolute certainty, and the trans- 

 mission of accidental qualities or especial ex- 

 cellence in any given particular, while always 

 possible, can never be depended upon with 

 certainty. If, however, we select parents both 

 distinguished for the same accidental variation 

 or accidental excellence the chances that this 

 will be transmitted to the offspring are, theoret- 

 ically, twice as great as when only one of the 

 parents is in possession of the desired quality; 

 and if in the produce from this coupling we see 

 manifestations of the desired tendency we may 

 unite the animals so bred with an increased 

 probability that they, in turn, will transmit it 

 to their offspring. It is mainly .by this process 

 of selection and coupling, with a view to the 

 attainment of certain desirable qualities, per- 

 severed in for many generations, that all noted 

 breeders have succeeded in moulding the forms 

 or establishing the breeds that have given them 

 celebrity. 



It must be borne in mind that the very proc- 

 esses of Nature w^hich make it possible for man 

 to effect improvement in any species of domes- 

 ticated animals conspire to make the work of 

 creating a new type of heterogeneous materials 

 extremely difficult. In making selections with 



