176 THE HORSE 



knowledge is not complete) ; neither should we dis- 

 card knowledge which is not absolutely correct. Most 

 knowledge is only relatively correct, but, if it is accurate 

 enough to aid us in our research and enlighten our 

 judgment, it is still valuable. The chemist cannot 

 make his analyses absolutely correct. Are they there- 

 fore of no value ! The steel rule may be shortened by 

 a little cool air, and lengthened by the mild rays of 

 the sun; is it therefore of no value in measuring 

 length and breadth because it is slightly inaccurate when 

 placed under slightly changed conditions ? While the 

 steel rule is never again absolutely a foot long after 

 it is made, except when placed in the same temperature 

 as that in which it was made, yet it varies only within 

 narrow limits. So, although it is seldom absolutely 

 accurate, it serves the world well for all uses to which 

 it is ordinarily put. 



We speak of the "science" of breeding. This term 

 is used loosely, and is too often made to cover both 

 the science and the practice of breeding. Science is 

 usually defined as ascertained facts, truths logically 

 arranged. For instance: it is a fact that like produces 

 like under exactly the same conditions. If two parents 

 are similar, or of nearly identical blood, and their 

 environment and potency similar, their offspring will 

 be similar to the parents. But it is also a fact that 

 precisely the same conditions, in all particulars, are 

 never present in the production of any two animals, 

 and, therefore, no two animals, though from the same 

 parents, are exactly alike. Thus it appears that the 

 forces of reproduction and growth, food and environ- 



