ADAPTATION. 233 



is a reason which every one recognizes as valid. It is ob- 

 servable, too, that the rapidity and completeness with which 

 an artificial power is lost, is proportionate to the shortness of 

 the cultivation which evoked it. One who has for many 

 years persevered in habits which exercise special muscles or 

 special faculties of mind, retains the extra capacity produced, 

 to a very considerable degree, even after a long period of 

 desistance; but one who has persevered in such habits for 

 but a short time has, at the end of a like period, scarcely any 

 of the facility he had gained. Here too, as before, suc- 



cessions of organisms present an analogous fact. A species 

 in which domestication continued through many generations, 

 has organized certain peculiarities; and which afterwards, 

 escaping domestic discipline, returns to something like its 

 original habits; soon loses, in great measure, such peculiari- 

 ties. Though it is not true, as alleged, that it resumes com- 

 pletely the structure it had before domestication, yet it 

 approximates to that structure. The Dingo, or wild dog of 

 Australia, is one of the instances given of this ; and the wild 

 horse of South America is another. Mankind, too, supplies 

 us with instances. In the Australian bush and in the back- 

 woods of America, the Anglo-Saxon race, in which civilization 

 has developed the higher feelings to a considerable degree, 

 rapidly lapses into comparative barbarism: adopting the 

 moral code, and sometimes the habits, of savages. 



§ G8. It is important to reach, if possible, some rationale 

 of these general truths — especially of the last two. A right 

 understanding of these laws of organic modification underlies 

 a right understanding of the great question of species. While, 

 as before hinted (§40), the action of structure on function 

 is one of the factors in that process of differentiation by 

 which unlike forms of plants and animals are produced, the 

 reaction of function on structure is another factor. Hence, 

 it is well worth while inquiring how far these inductions are 

 deductively interpretable. 



