ORIGIN OF VARIATIONS. 153 



yet we are certain of the fact that long disuse 

 weakens co-ordinations, and a cessation of stimuli 

 weakens associations and tends finally to obliterate 

 them. Therefore the withdrawal of those stimuli 

 sustaining that part of the inherited impulse which 

 causes the development of the peculiarities of a land 

 animal would, in the course of generations, destroy 

 that part of the inherited impulse. A gradual ret- 

 rogression must precede the addition of new parts. 

 Thus it seems probable that the legs of such car- 

 nivores as we have imagined would first go through 

 a period of shortening before they would assume the 

 breadth of flippers like those of a seal. 



It is not so difficult to imagine how, under such 

 conditions as I have just mentioned, a limb might 

 gradually grow smaller, its growth being checked 

 earlier and earlier in each generation. Since, in its 

 early development, the brain surpasses all other or- 

 gans of the body, we may suppose that the part of 

 the nervous system which governs the growth and 

 action of a limb would be developed — in fact, must 

 be developed — before the limb shall have attained 

 any size or functional power. But if, through un- 

 toward circumstances, that part of the nervous sys- 

 tem be not stimulated to activity, then its co-ordina- 

 tions which govern the growth of limb in both indi- 

 vidual and species, will gradually be weakened, and 



