DARWIN AND THE ORIGIN OF SPECIES 225 



may be diminished though not extinguished, we are unaware 

 of any reason for supposing that pouters, after a thousand 

 generations of true breeding, have acquired a fresh power of 

 doubling their crops, or that the oldest breed of Arabs are 

 likely to produce ' sports ' vastly surpassing their ancestors in 

 speed. Experiments conducted during the longest time at our 

 disposal show no probability of surpassing the limits of the 

 sphere of variation, and why should we concede that a simple 

 extension of time will reverse the rule ? 



The argument may be thus resumed. 



Although many domestic animals and plants are highly vari- 

 able, there appears to be a limit to their variation in any one 

 direction. This limit is shown by the fact that new points are 

 at first rapidly gained, but afterwards more slowly, while finally 

 no further perceptible change can be effected. Great, therefore, 

 as the variability is, we are not free to assume that successive 

 variations of the same kind can be accumulated. There is no 

 experimental reason for believing that the limit would be re- 

 moved to a greater distance, or passed, simply because it was 

 approached by very slow degrees, instead of by more rapid 

 steps. There is no reason to believe that a fresh variability is 

 acquired by long selection of one form ; on the contrary, we 

 know that with the oldest breeds it is easier to bring about a 

 diminution than an increase in the points of excellence. The 

 sphere of variation is a simile embodying this view; each 

 point of the sphere corresponding to a different individual of the 

 same race, the centre to the average animal, the surface to the 

 limit in various directions. The individual near the centre may 

 have offspring varying in all directions with nearly equal 

 rapidity. A variety near the surface may be made to approach 

 it still nearer, but has a greater tendency to vary in every other 

 direction. The sphere may be conceived as large for some 

 species and small for others. 



Efficiency of Natural Selection. Those individuals of any 

 species which are most adapted to the life they lead, live on an 

 average longer than those which are less adapted to the circum- 

 stances in which the species is placed. The individuals which 

 live the longest will have the most numerous offspring, and as 



VOL. I. Q 



