220 SPECULATIVE SCIENCE 



horse to run one part in five thousand faster than his fellows. 

 As a matter of experience, have our racers improved in speed by 

 one part in a thousand during the last twenty generations ? 

 Could we not double the speed of a cart-horse in twenty gene- 

 rations ? Here is the analogy with our cannon-ball ; the rate 

 of variation in a given direction is not constant, is not erratic ; 

 it is a constantly diminishing rate, tending therefore to a limit. 



It may be urged that the limit in the above case is not fixed 

 by the laws of variation, but by the laws of matter ; that bone 

 and sinew cannot make a beast of the racer size and build go 

 faster. This would be an objection rather to the form than to 

 the essence of the argument. The existence of a limit, as proved 

 by the gradual cessation of improvement, is the point which \ve 

 aim at establishing. Possibly in every case the limit depends 

 on some physical difficulty, sometimes apparent, more often con- 

 cealed ; moreover, no one can d priori calculate what bone and 

 sinew may be capable of doing, or how far they can be improved ; 

 but it is unnecessary further to combat this objection, for what*- 

 ever be the peculiarity aimed at by fancy breeders, the same 

 fact recurs. Small terriers are valuable, and the limit below 

 which a terrier of good shape would be worth its weight in 

 silver, perhaps in gold, is nearly as well fixed as the possible 

 speed of a race-horse. The points of all prize cattle, of all prize 

 flowers, indicate limits. A rose called * Senateur Vaisso ' weighs 

 300 grains, a wild rose weighs 30 grains. A gardener, with a 

 good stock of wild roses, would soon raise seedlings with flowers 

 of double, treble, the weight of his first briar flowers. He or 

 his grandson would very slowly approach the ' Cloth of Gold ' or 

 c Senateur Vaisse,' and if the gradual rate of increase in weight 

 were systematically noted, it would point with mathematical 

 accuracy to the weight which could not be surpassed. 



We are thus led to believe that whatever new point in the 

 variable beast, bird, or flower, be chosen as desirable by a fancier, 

 this point can be rapidly approached at first, but that the rate of 

 approach quickly diminishes, tending to a limit never to be 

 attained. Darwin says that our oldest cultivated plants still 

 yield new varieties. Granted ; but the new variations are not 

 successive variations in one direction. Horses could be pro- 

 duced with very long or with very short ears, very long or short 



