18 THE PRINCIPLES OF HEEEDITY 



aerial and perhaps even to a terrestrial life. The human fore-limb 

 affords another example. It differs greatly from an organ of 

 mere progression as in the horse ; yet all the stages by which 

 the arm of the man and the fore-leg of the horse have diverged 

 from a common ancestral type may be traced with considerable 

 clearness. It is still an organ of progression in the infant. 

 In this case, as in all others, Darwinians at any rate most of 

 them suppose, not that an entirely new structure has been 

 suddenly developed, but an old structure has been gradually 

 adapted by a change of shape to a new function. 1 



30. The rapidity of evolution, other things being equal, 

 varies with the stringency of the selection. It follows and 

 the deduction is confirmed by our experience of artificial 

 selection that, when the stringency of selection as regards 

 any character falls below a certain point, evolution ceases. 

 If it falls yet lower, particularly if there be complete cessation 

 of selection, the character in question undergoes deteriora- 

 tion. Thus the speed of thoroughbred horses has been 

 evolved by stringent selection ; if we cease to breed with care 

 the race would no longer evolve ; if we permitted indiscri- 

 minate breeding panmixia as it is called the race would 

 deteriorate. There is however a limit to the stringency of 

 selection, and therefore to the rapidity of evolution. The 

 death-rate must not exceed the birth-rate, or the race suffers 

 ultimate extinction. It follows again, since few individuals 

 are pre-eminent in many qualities and none in all qualities, 

 and since different qualities demand for their evolution 

 different sets of eliminating agencies, that a race of complex 

 plants or animals cannot undergo evolution in all its parts 

 at once; for then the birth-rate would fall below the death- 

 rate. The environments in which all races live change slowly, 

 and with each change of environment occurs a gradual change 



1 It is often said by its opponents that Natural Selection is able to 

 improve, but cannot create a structure. Thus we are told that 

 the electric organ of the torpedo cannot have been created by Natural 

 Selection, since in its feeble beginnings the organ cannot have been so 

 useful as to influence the survival rate. But its adherents reply that 

 whenever this objection is raised we may be sure it refers to some case 

 concerning which our knowledge, especially our historical knowledge, is 

 defective. Almost always it refers to some soft structure which has left 

 no fossilized remains from which the past may be traced. But whenever 

 our historical knowledge is sufficiently full we are able to perceive that 

 Natural Selection, through its power of modifying useful structures to 

 different but still more useful functions, may be an all-sufficient cause 

 for the creation of entirely new organs. It is claimed therefore that 

 the Darwinian, unlike the Lamarckian doctrine, affords a complete 

 explanation of all evolution. 



