268 PHYSICAL BASIS OF HEEEBITY 



has cut the knot by postulating an innate adaptive respon- 

 siveness of the animal to every critical situation that calls 

 out a response. The adherents of orthogenesis appeal, 

 apparently — in so far as they commit themselves — to 

 some sort of innate principle that causes advance in com- 

 plexity along one line, and they seem to hint at times even 

 along directed lines of adaptation. Still more elusive are 

 vague appeals made to some unknown principle — some 

 sort of mysterious element, some "Bion," resident in 

 living material and peculiar to it that is responsible 

 for evolution. 



We are not concerned with any of these so-called 

 agents, but there is a relation between chance and evolu- 

 tion shown by living things that has been largely neglected, 

 or at least vaguely referred to, even by natural selection- 

 ists, that is of fundamental importance when evolution is 

 treated as a phenomenon of chance. 



This relation may be stated in a general way as fol- 

 lows : Starting at any stage, the degree of development of 

 any character increases the probability of further stages 

 in the same direction. The relation can better be illus- 

 trated by specific cases. The familiar example of tossing 

 pennies will serve. If I have thrown heads five times in 

 succession, the chance that at the next toss of a penny I 

 may make a run to six heads is greater than if I tossed 

 six pennies at once. Not, of course, because five separate 

 tosses of heads will increase the likelihood that at the 

 next toss a head rather than a tail will turn up, but only 

 that the chances are equal for a head or a tail, so that I 

 have equal chances of increasing the run to six by that 

 throw, while if I tossed six pennies at once the chances 

 of getting six heads in one throw are only once in 64 times. 



Similar illustrations in the case of animals and plants 

 bring out the same point. If a race of men average 5 feet, 

 10 inches, and on the average mutations are not more than 

 two inches above or below the racial average, the chance 

 of a mutant individual appearing that is 6 feet tall is 

 greater than in a race of 5-foot men. If increase in height 



