THE EVOLUTION- 11 YPOTHES18. 351 



quite satisfactory, may not unnaturally be expected to make a 

 converse mistake. 



119. The hypothesis of evolution is contrasted with 

 the hypothesis of special creations, in a further respect. It is 

 not simply legitimate instead of illegitimate, because repre- 

 Bentable in thought instead of unrepresentable ; but it has 

 the support of some evidence, instead of being absolutely 

 unsupported by evidence. Though the facts at present as 

 signable in direct proof that by progressive modifications, 

 races of organisms that are apparently distinct may result 

 from antecedent races, are not sufficient ; yet there are nu 

 merous facts of the order required. It has been shown 

 beyond all question that unlikenesses of structure gradually 

 arise among descendants from the same stock. We find that 

 there is going on a modifying process of the kind alleged as 

 the source of specific differences : a process which, though 

 slow in its action, does, in time, if the circumstances demand 

 it, produce conspicuous changes a process which, to all 

 appearance, would produce in the millions of years, and under 

 the great varieties of conditions which geological records 

 imply, any amount of change. 



In the chapters on &quot;Heredity&quot; and &quot; Variation,&quot; con 

 tained in the preceding Part, many such facts were given ; 

 and plenty more might be added. Although comparatively 

 little attention has been paid to the matter until recent times, 

 the evidence already collected shows that there take place in 

 successive generations, alterations of structure quite as 

 marked as those which, in successive short intervals, arise ir 

 a developing embryo nay, often much more marked ; since, 

 besides differences due to changes in the relative sizes oi 

 parts, there sometimes arise differences due to additions and 

 suppressions of parts. The structural modification proved 

 to have taken place since organisms have been observed, is 

 not less than the hypothesis demands bears as great a ratio 



