DARWIN AND THE ORIGIN OF SPECIES 249 



each classifier would have pet marks by which to decide the 

 question, in which case the new principle would not be of much 

 practical use; yet if the theory were really true, in time the 

 marks of common ancestry would probably come to be known 

 with some accuracy, and meanwhile the theory would give an 

 aim and meaning to classification, which otherwise might be 

 looked upon as simply a convenient form of catalogue. 



If the arguments already urged are true, these descents from 

 common ancestors are wholly imaginary. ' How, then,' say the 

 supporters of transmutation, ( do you account for our difficulty 

 in distinguishing, d priori, varieties from species ? The first, 

 we know by experience, have descended from a common ances- 

 tor : the second you declare have not, and yet neither outward 

 inspection nor dissection will enable us to distinguish a variety 

 from what you call a species. Is not this strange, if there be 

 an essential difference ? ' 



No, it is not strange. There is nothing either wonderful or 

 peculiar to organised beings in the difficulty experienced in 

 classification, and we have no reason to expect that the dif- 

 ferences between beings which have had no common ancestor 

 should be obviously greater than those occurring in the descend- 

 ants of a given stock. Whatever origin species may have had, 

 whether due to separate creation or some yet undiscovered pro- 

 cess, we ought to expect a close approximation between these 

 species and difficulty in arranging them as groups. We find 

 this difficulty in all classification, and the difficulty increases as 

 the number of objects to be classified increases. Thus the 

 chemist began by separating metals from metalloids, and found 

 no difficulty in placing copper and iron in one category, and 

 sulphur and phosphorus in the other. Nowadays, there is or 

 has been a doubt whether hydrogen gas be a metal or no. 

 It probably ought to be so classed. Some physical properties 

 of tellurium would lead to its classification as a metal; its 

 chemical properties are those of a metalloid. Acids and bases 

 were once very intelligible headings to large groups of sub- 

 stances. Nowadays there are just as finely drawn distinctions 

 as to what is an acid and what a base, as eager discussions 

 which substance in a compound plays the part of acid or base, 

 as there can possibly be about the line of demarcation between 



