14 THE CLASSIFICATION OF BIRDS 



It is manifest that if either of these propositions be accepted 

 the true classification of Birds becomes a matter of opinion, incap- 

 able of proof, and that we can only accept the system as least 

 likely to be wrong which involves the independent acquirement, 

 or the acquirement and subsequent loss, of very few or very 

 unimportant characters. It is easy to count the number of char- 

 acters ; it is very difficult to say which are easy and which are 

 difficult to acquire or to lose. 



It is a disputed point amongst biologists whether characters 

 acquired after birth are or are not transmitted to offspring. Very 

 little satisfactory evidence has been produced on either side, but 

 much discussion has arisen, and the question appears to be one 

 that will give occasion for much difference of opinion for years to 

 come. The subject is of great importance in its relation to the 

 origin of species, but it has no bearing upon their classification. 

 The extent to which characters can be independently acquired is 

 an entirely different question, and is one which lies at the root of 

 all attempts at classification. It is a difficult problem. The mass 

 of evidence is overwhelming so enormous that it is almost impos- 

 sible adequately to grapple with it. It cannot be too emphatically 

 asserted that characters derived from inheritance are all-important 

 as bases of classification, whilst characters independently acquired 

 are absolutely valueless. 



All attempts to find a natural classification of Birds must be 

 based upon the laws of heredity. It is a universal law that, in an 

 overwhelming majority of cases, parents transmit their characters 

 to their offspring ; but were this law absolute the development of 

 new species would be impossible, and the assumption that all 

 existing species had been specially created would be inevitable. 

 It is unquestionably true that the characters of an individual may 

 be slightly changed by use or disuse, or by other effects of environ- 

 ment ; but it is an open question whether these slight changes of 

 character can or cannot be transmitted to offspring. The theory 

 of Natural Selection assumes that the transmission of characters 

 is not absolute, but is subject to slight variation"; and inasmuch 

 as the laws which govern these variations are absolutely unknown, 

 the variations are provisionally called fortuitous variations. It 

 further assumes that the possession of even a very slight change 

 of character in a direction favourable to longevity may give to the 

 possessor the opportunity of producing more offspring than "the 

 very slightly less favoured individuals. It still further assumes 



