
INTRODUCTION. 
TuE following is an attempt to classify existing Birds in groups which are capable 
of being diagnosed. I have divided the Class Aves into six Subclasses, fourteen 
Orders, and thirty-six Suborders. These Subclasses, Orders, and Suborders are 
diagnosed by a series of characters each of which is believed to be found in every 
species contained in the group diagnosed, and the combination of which is believed 
not to occur in any species outside that group. In order to show that each 
combination is diagnostic, a list of the thirty-six suborders is attached to each 
diagnosis, and a star is placed under each suborder in which every species it 
contains is excluded by the character referred to by a letter at the side of the 
column in which the star is placed. In most cases a few additional characters are 
given to strengthen the diagnosis. In selecting characters I have been careful to 
choose those which are supposed to denote affinity—in other words, to be due to 
inheritance from common ancestors; and as far as possible to avoid such as only 
denote analogy, that is to say such as are only instances where like causes have 
produced like effects. It is, however, very difficult to tell which characters have 
been inherited and which have been independently acquired. In many cases it is 
only possible to guess, in others it is absolutely impossible to form any opinion at all. 
It may perhaps be fairly assumed that the more complicated a character is, the less 
chance is there that it can have been independently acquired by two groups. It 
may also be taken for granted that it is very much more difficult even for a simple 
character to be independently acquired than to be independently lost. 
It must be remembered that, so far as the individual is concerned, every bird 
inherited all its characters, whether they may be adapted to its present needs or 
not, but that all characters were originally acquired either by the species or its 
ancestors because they were originally adapted to its needs. 
