xiv INTRODUCTION. 



structural characters are most affected by the nature of the food and the 

 necessity or otherwise to migrate. It will at once be seen that the former 

 set of causes are much more constant than the latter in the Palsearctic 

 Region. There is no reason to suppose that before the existence of man in 

 this region much change took place in the enemies against which birds had 

 to contend ; nor has it ever been suggested that the tastes of female birds 

 are as fickle as those of the females of some of the more highly developed 

 animals of the Palsearctic Region ; whilst, on the other hand, there can be 

 no doubt that both the food and the migrations of birds must have been 

 affected to an enormous extent by the changes of climate consequent on 

 the coming on or passing away of glacial epochs. 



Our ignorance of the comparative value of generic characters appears to 

 me to be absolute ; and, inasmuch as naturalists have agreed that the name 

 of a bird is to be binomial, a combination of the generic and specific names, 

 the wisest course is probably to group species together into convenient 

 genera, which may assist the memory, taking care to satisfy ourselves that 

 the species in each genus are connected together by closer links than those 

 which connect them with species in other genera. The lines which Nature 

 has drawn between different genera are caused by the extinction of inter- 

 mediate species, or by the wideness of the differentiation which has taken 

 place between them, which is generally, though not necessarily, a proof of 

 the length of time which has elapsed since their original separation. All 

 we have to guard against is that the lines which separate our subgenera 

 shall be narrower than those which separate the genus from the nearest 

 allied genera. 



Our next business is to group our genera into families, the largest of 

 which may be conveniently divided into subfamilies. 



So far we shall find it pretty fair sailing in our attempts to classify 

 British birds ; but when we come to group our families into orders, the 

 difference of opinion amongst ornithologists is so great as to the value of 

 characters (which date back to such remote ages) as a sign of relationship 

 or community of origin, that we are entirely at sea, and can only shrink 

 from attempting to decide where doctors disagree. To show the great 

 divergence of opinion amongst ornithologists, it is only necessary to com- 

 pare the various modern attempts at a scientific classification of birds, 

 which will be found to differ from each other in almost every important 

 respect, so that it is obvious that any change in the generally received 

 classification would be at least premature. Most of these classifications 

 are open to the fatal objection that they are attempts to make a linear 

 series, beginning with the most highly specialized birds and ending with 

 the least so ; whereas a true classification must be a chart in which the 

 most highly specialized birds are in the centre, and the least so at the cir- 

 cumference, where they lead on to the forms most nearly allied to birds. 



