For a classification to have a scientific value, it is not enough to 

 associate such species as bear only obvious marks of external 

 resemblance; characters must also be found which point. to genetic 

 relationships in remote periods of the earth's history, the presence of 

 which connotes the co-existence of many others. Even Linnaeus fell 

 into errors which all modern science rejects ; the grallatorial pratincole 

 can no longer be regarded as a kind of swallow, nor the accipitrine 

 cariama of South America as allied to the rail-like jacana. 



The most recent and elaborate exposition of the affinities of birds 

 on some such solid foundation as this was given by the late Professor 

 Sundevall in his * Methodi Naturalis avium disponendarum Tentamen' 

 (Stockholm, 1872-3), and it is for this reason that his views, which, in 

 their outcome at least, have received a wide acceptance from modern 

 ornithologists of all countries, have been adopted here. It would be 

 obviously unfair to criticize this or any classification as a whole from 

 its application to any smaller collection of birds than those of the 

 entire globe. Suffice it, however, to say that no one can study 

 Sundevall's observations without materially adding to his store of 

 knowledge. 



When the impossibility of a linear series of any living creatures 

 bearing complex relationships 011 every side is fully recognized, it is 

 clear that the first step in classification is to define certain groups. 

 The old division into Land and Water birds is a familiar instance, a 

 division, however, too vague to be of any use. But if the young of a 

 great variety of birds are observed from their earliest existence, some 

 are seen to begin life in a much less developed state than others. One 

 group, called thence Gymnopaedes, are hatched, in a more or less 

 elaborately-constructed nest, nearly naked, helpless, and utterly 

 unable to seek their own food, or even to support themselves upon 

 their legs ; the other, called Dasypaedes, are hatched, in a nest often 

 of -the simplest form, wholly covered with down, and able to move 

 about and feed themselves from their very first exclusion from the 

 egg. Here is at once the first step towards a scientific dichotomy ; it 

 remains, in the next place, to subdivide these groups into orders. 



Taking now the Gymnopaedes, for instance, on one side can be 

 placed the order Oscines, with a strong hind-toe separately movable, 

 its claw longer than any other, and the greater wing-coverts not 

 reaching to the middle of the secondary quills ; on the other the 

 order Volucres, with a hind-toe weak (or even absent), not separately 

 movable, its claw small, and the greater wing- coverts (except in some 

 woodpeckers) reaching well beyond the middle of the secondaries. 



Then, regarding the Oscines, which order contains about a quarter 

 of all known species of birds, the majority have a double horny entire 

 plate at the back of each tarsus ; while the rest, of which the larks 

 and the hoopoe alone are British, have the same part of the tarsus 

 covered with separate scutella or scales. The possession of nine cubital 

 quills, or of ten, constitutes another important distinction. 



The Volucres, again, are divided according to whether the fourth (or 

 outer) toe can be turned backwards or not. 



The morphological value of such assemblages as these is obvious to 

 the comparative anatomist, when he sees how many other independent 



