r'TRODUCTiON. xiii 



As already indicattHl, the ranj^e of variation witliiii this class is extremely 

 limitfd; and if «»nr views respecting,' the taxononiic value of the suh- 

 divisiuns are inHuenced hy this condition of thin<;s, wu are ol)li<j;ed to deny 

 lr» the i^n-oups of living birds the right which has generally been conceded 

 of ranking as orders. 



Tlie ureatest distinctions existing among the living members of the class 

 are exhibited on the one liand ))y tlie Ostriches and Kiwis and the rehited 

 forms, and on the other by all the remaining birds. 



Tliese c(»ntrasted groups have been regarded by Professor Huxley as of 

 ordinal value; but the difUnences are so slight, in ccmiparison with those 

 whieh have received ordinal distincticm in other classes, tliat the expe- 

 (Uency of giving tliem that value is extremely doubtful ; and chey can 

 be coml)ined into one order, which may approj^riately bear the name of 

 Ell rhipidn I'll . 



An ol)jection has been urged to this depreciation ef tlie value of the sub- 

 divisions of the class, on tlie ground that the peculiar adaptatii n for flight, 

 which is the prominent characteristic of birds, is incapable of being combined 

 with a wider range of form. This is, at most, an explanation of the cause 

 of the sliLrht range of variation, and should not therefore affect the ex])osi- 

 tion ot'the A/r/ (thereby admitted) in a classification based on morphological 

 characteristics. JUit it must also be borne in mind that flight is by no 

 means inc<>nn)atible with extreme modifications, not only of the organs of 

 flight, but of other parts, as is well exemplified in the case of bats and the 

 extinct pterodactyls. 



Nor is the class of birds as now limited confined to the single order of 

 Mhich only we have living representatives. In fossil forms we have, if the 

 dilferences assumed l)e confirmed, types of two distinct orders, one being 

 rej)resented by the genus Air]uropteri/.r and another by the genera IcJitlu/oniis 

 and Aimtoraiii of Marsh. The first has been named Snvrura' by Hieckel ; 

 the second Lliflnfornitlnilrs by ^larsh. 



(.V>m})elled thus to (piestion the existence of any groups of ordinal value 

 among recent birds, we ])roceed now to examine the grounds upon whieh natu- 

 ral subdivisions should be based. The prominent features in the classification 

 of the class until recently have been the divisions into groups distinguished 

 i>y their adaptation for different modes of life ; that is, whether aerial or for 

 progression on land, for wading or for swimming ; or, again, into Land and 

 \Vat(;r l>irds. Such groups have a certain value as simidy artificial combi- 

 nations, l>ut we m\ist not be considered as thereby committing ourselves to 

 such a system as a natural one. 



The time has scarcely amved to justify any system of classification 

 hitherto proposed, and we can only have a sure foundation after an exhaust- 

 ive study of the osteology, as well as the neurology and splanchnology, of the 

 various members. Enough, however, has already been done to convince us 

 that the subdivision of the class into Land and Water Birds does not express 



