INTRODUCTION. 



The number of specimens of FringillidcB recorded in the present 

 volume is 9443, exceeding by several thousands the contents of any 

 previous volume of the ' Catalogue of Birds.' The number of species 

 recorded is 559, of which only 30 are unrepresented in tbe Museum 

 collection, which contains likewise the types of no less than 125 

 species. 



The line of demarcation between the families FringillidcB and 

 Tanagridce seems to be au extremely arbitrary one, and many genera 

 included by me as Finches are just as likely to be Tanagers, if there 

 is really a definable character for the separation of the two families. 

 I allude more especially to such forms as Piezorhina, Ohamceospiza, 

 Coryphospingus, &c. Some distinctive cbaracters may ultimately 

 be discovered in the anatomy of the FringillidcB and Tanagridce 

 which will serve to separate them ; but at preseut the whole classi- 

 fication of these birds is highly unsatisfactory. 



Equally difficult is it to arrive at a satisfactory conclusion with 

 regard to the value of the various subspecies and varietal forms 

 found in North America. It is probable that these can only be 

 properly studied in American museums, although, thanks to the 

 liberal donations of the United States National Museum, Mr. G. B. 

 Sennett, Mr. H. K. Coale, and other correspondents, I have had a 

 better series of North-American birds at my disposal than has ever 

 been my lot before. Some of the races recognizer! by the American 

 Ornithologists' Union in their authoritative ' Check-list ' I have 

 been able clearly to recognize. In other cases the series of skins has 

 been sufficient to show an absolute connection between several forms 

 of Nearctic Finches : whilst some of the differences put forward by 

 American ornithologists appear to me to be insufficient to recognize 

 many of their subspecies. When the British Museum possesses a 



