The Zoologist— February, 1807. 585 



Willnghbifs, 1678.— Although by no means complete, yet unques- 

 tionably the best that had hitherto been promulgated, and meritorious 

 from the fact of its being framed nearly two centuries ago, with no 

 other assistance than the meagre compilations of preceding authors. 

 It is also without doubt the basis on which the ornithological classifi- 

 cation of Linnaeus was founded. Nevertheless, it is confused, obscure 

 and full of errors, especially as regards the water birds. The 

 distinction, however, between the land and water birds is well 

 preserved. 



Brisson's, 1760.— Unnecessarily long, no less than twenty-six 

 orders and one hundred and fifteen genera, and too fantastical in the 

 characteristics of the orders. Moreover, birds belonging, naturally, 

 to the same genus have been placed not only in different genera, but 

 even in different orders. 



Linn<BUs\ 1766.— Violation of natural order in division of land or 

 perching birds into two orders placed wide apart, and the introduction 

 of aquatic orders between. In this respect inferior to Willughby. 

 Nevertheless the best artificial arrangement. 



Latham's, 1790.— Commendable for its brevity, nine orders, but 

 containing what would now be considered mistakes. For instance, 

 Recurvirostra placed in order Palmipedes, instead of Grallatores : 

 Glareola, although placed by some (and amongst others Latham) in 

 order Grallatores, and by others in order Palmipedes, should, I take it, 

 be found in neither, for having some characteristics of each, as well as 

 others peculiar to itself; it should more correctly hold a separate and 

 intermediate rank. 



LacepUds, 179S/v"; Longer even than Brisson's, having no less than 

 thirty-nine orders aatfone hundred and thirty genera. Very confused, 

 too, "and difficult _fc remember, and this is an essential point. The 

 characteristics of the various orders appear too fantastical, and several 

 orders might very well be comprised in one, for in the chief features 

 they agree, but apparently because they do not agree in one or two 

 trifling details Lacepede has separated them. 



DumerWs, 1806.— Somewhat similar to the more recent classification 

 of Yarrell, but differing in several respects and inconsistent in many 

 others. Yarrell has five orders, Dumeril six, the additional order 

 being that of " Grimpeurs " (Scansores), from which, however, he 

 omits Certhia, and places it not only in a different order, but in the 

 same family as Alcedo ! He omits Sylvia, Parus, Motacilla and 



SECOND SERIES — VOL. II. l 



