844 



THE AMERICAX XATirRALIST. [Vol. XXXVIII. 



of classification, and this is rendered vastly more useful from the 

 fact that many of the characters of the suborders are con- 

 veniently footnoted in the same arrangement. He also em- 

 ployed the ' circle system ' in several plates in his memoir, as 

 well as a modification of the ' radiating line system,' and both 

 with more or less effect (Plates IX-XII). This plan of pre- 

 senting group characters is excellent in many particulars, and in 

 connection with a modified linear scheme it has been employed 

 by Cope in his Classification of Aves {Amer. Nat., Vol. XXIII, 

 No. 274, Oct., 1889, p. 869), and also by Hans Gaclow {P. Z. S., 

 1889). Still earlier than these it was used by Garrod {P. Z. S., 

 1874). 



Turning to the plan of the ' Phylogenetic Tree,' it has, 

 among other authors, notably been employed by Reichenow 

 {Vdgel.der Zoologichen Garten, 1882) and Max Furbringer 

 ('* Untersuchungen zur Morphologie und Systematik der Vogel " : 

 ' Bijdragen tot de Dierkunde; Amsterdam, 1888. Tafln XXVI, 

 XXVII). Reichenow's "Tree " is a very crude representation, 

 and needs no special description in this place. It has, however, 

 an historical interest for us, and a good copy of it may be found 

 in Dr. Sharpe's ' Review' on page 23. A far more elaborate 

 and widely known representation is Furbringer's ' Phylogenetic 

 Tree of Birds.' This brings up the main stem from the rep- 

 tilian root-stock, from which, as it ascends, are thrown off the 

 now dead limbs of a number of more or less known or even 

 hypothetical groups of avireptilian or reptilioavian forms. Far- 

 ther up, more extensive branches carry out the evolution of the 

 main ostrich types, and soon above these the generous and com- 

 plete development of the tree,— its many more or less closely 

 divitling and branching limbs, boughs and stems, showing all the 

 mam modern or existing groups as they are supposed to have 

 been -ivcn off from each other, or else to have arisen from the 

 main trunk. This tree is horizontally divided by imaginary 

 planes at certain points, dividing it into upper, lower and middle 

 ^eetl..ns or hon/.ons. X'iewing the projections of these ver- 

 tually we have tlie circular sections of the various branchings 

 of the tree before us, and these will again give the relations to 

 each other (.t the various groups. Three such projections are 



