1892.] CLASSIFICATION OF BIRDS. 235 



related to the CoraciidEe, and are (as indicated by the Opisthocomus- 

 Gallidse connexion) the lowest of the three groups of Psittaci, Cora- 

 ciid?e,and Coccyges. Cuculidse, as well as Coraciidse, are zoophagous, 

 chiefly insectivorous. The Striges, as a lateral branch of the lower 

 Coraciine stock, explain the considerable number of characters which 

 connect the Striges with the Coccyges, 28 against 12, and with the 

 Psittaci, 22 against 18. In our hypothetical tree the Psittaci would 

 combine with the Coccyges into one bigger branch — Cuculiformes ; 

 the Psittacine twig to stand between that of the Musophagidse and 

 looking towards the branch of the Strige-s, which again come out of 

 the bigger branch of the Coraciiformes. This big branch and that 

 of the Cuculiformes would ultimately combine into a still bigger 

 branch ; below this bifurcation would come oif Opisthocomus and 

 lower still that of the Gallidge. Thus the Psittaci permit us a glimpse 

 at a large part of the Avine tree, namely at that big branch which 

 downwards points towards the Galliformes and towards the Gallo- 

 Ralline and Rallo-Limicoline region of the tree, while the same 

 branch upwards ends not only in all the so-called Picarias but also 

 in the Pico-Passeres. 



The laborious process exemplified in these comparisons was 

 applied to all the families and was not without results, because certain 

 families were gradually found to assume a central position 

 towards which a number of others gravitated. Thus, for instance, 

 the Coraciidse had to be compared with not less than 10 other 

 families ; the Gallidse with 8, &c. Notoriously difficult forms, as, 

 for instance, Trogons and Colies, naturally caused more trouble 

 than others, since the number of comparisons had to be increased. 



The result of all this is the following classification. In the mode 

 of denomination of the various smaller and larger combinations I 

 have followed Fuerbringer's plan. I am sorry that my classification 

 does not bear greater resemblance to his, but nobody who has really 

 studied Fuerbringer's work will fail to perceive that I stand on my 

 friend's shoulders, or rather on his two fundamental volumes in 

 addition to my own work. 



The subfamilies, which have been mentioned whenever desirable, 

 end in ince. Tlhe families end in idee. 



The Famihes are combined into Suborders, indicated by substantival 

 names. The Suborders are combined into Orders, ending 'u\ formes, 

 with a Latin substantive stem. The Orders could, if necessary, be 

 combined into Phyla, ending in morphce, with a Greek substantive 

 stem ; these would correspond with Fuerbringer's Orders, while 

 my Orders are equivalent to his Suborders. 



The whole of the Class Aves has been divided into 2 Subclasses, 

 to which the names of Arrhornithes and Neornithes have been given, 

 the equivalent names of Saururae and Ornithurae being objectionable, 

 because there is no difference in the skeletal part of the tail of 

 ArchcBopteryx and that of the Ratitse, Crypturi, and Hesperornithidse. 

 The Neornithes are separated into the two Divisions of Ratitse and 



