VOl 'i9S XI1 ] von Ihering, The Dendrocolaptidce. 151 



genera a profound median sulcus is developed. The configuration 

 of the skull depends in a great measure upon the breadth of the 

 interorbital part of the frontal bone and the proportion of this 

 to the greatest breadth of the skull (considered as 100) varies from 

 16 to 50, the absolute measure being in Synallaxis spixi 2.4: 14.3 

 mm. and in Campylorhamphus trochilirostris 8.8: 17.2 mm. As 

 already suggested by Furbringer the study of the variations in 

 the nostrils of the Dendrocolaptidce has shown that this is a 

 character of secondary value. 



The importance which is given in ornithological literature to 

 such terms as holorhinal and schizognathous represents an inherit- 

 ance from the past century. When Huxley in 1867 published his 

 classic treatise on the classification of birds it seemed as if the skull 

 was to attain the same importance in the classification of birds as 

 it had already reached in the mammalian system. 



Six years later Garrod gave to the structure of the nostrils the 

 same importance in avian classification as Huxley had given to the 

 palate structure. And now we ask what is the situation to-day? 



The results set forth in this paper with reference to the schizo- 

 rhiny of the Dendrocolaptidce confirm the opinion of Furbringer 

 as stated above; who also (I. c. p. 1034) rejects Huxley's groups 

 based on palate structure. Beddard (1. c. p. 140) also points out 

 that the maxillo-palatine classification is not really satisfactory 

 from a systematic point of view and adds that it is rendered 

 harmless by the fact that the groups are really not as hard and 

 fast as might be supposed from text books in general. 



In this, however, I cannot agree with Beddard as generalizations 

 of this sort, rejected by the most competent morphologists, often 

 persist with tenacity in our systematic literature and in many 

 instances hinder the zoologist from following his own inclination. 

 If in studying any family in the zoological system we take one 

 anatomical character as a basis for the arrangement of the genera 

 or species we construct a system which is entirely changed if we 

 make use of some other character. Skull or pel vis,\ sternum or 

 syrinx, pterylosis or muscles — - in nearly every case we obtain 

 a different arrangement. 



The result of the exclusive application of certain anatomical 

 characters is seen in Garrod's classification of the Psittaci, which 

 has been accepted by Beddard, in which the South i\.merican 



