ART. 5. FOSSIL BIRDS FROM ARIZONA WETMORE. 5 



ranging now in South America, southern North America, southern 

 Africa, India, and Burma^ — that, lacking skeletons of some of the 

 species of the genus, I should have hesitated to describe this fossil 

 bird as new were it not that it represents a species distinctly smaller 

 than any of those known to-day. The dimensions of the fossil hu- 

 merus are less than those of D. arcuata^ that species and D. javamica 

 being the smallest of modern representatives of the group. 



A fossil tree-duck, Dendrocygna validipinnis, has been described 

 by C. W. de Vis ^ from what are said to be " post-pliocene deposits " 

 in Queensland, Australia. This species need not figure in the present 

 comparison as it also is a larger form, compared by the describer Avith 

 the Australian tree-duck, now known as Leptotarsis eytoni. If 

 validipennis is correctly allocated as a tree-duck, as seems from the 

 figures and description to be the case, then it is probable that the 

 skeleton of Leptotarsis offers some difference from that of Dendro- 

 cygna in the form of the internal and external tubercles, the crista 

 superior and the head of the shaft. Material is not at hand to verify 

 this supposition. 



In the diagnosis given for Dendrocygna eversa comparison has 

 been made with D. hicolor. The fossil D. eversa, differs from all 

 the tree-ducks at hand (as listed above) in smaller size, position 

 of deltoid ridge, and angle formed by crista inferior at junction 

 with side of shaft. In the other diagnostic characters assigned 

 eversa agrees with arhorea in sharp ridging of upper end of shaft, 

 and with arcuata and arhorea in the amount of overhang of the hu- 

 meral head. 



Dendrocygna eversa will stand for the present as the smallest 

 of known tree-ducks. 



The humeral head in Dendrocygna resembles that of geese 

 (Branta, Anser, Chen, and Philacte) and differs from ducks {Dafila, 

 Marila, Ensmafura, Histnonicus, Oidemia, etc.) and Mergansers 

 {Mergus and Lophodytes) in that when the bone is viewed from in 

 front no overhang of the caput humeri is evident in the cleft of the 

 incisura capitis, in the compressed ridge at the upper end of the 

 shaft on the posterior side, and in general narrow form of the hu- 

 meral head. Dendrocygna differs from the geese mainly in the 

 relatively lessened bulk of the caput humeri, in the position of the 

 nutrient foramen nearer the middle of length of shaft, and in the 

 sharper angle on the margin of the crista superior, differences more 

 or less intangible, that do not hold definitely for all the genera of 

 geese as a group. Anser albifrons in particular strongly suggests 

 Dendrocygna. 



iProc. Linn. Soc. New South Wales, ser. 2, vol. 3 (for 1888), 1SS9, p. 1282, pi. 34, 

 figs. 5a, 5b, and 6. 



