320 



NEW YORK STATE MUSEUM 



one is arrived at, and it has the aborted character of a caudal 

 vertebra. These last are thoroughly nonpneumatic in character, 

 with thick, stumpy neural spines and an almost complete absence of 

 haemal ones. On the first caudal vertebra the lateral apophyses are 

 short and thick ; to be somewhat better developed on the second one ; 

 best on the third; to slightly abort again on the fourth and fifth; 

 while the last one, or the sixth, lacks them entirely. I find the 

 pygostyle of the wild goose to be of good size, it being a thin, sub- 

 oblong plate of bone, that is thickened and longitudinally grooved 

 along its ventral margin, and sharpened along its dorsal edge. 



Doubtless, at least three or four, or maybe more, tail vertebrae 

 enter into its composition. The last two caudals of Anser a 1 - 

 b i f r o n s possess bifurcated haemal processes ; and the anteroin- 

 ferior angle of the pygostyle in both Chen and Anser is similarly 

 bifid. In the former genus the ilioischiac notch is sharp and tri- 

 angular, and its apex may be bridged over so as to create a minute 

 foramen in advance of it. Anser has this notch large, and nearly 

 as wide as long, and in this bird the ilia extend somewhat further 

 back than the ischia, the very reverse of what we find in the tree 

 ducks. Turning to the pelvis of these last named birds, and 

 regarding the bone of the specimen of Dendrocygna 

 autumnalis at hand, one is at first struck with its narrowness 

 throughout; with the unusual depth of its lateral, postacetabular 

 walls ; and with, as it were, the general drooping of its hinder 

 moiety. All this is quite unlike what we find in the pelves of 

 ordinary geese, or even in any of the ordinary ducks, as, for ex- 

 ample, those of the genus Anas. Agreeing with Chen, Anser and 

 Branta in the forepart of its preacetabular region, the first difference 

 is seen in the slight perviousness of the ilioneural canals behind. A 

 propubic spine is present as in the geese, but the obturator foramen, 

 which is here of very small size, tends to merge completely into the 

 obturator space [see fig. 40]. An ischiadic foramen is com- 

 paratively much smaller than in the true Anserinae, and the distal 

 ends of the long, slender pubic rods are not at all paddle-shaped 

 as in Rranta and the smaller geese. Either ischium has its pos- 

 terior third much expanded, and its distinctly rounded hinder 

 margin projects distally beyond the truncate border of the ilium 

 above it. Agreeing with all true representatives of this suborder, 

 the mesial surface of either ischium is longitudinally marked from 

 one end to the other by a raised and rounded welt of bone, that 

 lends great strength to this part of the pelvis. This ridge merges 



