OSTEOLOGY OF BIRDS 317 



there seem to be but 10, as is the case in some ducks. Throughout 

 the entire cervicodorsal region of the spine in Branta the neural 

 canal is small and subcylindrical in form, and this holds true for 

 other forms we now have under consideration here. Viewing the 

 thoracic skeleton of Branta canadensis upon direct lateral 

 aspect, it is to be observed that the dorsal vertebrae when thus ar- 

 ticulated /// situ are very closely interlocked indeed ; their 

 neural spines are not lofty and are much elongated, with many long 

 osseous spiculae interlacing among them, as well as strapping to- 

 gether the broad transverse processes. The centra are large and 

 strong, being somewhat transversely compressed, except at their 

 articular ends, which are broad and enlarged. Low, feebly devel- 

 oped haemal spines are present, while the articular facets for the 

 ribs are w^ell concaved and constitute marked features of these 

 vertebrae. Although of some length the first pair of free ribs lack 

 cpiplcural appendages; the second and longer pair of free ribs 

 possess them, though they are small, and as in case of all the rest 

 of the series, firmly coossified with the rib. They are thin and 

 fiattish on all four dorsal ones, and in any case do not quite fully 

 overlap the rib next behind them. The anterior pelvic rib possesses 

 a semiaborted epipleural process also, but all the other pelvic ones 

 lack them. Of these last there are four pairs, and, studied upon 

 this lateral aspect, they are seen to become more and more slender 

 as we pass from before, backward. They are curiously bent hi 

 front of the pelvic acetabulum, indeed the ultimate rib is a mere 

 wirelike rod of bone as it were, that for its vertebral moiety passes 

 very close to the lower border of the pubic element of the pelvis ; 

 is in intimate contact with the propubic process; then goes across 

 as a thread of bone, to merge by its thoracic end with the nether 

 aspect of the ilium near the vertebra to which it belongs. All these 

 ribs, both dorsal and pelvic, articulate below with well developed 

 costal ribs, the last pair of which are " floating " ones, and come 

 nowhere near the sternal border. These sternal ribs, or at least 

 the first five pairs of them, are highly pneumatic, having their distal 

 ends gradually enlarged, and much compressed from side to side, 

 the reverse being the case with their sternal extremities, that 

 is, in so far as their flattening is concerned. Thus it will be seen 

 that Branta canadensis has fi-z'c pairs of pelvic ribs, while 

 I find that Anser albifrons has but four, and. Chen and 

 Dendrocygna still fewer, having each but tliree pairs. 



