OSTEOLOGY OF BIRDS 287 



the 14th is arrived at, where it is quite distinct, becoming more so 

 in the 15th, after which this character is well developed in the dorsal 

 region. As in man)- other ducks, the mid cervical vertebrae are 

 very broad and short, and throughout this region of the spine they 

 are closely locked together, due to their much aborted pre- and 

 postzygapophyses. On the ventral aspect, the carotid canal is 

 never closed in, but on the contrary in the 6th to the 9th it is 

 widely spread open in the transverse direction, and the parial para- 

 pophyses are quite inconspicuous. The lateral vertebral canals are 

 covered for nearly the entire length in the first eight or nine cer- 

 vicals, thus affording very thorough protection to the vessels they 

 pass during life, forming as they do in this region a nearly con- 

 tinuous tube upon either side. In a manner similar to what was 

 found in Clangula, the haemal spines of the dorsal vertebrae are 

 strongly developed, and they are represented by a single median 

 platelet of bone in the 12-15 cervicals, to become trifid on the 16th. 

 In the dorsal region of S . mollissima the vertebrae are all 

 closely locked together, while a network of bony spiculae lash 

 across the superior surfaces of the broad transverse processes, and 

 even pass on to the neighboring surfaces of the bodies of the ribs. 

 This is also the case in P . s t e 1 1 e r i and other ducks. In long 

 necked species, such as Dafila for example, the characters of the 

 vertebrae are materially changed in the matter of form, for in the 

 mid cervical region they become more or less elongated and some- 

 what cylindrical in shape, with the carotid canal narrowed very 

 much, although even here it is open for its entire length. 



Finally, it may be said, that, although as a whole the vertebrae 

 making up the spinal column in the Anatinae agree in their gen- 

 eral characters, there are nevertheless to be found certain minor 

 and unimportant characters in these bones of any species, irre- 

 spective of its genus, which for that species are quite distinctive 

 and constant. 



Further along*, and after the remainder of the skeleton has been 

 examined in detail this matter will be referred to again. 



Turning our attention now to the consideration of the pelvis, 

 we find this compound bone in Spatula presenting upon its 

 dorsal aspect the following points for our examination : The ilio- 

 neural canals are completely closed in by the ilia meeting and 

 anchylosing with the crista of the leading sacral vertebrae. This is 

 the case, I believe, throughout the entire suborder. On either side 

 of this the preacetabular portion of the ilium is longitudinally con- 



