332 BULLETIN UNITED STATES GEOLOGICAL SURVEY. \rol.\l. 



assist in iuclosing, and with tliis change the succeeding vertebice have 

 their diapophyses much elevated to meet the internal iliac margins. 

 This section consumes four additional vertebrae, the centra of which go 

 to make up the latter moiety of the cavity for the " ventricular dilatation" 

 of the myelou, and they show the double foramina on either side, one 

 above another, for the separate exit of the motor and sensory roots of 

 the sacral plexus. 



A double row, i. e., one on either side of subcircular vacuities, exists 

 here also among the transverse processes (Plate XIII, Fig. 90, Tefrao 

 canadensis^ S ). It is through this portion of the sacrum that we observe 

 in the chick the greatest amount of tardiness in sealing up of the neural 

 tube above by the superior union of the engaged neurapophyses. 



The remaining eight vertebrae become much compressed with ex- 

 panded processes that rarely allow apertures to remain among them, 

 forming an excellent mid-section to the broad and capacious pelvic 

 cavity, with nearly all signs of its original formation obliterated on the 

 outer and superior aspect. • 



The neural canal is distinctly circular as it enters the sacrum an- 

 teriorly, becoming only slightly flattened as it nears the coccyx. 

 Above we find the neural s]iine confluent with the ilia anteriorly 

 along its summit, and some additional bone deposited posteriorly in the 

 way of their lateral plates, to bridge over the ample "iho-neural" 

 canals. Opposite the " gluteal ridges " the bones are yet firmly knit, 

 but for the remaining i)art of the sacro-iliac suture the interested bones 

 can be said only to snugly meet each other. Cupidonia alone has quite 

 an interspace present (Plate XII, Figs. 83 and 84). 



The sacral wedge is quite thoroughly permeated by air, which enters 

 through foramina in the vertebrae in localities similarly situated to 

 those described in speaking of the anterior part of the column. 



In Centrocerciis sometimes the first sacral vertebra bears a well-de- 

 veloped hypapophysis, and there may even be some evidence of this 

 process on the second segment. The expanded extremities of these 

 median i)rocesses are connected along their inferior margins by a deli- 

 cate ligament of a firmer consistence than that membrane, which fills in 

 the vacuities between the processes to which it is a limiting border. 

 Xow, it depends how far this ligament is conducted backwards as to 

 how many of the anterior sacral vertebrae bear hypapophyses, as from 

 its attachment to the hypapophysis (we have never seen it commence 

 on the cervicals, though the directions assumed by their processes bear 

 it out) of the first dorsal it completes a long shallow arc of an ellipse, 

 in which the lower margins of the hypapophyses are found and assist 

 to complete. This semi-osseous, semi-membranous, attenuated median 

 plate dips down into the thoracic cavity in the living bird, for some 

 little distance, as an interpneumonic septum. 



The number of free caudal vertebrae in the adult Sage Cock is five, 

 and to these is to be added the ^pygostyle. They all have pretty much 



