[85] THE OSTEOLOGY OF AMIA CALVA. 831 



neural and haemal arclies remain free througliout life, articulating with 

 them upon certain facets that are overlaid by their cartilage. 



I fail to find a pair of ribs attached to the first free vertebra or 

 what now corresponds to the "atlas." Its neural arch has an inde- 

 pendent spine, articulating- with it, and directed backwards. A simi- 

 lar spine, only longer, is found in a like position on the neural arch of 

 the second and third vertebrje. Three or four others follow in sequence 

 behind these, but they have no apparent connection with the neural 

 arch of the vertebrae. The second vertebra supports a delicate pair of 

 ribs, which articulate directly with the sides of its centrum. In the third 

 segment a small pair of parapophyses have made their appearance, and 

 the ribs of this vertebra articulate with their outer extremities. These 

 parapophyses are characteristic of the vertebrae to the thirty- seventh 

 inclusive. They are always directed downwards and outwards ; are 

 longest in mid series, but as they proceed backwards are situated lower 

 down on the centrum of the vertebra. The ribs are long and slender 

 and become more so as we proceed towards the tail ; in every case they 

 articulate with the extremity of the parapophyses. 



The extremity of the neural spine of the sixth vertebra in Amia is 

 bifurcated, and this feature is present for about two-thirds the way 

 down the column ; these spines being directed upward and backward, 

 with the ones over the middle of the abdominal cavity more decidedly 

 backwards, though the rear spines are the most deeply bifurcated. 

 Twenty of the ultimate ones are simple in their structure. Not very 

 well marked parapophyses are found upon the thirty-eighth vertebra, 

 and this segment is without a pair of ribs. The neural arches inclose 

 quite a capacious neural canal, and their bases articulate between each 

 consecutive pair of vertebrae, these latter having a form to accommo- 

 date themselves to this unique condition (Fig. 26). ^o neural arch 

 is found upon the forty-fifth vertebra, 'and from that onwards they only 

 occur upon the alternate segments. In the thirty-ninth vertebra, what 

 would at first appear to be the parapophyses in the anterior part of the 

 column, are here much larger, freely articulated, and inclose a canal by 

 the union of their extremities beneath, in the medium plane. These 

 also skip the same vertebra as the neural arches do above tTiem on the 

 column 5 fourteen of them also support a free spine from their mid points 

 below. After this they are united and pass round the bent-up verte- 

 bral column, becoming broader and gradually shorter, where they sujj- 

 port the caudal fin rays (Fig. 25). The last six or seven of these 

 hcemal spines appear to be anchylosed with the vertebrae. 



I count in my specimen before me, fifty-three bony rays in the long 

 dorsal fin; these branch above, and the ultimate ones branch a second 

 time. These rays are supported by an equal number of interspinoiis 

 hones, through the intervention of little ossicles that pass obliquely 

 from one to the other (Fig. 25). All this part has been quite correctly 

 figured by Franque, but this author overlooked a series of delicate little 



