4 1 2 Bulletin American Museum of Natural History. [Vol. XIX; 



lateral spines I have been unable to determine with any cer- 

 tainty; but they are possibly distinct bones. 



Dorsal and anal fins are present, but their rays are dis- 

 turbed. 



What are probably vomerine teeth are presented. There 

 are three rows on one side of the midline and one row on the 

 other, and there were not less than 5 rows. Those of one of 

 the rows farther from the midline are compressed laterally 

 and each forms a longitudinal cutting edge. The other teeth 

 have a part of their triturating surface mammillated and the 

 borders finely crenulated. 



No. 45 1 6^ (3703) is a fish which has been spread out either 

 by crushing or by inflation by gases during decomposition. 

 The space occupied by the notochord is enlarged so that the 

 bases of the neural arches and those of the ribs are from 6 to 

 II mm. apart. This specimen shows that both the neural 

 arches and the hsemal arches had their proximal ends ex- 

 panded against the notochord. The expansions of the neural 

 arches join and form a continuous covering for the upper side 

 of the notochord. The ends of the haemal arches probably did 

 not come into contact with each other or with the neural arches. 



Beneath the dorsal rays I count 10 interneural supports. 

 There appear to have been 8 anal rays. There are traces of 

 both pectoral and ventral fins. In one gill chamber are seen 

 the impressions of four series of gill filaments. 



No. 4516a (3666) shows the fish as seen from the side, and 

 thus gives us an idea of the elevation of the head and body. 

 The total length is 1 20 mm. The height of the body at the 

 pectoral spine is 30 mm. Only a faint impression of the 

 occipital spine remains. The rays of the dorsal fin cannot 

 be counted. The lower rays of the caudal are the longest and 

 the fin ends rather bluntly. The anal appears to comprise 

 8 rays. The ventrals are distinctly displayed. There is a 

 continuous line of bones along the upper side of the noto- 

 chordal region, the bases of the neural arches. Eleven neu- 

 ral arches are counted from the middle of the back to the 

 base of the caudal fin. Vomerine and splenial teeth are seen, 

 but the number of rows cannot be determined. 



