366 The Ohio Naturalist. [Vol. VIII, No. 8, 



The relations of the dorsal to the dorso-laterals is shown in 

 both restorations. In the specimen the postero-dorso-lateral 

 and antero-dorso-lateral are firmly united in their original rela- 

 tion. Some fragments are missing from the upper margin and 

 are restored in outline in the drawings. The right margin of the 

 dorso-median is missing in the specimen but its position and shape 

 are almost perfectly indicated on the dorso-laterals. Outside of 

 the dotted line in figure 2 the antero-dorso-lateral is restored. 

 As the left postero-dorso-lateral is crushed down on the right 

 dorso-laterals their relation to one another is obscured. The 

 writer is not satisfied with the outline of the antero-dorso-lateral 

 as it is shown in the drawing, but the bones are undoubtedly in 

 their natural association and the antero-dorso-lateral is the only 

 bone to fill the space in front of the posterior bone. The over- 

 lapping part of the antero-dorso-lateral is very thin. 



The relation of the antero-dorso-lateral to the skull is deter- 

 mined beyond controversy by the specimen under discussion and 

 by two other specimens in the Museum of Oberlin College. The 

 relation is determined by placing the dorso-median and dorso- 

 laterals in association and placing the median line of the dorso- 

 median in the median line of the skull and the articulating part 

 of the antero-dorso-lateral in its socket. The anterior edge out- 

 side of the articulation overlaps the depressed edge of the poster- 

 ior part of the skull for about two centimeters. A specimen of 

 Dinichthys intermedins shows the same relation, and in a speci- 

 men of Dinichthys recently collected from the Huron shale in 

 which both antero-dorso-laterals are preserved in their natural 

 relation to the skull the same relation is shown. 



The position of the clavicular and its relations to other bones 

 is definitely shown and is represented in both restorations. 

 The main articulation is with the depression in the antero-dorso- 

 lateral. Its anterior edge overlaps for more than two centimeters 

 the depressed posterior edge of the skull though it does not arti- 

 culate with the skull. The posterior end of the suborbital rests 

 against a large part of the anterior edge of the clavicular between 

 its two anterior projections. The outer of the anterior projections 

 which just reaches the lower edge of the suborbital probably 

 supported a lateral appendage. It diverges from the inner projec- 

 tion at an angle of about forty degrees but soon curves inward and 

 runs nearly parallel with the inner part. The distal end of this 

 projection is a separate bone. It is sometimes ankylosed with 

 the rest of the bone but is detached in many cases. It should be 

 classed as a distinct skeletal element. If it is homologous with 

 any bone of other vertebrates the writer is in doubt about the 

 homology. The inner anterior projection of the clavicular ex- 

 tends further forward than the outer, and the anterior end sup- 

 ports the mandible. In the restorations the inner part is dis- 



