450 CHARACTERS OF PALEOZOIC FISHES COPE. 



stated. Its presence is iudieated by tlie longitudinal transversely 

 concave inferior surface of the element called jugal below. This 

 articular surface might have supported some form of tooth, but as no 

 such have been found associated with the rather abundant remains 

 of Macropetalichthys, it is more probable that a distinct element was 

 attached to this surface. 



As is well known, the superior surface of the head-shield is divided 

 into symmetrical tracts by well-marked lines. These areas have been 

 regarded as the osseous cranial elements, and have been named by 

 Newberry in correspondence with those of higher vertebrata.* The 

 lines referred to, however, are not sutures, but tubes which belong to 

 the lateral line system; and they traverse the centers of the true bony 

 elements instead of bounding them. They join at the centers of some 

 of the elements, and in such cases mark the points of origin of the 

 osseous radii, whose direction they follow. The direction of these 

 tubes is as follows in the present species, and approximately in all the 

 other members of the genus: In the first place there is a frontal lyra, 

 whose branches are parallel for a distance in front of the orbits (as far 

 as the specimen is preserved), and which begin to converge at a point 

 a little in front of the anterior border of the orbit. They join on the 

 middle line about half an orbit's diameter behind the line connecting 

 the posterior borders of the same. From this point they diverge at an 

 angle a little greater than 90 degrees to a [)oint immediately behind 

 the superior border of the orbit, and nearly two orbits' diameter pos- 

 terior to the latter. From this point two lines diverge, one toward the 

 externo-posterior angle of the skull, the other downwards and forwards 

 at an angle a little over 00 degrees from the other branch. The lines 

 are all perfectly straight except those of the lyra, which are bent just 

 in front of the anterior border of the orbits. That these lines represent 

 tubes is readily seen where they are broken across. That of the lyra 

 has a subtriangular section. Below it, in front of the orbit, is a smaller 

 one of round section which the fracture of one side enables me to trace 

 as far as opposite the anterior border of the orbit. 



In their distribution these tubes do not nearly resemble those of 

 Homosreus as represented by Traquair. t A closer resemblance can be 

 traced to those of Coccosteus|, of Dinichthys, and especially to those 

 of Titanichtliys.§ The lateral branches of tlie frontal lyra unite posteri- 

 orly at an angle in Dinichthys terrellii, are slightly separated by a trans- 

 verse tube in Titanichthys agassizii, and are more widely separated in 

 Goccostcus decipiens. In all three, divergent branches extend posteri- 

 orly, as in Macropetalichthys. In the three forms mentioned, these 

 posterior branches send, anteriorly and exteriorly, a branch from a 

 point close to the posterior border of the skull, on each side. This mar- 



• Tho Paleozoic Fishes of North America, 1890, p. 43. 



\ Geological Magazine, 1881>, p. 1, pi. i. 



I Tra<|nair, loc. cit. 



^ Newberry, I. c, i)l.s. i and iii. 



