488 Edward Phelps Allis jr., 



can be said is, that the main canal probably traverses a lateral extra- 

 scapular, two squamosals and a postfrontal; and the supratemporal 

 canal a lateral and a mesial extrascapular on either side, and perhaps 

 also the so-called medial occipital bone, for a few pores are shown in 

 this latter bone. Bone No. 4 of Pander's descriptions is considered 

 by that author as the frontal, but it can not alone represent that bone, 

 for it is not traversed by the supraorbital canal. Certain of his bones 

 7 and 8 must accordingly be included in the frontal. 



In the Pteraspidae, according to Woodward [80], a well deve- 

 loped system of lateral canals existed, there being, according to that 

 author, in both Pteraspis and Holaspis, four so-called longitudinal 

 canals and a number of connecting canals. One of the longitudinal 

 canals, on each side, lies near the lateral edge of the median plate of 

 the cephalic buckler, and the other near the mid-dorsal line of the same 

 plate. Certain of the connecting canals connect the two mesial longi- 

 tudinal canals with each other, the others connecting those canals with 

 the marginal canals. All of these canals communicate with the outer 

 surface by short diverticula, "alternately right and left" along the 

 canal, each diverticulum opening onto the outer surface by a single 

 pore, these pores thus forming, along each canal, a double series of 

 openings on the external surface of the buckler. The canals have, in 

 their general course, a certain resemblance to the canals in the Batoidei, 

 but there are not sufficient data to warrant a comparison. 



In Acanthodes Bronni, there are in all probability, according to 

 Reis [62, p. 195], three longitudinal latero -sensory canals on the top 

 of the head, one of them being median in position, and also a supra- 

 temporal and a suborbital canal. Each of these canals is said to lie 

 in a bony half-canal on the inner and not on the outer surface of a 

 line of modified scales. In the closely related Traquaria but one canal 

 is shown by Fritsch [38] on either side of the head, and it is appa- 

 rently the main infraorbital. It apparently lies on the outer and not 

 on the inner surface of elongated semi-cylindrical scales. Nothing re- 

 sembling this arrangement of canals is known in living fishes. 



In Chondrosteus acipenseroides the supraclavicular is said by 

 Traquair [74, p. 256] to have its "upper extremity obliquely perforated 



