186 MR. E. S. GOODRICH : RESTORATIONS OF 



surface, while further forward there are indications of a Splenial showing 

 on the lower outer surface. Of the existence of this element, however, 

 I could not make certain. 



Special attention was devoted to the course of the lateral-line system. It 

 is not included in the figures of Watson and Day ; but Pander studied it with 

 considerable success. Recently it has again been figured by Collinge (3) 

 without, however, adding much to Pander's results. These authors seem 

 to have indicated in their restorations the distribution of the lateral-line pores 

 rather than of the canal itself. The distribution of the two is by no means 

 always the same, since the pores often stray far from the canal with which 

 they are connected by delicate branches, and may be dotted about somewhat 

 irregularly. The double broken line in my figures indicates the course of 

 the canals only, as they are often beautifully revealed in the fossils. 



The main lateral-line canal passing from the body scales enters the hinder 

 region of the tabulare (often called supratemporal), runs forward through 

 the supratemporal (pterotic), intertemporal, and postfrontal. About the 

 middle of the tabulare it gives off a transverse occipital branch which joins 

 its fellow from the opposite side in the median dermal supraoccipital (post- 

 parietal). That portion of the canal which lies between the origin of the 

 transverse occipital and the origin of the infraorbital branch may be called 

 the temporal canal. It is generally considered to belong to the infraorbital. 

 In the postfrontal the canal branches into an upper supraorbital and a lower 

 infraorbital canal. The former proceeds along the margin of the frontal to 

 the rostral shield, where it describes an elegant curve and appears to end 

 close to the nostril. I could find no anterior commissure; if such exists 

 it must be on the ventral surface of the snout. The hitherto unrecognized 

 junction of the supraorbital canal in the postfrontal bone with the infraorbital 

 canal which runs up through the postorbital from the jugal I have been able 

 to trace quite clearly. In the jugal a horizontal branch is given off from 

 the infraorbital canal ; it. passes backwards across the squamosal, and 

 apparent])' joins the preopercular canal which enters the angular, and 

 proceeds forwards to the front end of the dentary. 



These lateraldine canals are of special importance when comparing the 

 skull of fishes with that of the primitive Tetrapoda. Pollard (11), Baur (2), 

 Allis (1), and especially Moodie (9), have already made use of the canals 

 in Amia and Poli/pterus in comparison with the Stegocephali, where the 

 course of the lateral line is often marked by grooves. But these modern 

 fish are specialized in many respects, and the pattern of Osteolepis agrees 

 more closely with the Stegoeephalian (see Moodie's figures, 9). However, 

 it is not proposed to enter into a detailed discussion of the comparison in this 

 paper, but it may be pointed out that the horizontal squamosal canal 

 mentioned above is characteristic also of the Stegocephali. It is possibly 



