466 Edward Phelps AUis jr., 



trary to Parker's statement [57, p. 96]. the name "frontal postérieur" [32, 

 vol. VIIIj pi. 1). In all of these fishes, also, it seems probable that the post- 

 preoperciilar end of the squamosal lodges a postfacial sense organ. A supra- 

 temporal canal is present, in all, and it is enclosed, in all excepting Cottus, 

 in an extrascapular that is represented by either one or several ossicles. 

 In each of the three species of Cottus mentioned, the suprateni- 

 poral canal traverses a part of thé parietal. In both Cottus aeneus 

 and C. octodecimospinosus, there is one extrascapular ossicle that lodges 

 a section of the main infraorbital canal, and a second ossicle which 

 encloses the lateral portion of the supratemporal canal. Mesial to 

 this latter ossicle, the supratemporal canal is enclosed for a short 

 distance in the superficial layers of the parietal, beyond which bone 

 it lies, as a simple dermal canal, on the dorsal surface first of the 

 parietal and then of the supraoccipital, until it joins, in the middle 

 line, its fellow of the opposite side. This median, dermal portion of 

 the commissure lies along the hind edge of the dorsal surface of the 

 skull, in a depressed portion formed by the gouging out, as it were, 

 of the dorso-posterior edge of the skull between the two parietal pro- 

 cesses. The two extrascapular ossicles on each side of Cottus aeneus, 

 which fish I have recently had examined here, each lodge a sense organ, 

 one of which belongs to the main infraorbital line and the other to 

 the supratemporal line. Two other organs are found in the remaining 

 portion of the supratemporal canal but their exact positions could not 

 be definitely determined in the material I had at my disposal. One 

 organ lies quite certainly in that section of canal that is enclosed 

 in the parietal, and the other apparently immediately beyond that 

 enclosed section, on the dorsal surface of the parietal. One extra- 

 scapular ossicle, on each side, has thus here quite certainly fused 

 irrecognizably with the parietal, to form' part of the parietal process, 

 and a second one has perhaps fused with that part of the parietal 

 that lies mesial to the process; but this second ossicle would, in that 

 case, be a concave scale-like or semi-cylindrical ossicle, and not a 

 tubular one. In the accompanying figui-e, fig. 35, the canals are shown 

 in Cottus octodecimospinosus, but the sutures separating the cranial 

 Ijuues are not shown, as they were not given distinctly in my sketches. 



