190 PEOF. W. K. PARKER ON THE MORPHOLOGY 



The supratemporal region of each parietal is hilobate ; it lies over the anterior canal 

 (a.sc) ; behind this expanded part each bone slowly lessens towards its fellow at the 

 mid line ; they leave much of the supraoccipital region naked. 



The squamosals (fig. 2, sq) are mere splints binding the suspensorium, and running 

 along the side of, and somewhat above, the ear-mass. 



On the floor of the skull we see a very long, delicate, inferiorly bulging parasphenoid 

 (fig. 2, pa.s) ; in the space between the fore part of the two vomers it is very splintery ; 

 it gradually narrows to the foramina ovalia, and then spreads into a large ox-face of 

 bone. This basitemporal part has a pair of large rounded lobes, with a lesser pair both 

 in front and behind these : the end is transverse and somewhat emarginate ; it reaches 

 much further back than the parietals. 



The mandibular arch (fig. 3) is a delicate and very elegant bow, much unlike the 

 same arch in the stout Newt just described, and very larval in character. The Meckelian 

 core of cartilage (mk) is unaltered by its investments ; it reaches the chin inside the 

 dentary (d), and its thick hinder part lies in the articularc (ar) as in a trough. Tbat 

 bone is scarcely more than three fifths the length of the mandible, and the splcnial (sp) 

 is very small and lies on the fore end of the articulare. The condyle is oval and large, 

 as in larval Caducibranchs ; it rolls loosely on the slightly scooped condyle of the sus- 

 pensorium. 



As in the Menopome and Siren, the hyobranchial apparatus has undergone but little 

 change ; there has been no absorption of the hinder arches, and their denticulations 

 remain : these things all show that this type is not a true Caducibranch. 



The hyoid and 1st branchial are much alike, but the latter is the longer and slen- 

 derer ; they both remain soft. Their subdivision differs ; for the ccratohyal (c.hy) cuts 

 off a hypohyal (h.hy), here unusually large ; but these two together correspond morpho- 

 logically with the ccratobranchial ; for the epihyal is suppressed. The thick 1st cerato- 

 branchial (e.br) is attached to the middle of the 1st basibranchial, a long oval cartilage 

 (b.br 1 ), and the 2nd ceratobranchial (e.br 2 ) to its end. This last bar is as long as 

 the one in front of it, but is very much slenderer ; it also has acquired an imperfect 

 bony shaft. Its epibranchial piece (e.br 2 ) carries the 3rd and 4th arches, which are 

 composed, as usual, of merely the epibranchial segment (e.br' i & 4 ). The 2nd basi- 

 branchial (b.br 2 ) has begun the usual metamorphosis ; its fore half has become mem- 

 branous, and the hind half is ossified, all but the outer lobes, the only parts that are left 

 in Notophthalmus and Cynops (Pis. XVII., XVIII.) . The presence of three bony 

 centres, and the absorption of the hindermost median bar in front, are the most marked 

 evidence of metamorphoses of these bars. 



Considering our great poverty as to existing Perennibranchiate and Cryptobranchiate 

 Urodeles, this curious sharp-toed Japanese Salamander turns out to be a form of no 

 little importance. It is a very useful stepping-stone from the lower to the higher kinds 

 of its own " Order." 



