OF THE SKULL IN THE AMPHIBIA UEODELA. 191 



Species 6 Sr 7. — The Skulls (larval and adult) of Spelerpes rubra, and larval skulls of 



Spelerpes salmonea. 



These North- American Caducibranchs measured as follows : — The adult S. rubra 4^ 

 inches long, and the young (with gills) 1 inch 10 lines ; the larval specimens of S. salmonea 

 measured 2 inches 1 line and 3f inches. The larva3 of this genus reminded Prof. Mivart 

 (' On the Common Prog,' p. 48) of Menobranchus ; they struck me as being like Proteus. 

 We were both right ; for they strongly resemble and very accurately represent both those 

 Perennibranchs. The youngest of these larvse (S. rubra) was of considerable length, 

 namely If inch, and yet the structure of its skull is extremely simple, and differs in 

 nothing essential from that of the adult Menobranchus lateralis (compare PI. XX. 

 figs. 1 & 2 with that type : see Huxley, P. Z. S. 1874, pis. xxix. & xxx.). 



This larval skull is curiously illustrative of the extreme variability of the time in 

 which ossification is set up, and harmonizes with what I find in the Batrachia, namely, 

 that of the most generalized kinds, some skulls are most feebly ossified, and others 

 are a mass of dense bone, with scarcely a suture left. 



Here there is no distinction even between the prootic and exoccipital ; all the occipito- 

 otic region on each side is one undistinguishable tract of bone (PI. XX. figs. 1 & 2, pro, 

 e.o) ; and these bony territories have met in the supraoccipital region, although they 

 have a broad synchondrosis below. There is a character here which strikes the eye at 

 once, namely, that the foramen magnum is nearly as wide as the average width of the 

 cranial cavity. This is surely a low character. 



The condyles (pec) look obliquely upwards, like ordinary " pre-zygapophyses," and 

 between them there is a very wide crescentic notch for the " odontoid vertebra." There 

 is a rudiment of a bony roof to the skull running forwards from the supraocciptal region. 

 This bony tract first grows inwards as a lobe, and then stops short over the Gasserian 

 ganglion ; thence the skull-wall is unossified, and only behind does it lean over the 

 brain. 



The whole configuration of the semicircular canals (fig. 1, a.sc, p.sc) is stamped upon 

 the bony periotic. These structures are very large. 



It will be seen that above, in front of the two ampullae, there is a tract of cartilage ; 

 this is an upgrowth from the basal plate (trabeculo-parachordal), and is normally 

 related to the pier of the mandibular arch, the " suspensorium." On the lower face 

 of the ear-capsule there is some cartilage left ; this is a trilobate tract beneath the 

 " sacculus ;" outside this tract is the large fenestra ovalis (fs.o), with its correspondingly 

 large stapes (st). 



A considerable blood-vessel is seen entering into (? emerging from) the membranous 

 labyrinth at the inner edge of the regularly ovoidal stapes, whose narrow end looks 

 forwards and outwards. 



There is a short canal for the facial nerve (fig. 2. 7), but it soon escapes from the 

 capsule, and runs outwards and forwards beneath the apex of the otic process (ot.p), some 

 distance in front of the stapes. 



The 9th and 10th nerves (9, 10) are seen escaping from the skull through their twin- 

 passage, between the condyles and the ampulla of the posterior canal (p.sc). 



26* 



