OF THE SKULL IN THE AMPHIBIA UEODELA. 203 



gether shorter and has a better finish, with denser and more polished bones ; it evidently 

 belonged to a more aged individual, as it has lost its sphenoidal teeth. 



Tbe occipital condyles (PI. XXI. figs. 6-8, oc.c) are very large and pedunculated ; tbe 

 facets look more downwards than upwards ; tbe space between them is a measure of tbe 

 size of the intercalary (odontoid) vertebra. Tbe occipital arch is, on each side, entirely 

 confluent with tbe auditory mass ; above (fig. G) the arch is finished by suture, below (fig. 7) 

 by a narrow tract of cartilage. 



Tbe auditory masses are much less rugged than in Spelerpes, tbe divisional fossa 

 and tbe swelling arches of the semicircular canals being more bidden by tbe more perfect 

 bony encasement ; and tbe parietals (p) are imbricated on the hind skull more smoothly 

 and with more exquisite finish. 



Tbe cartilaginous tract between the prootics and the sphenethmoids (sp.e) is very small, 

 just a little rim to the bole for tbe trigeminal nerve ; but on the outer region above 

 there is a tract of cartilage where the otic process has coalesced with the capsule 

 (PL XXI. figs. 6, 8 and PI. XVIII. fig. 9, ot.p). 



But below (PL XXI. fig. 7 and PL XVIII. fig. 9) tbe quadrate has run its bony sub- 

 stance into the apex of tbe process ; and between this part tbe horizontal canal (h.sc) is 

 completely encased in bone. The sulcus between the conjoined and the single canals 

 above (fig. 8, a.sc, p.sc, h.sc) is shallow and crescentic ; the squamosal (sq) docs not climb 

 on to this space nearly so much, but is more lateral. 



Below (PL XXI. fig. 7 and PL XVIII. fig. 9) the fenestra ovalis is sublateral, but it 

 is less, as also its enclosed stapes (fs.o, st) ; that plate is roughly pyriform in outline, is 

 unossified at its rim, and has on its broad hinder part a boss for the muscular fibres that 

 are inserted in it. Also on the outer margin in front of the boss there is on the right 

 side a bony columellar handle, which is attached by ligamentous fibres to the under 

 surface of the " tegmen " (PL XXI. fig. 7 and PL XVIII. fig. 11, co, st)*. But on the 

 left side (PL XXI. fig. 7 and PL XVIII. fig. 9) this rod is smaller, and its hinder 

 end is attached by a ligament nearly its own length to the stapes ; the fibrous tissue is 

 continued from its fore end to tbe tegmen and otic process. This is the upper end of 

 the " suspensorio-stapedial ligament," such as is found in other Urodeles. 



This curious onesidedness and irregularity, and the free development of this element 

 in the llenopome, leave no doubt in my mind as to its nature ; the facial nerve (7) 

 passes out beneath the otic process, further forwards, in front of this small columella. 



The sphenethmoids (sp.e) reach from the rim of the foramen ovale to a short distance 

 from the nasal region ; they are fenestrate for the optic nerve (2), and they form a hard, 

 strong wall to this well-built little skull. 



Only the narial rim (e.n) and the etbmo-palatine (e.pa) can be seen without dislocating 

 the bony plates, which are most artistically fitted and compacted together. 



The condyles of the suspensorium are behind the foramina ovalia as much as they are 

 in front of it in Spelerptes ; the quadrate (q) ossifies up to the pointed apex of the otic 

 process (PL XVIII. fig. 9, ot.p), but the front of the suspensorium above is soft. 



From the front a similar, but smaller pterygoid process (e.pg) runs forwards and out- 



* In PL XVIII. both figs. 9 and 11 arc reversed. 



