236 
MR. W. K. PARKER ON THE STRUCTURE AND 
69. Pliryniscus varivs .—Adult female ; If inch long. Costa Rica. 
The last species had the average relation of length to greatest breadth ; in this we 
have, suddenly, one of the longest skulls in the group, the breadth being little more 
than three-quarters of the length ; the general outline is that of a parabola (Plate 41, 
figs. 6-9). In some things the asymmetry is like that of the last kind, in others 
the largest bone of a pair is on the opposite side. The ossification is more intense and 
Salamandrian than in P. cruciger. The skull is still more depressed, and the deve¬ 
lopment of some parts has suffered severer arrests. The fitting on of the head to the 
spine is different, in that there was an almost directly 'posterior aspect of the occipital 
condyles ; here (figs. 6-8, oc.c.) they are postero-irferior; the foramen magnum {fra.) is 
still more oblique (superior in aspect) than in the last kind. Recently metamorphosed 
young of normal Anura show nearly as great an extension of the parotic region 
as this species; for here the horizontal canal runs into the substance of the thick 
unossifiecl “tegmen” (t.ty.). That selvedge, a band above the huge optic fenestra 
(II.), the rim of the duck-billed snout, and the edge of the fenestra ovalis, are 
unossified ; very little else of the cranium proper remains soft. Measured along the 
axis, there is but little difference in the relative extent of the fore, middle, and hind 
skull; the fore skull is somewhat longer than the rest, on account of the projection 
of the snout. 
The semicircular canals project well from the bony hind skull; the supraoccipital 
region is short, the two oval and the large heart-shaped fontanelles (/>., fo'.) are very 
large; there is just an overlapping margin as a rudiment of the tegmen cranii running 
along the sides, and enclosing the fontanelles. The mid skull is wide, making it look 
short; the temporo-postorbital region is very wide. The mid skull bulges in the 
middle (fig. 8), but shrinks in front; the floor of the nasal region is like a highly 
arched “ hard palate ” in a human skull. The optic fenestire (II.) occupy nearly a 
third of the orbital tract, and are almost as large as their counterparts for the setting 
in of the auditory capsules ; the “ serial homology ” of these spaces can be well seen 
in such a skull as this. The girdle-bone extends from this space to nearly the verge 
of the snout, and thus occupies three cranial regions, besides taking in the etlnno- 
palatines {e.pci.), and uniting with the palatine ectostosis {pa.). The pro-rhinals 
(Plate 40, fig. 10, p.rh .) are small, and the fore palate is narrower than in the last. 
There is a narrow tract of cartilage on the edge of the upper surface of the pterygoid 
(fig. 6, pgf and the facet of the pedicle {pd.) and most of the quadrate {q.) are unossified. 
The condyles (q.c.) are reniform, long, and very oblique, the left more than the right. 
The mandibles (fig. 9) show a long condyle also, and very large mento-Meckelians 
{m.mk.) ; the dentary {cl.) is very long, and the articulare {ar.) a mere trough for the 
cartilage {mk.). This is neither annulus tympanicus, nor columella ; the stapes (figs. 7 
and 8, st.) is large, oval, and umbonate. 
The stylo-hyal (figs. 7-9, st.h.) is adherent ; it is a narrow band, very slightly 
