FOSSIL VARANIDAE AND MEGALANIDAE. 
895 
and eleventh, they gradually disappear being almost totally absent on the 
twentieth. If a generalization of this phenomenon may be allowed and 
thus also applied to our fossil Varanus, the dorsal vertebra in question 
might be determined as belonging to the region in the vicinity of the 
eleventh vertebra, which also from other considerations appears to be the 
most likely supposition. The facies costalis is intact on the left side only, 
it is however considerably worn down and therefore of tolerably smooth 
surface, not exhibiting the three protuberances so striking in V. griseus, 
corresponding to which we find on the outline of the facies costalis but 
three feeble convexities, the parts of which are separated from eachother 
by two horizontal strangulations. In some places on its periphery the ball 
presents trifling .fractures ; it is well developed and broadly rounded; 
corresponding to the articulating cavities (as viewed from above) it is long 
and rather flat ; the centre, before the basal part of the ball, is bilaterally 
slightly hollowed; immediately after this we find on both sides a curved- 
edged extension protruding laterally, forming the limit between the centre 
and its cap-like ball, the latter ending accordingly in two small lateral 
corners or points. Compared to the total height and the volume of the 
vertebra, the spinal canal is narrow, proportionally much smaller than in 
F. griseus. The upper arch is of considerable width, after the proc. obliqui 
although slender, being however less delicate than with F. griseus Daud. 
The proc. obi. anteriores gradually rise in lateral direction ; they are com¬ 
pressed at their base and flatten down, further on following a less vertical 
direction than in F. griseus; ellipsoid-shaped in their superior region, on the 
anterior parts (in the same level with the spinal canal) they exhibit a sub- 
rhomboidal form, the inner side of this anterior part being grooved in the 
vicinity of the spinal canal, and on its upper edge protuberating forwards 
ridgelike. This ridge continues above the spinal canal in the tectiform central 
part of the upper arch. (Sse for these details PI. I, Fig. 12 & Pl. II, Fig. 8). 
The groove-like furrow just mentioned (sulcus connectens mihi) 
leading from the anterior lateral part of the arcus vertebrae (= the anterior 
part of proc. obi. ant. just referred to and designated by me as facies 
subrhomboidales) to the spinal canal, is practically completely 
missing in F. griseus (Pl. II, Fig. 9) since in F. marathonensis the obtuse 
angle above the sulcus connectens — embraced by the anterior border of 
the tectiform front part of arcus vertebrae and the anterior edge of 
proc. obi. ant. — does not exist in F. griseus, the proc. obi. ant. of this 
latter rising more vertically, describing thus an angle of almost 90°, leaving 
hereby no space for an eventual furrow below it. From the phylogene- 
tical point of view this phenomenon might be regarded as a reduction of 
space, the cause of which remains to be established. The edge correspon- 
