Description of Paleeospondylus Gunni. 91 
of doubt. The supposed ring is the result of weathering, 
so that the resemblance between Palwospondylus and 
Petromyzon is not quite so close as the presence of such a 
structure would have led us to suppose. 
Nevertheless, the total absence of any trace of lower jaw, 
together with the entire appearance of the cranium, is highly 
suggestive of the idea that we have here to deal with an 
agnathous suctorial vertebrate, related in some way to the 
modern Marsipobranchii. What of the organs of special sense? 
There can be no reasonable doubt that the lateral parts of 
the hinder moiety of the cranium enclosed the auditory 
organs. In Fig. 1, what I suppose to be the auditory capsules 
are shown broken into at various places by the process of 
weathering. The ventral convexity of the lateral parts of 
the anterior half of the skull, of the part on each side which 
I have compared to the palatine cartilage of the lamprey, 
most probably also corresponds to the position of the eye. 
When these convexities are worn away, as in Figs. 1, 4, 
and 5, what seems to remain of the front part of the 
skull is a comparatively narrow portion traversing antero- 
posteriorly the space enclosed by the now ring-like circum- 
ference, and dividing that space into two smaller and lateral 
spaces, which at first sight one might be tempted to look 
upon as paired nasal chambers. But these seem to me to be 
certainly only the orbits broken into from below, and con- 
sequently I- here agree with Mr Woodward, that there is 
“no certain evidence of nasal capsules.” Consequently it is 
not impossible that the nasal organ was single, as in recent 
Marsipobranchii. 
Immediately behind the head there is an appearance which 
I at first interpreted as a small median shield covering the 
first half dozen vertebre on the dorsal aspect (see Fig. 2; also 
«x in Woodcut), in which interpretation Mr Woodward also 
seems to concur. But since I wrote my previous notice I 
have examined several specimens, for example that repre- 
sented in Fig. 1, in which it is plain that the supposed 
median shield consists of two narrow lateral pieces, which 
may be divaricated at an angle from each other. Ordinarily, 
however, the two pieces lie so closely in apposition to the 
