1886.] An Interesting Connecting Genus of Chordata. 1031 
must be kept well apart from them on this account. The rela- 
tions of all these forms to known types may be tentatively repre- 
sented in the following scheme: 
Class TUNICATA., 
Order ANTIARCIIA, Vent posterior; ? mouth as well as protostome present. 
Fam. Bothriolepidide. Caudal region absorbed. 
Fam. Prerichthyide. Caudal region present. 
Class AGNATHA. ithout lower jaw or scapular arch. 
s. 
subclass MONORRHINA (Marsipobranchii). A single median nareal orifice. 
Order HyPERroARTI (Myxinidæ). 
Order HYPEROTRETI (Petromyzontidæ). 
Subclass DIPLORRHINA. Two median nareal orifices. 
Fam. Mycteropide. Cephalic and ventral bucklers. 
ass PIS 
| i Subclass 1. HOLOCEPHALI. 
se 3. SELACHII, 
sc 
4. TELEOSTOMI. 
Order PLAacocanorner. The structure of the fins of this order being unknown 
it cannot be referred to either of the three primary divisions (Crossoptery- 
gia, Chrondostei and Actinopteri) with certainty. Supposing it to belong 
to the last named, it agrees best with the Isospondyli, but apparently differs 
in the lack of some of the elements of the suspensorium of the lower jaw. 
There is no sufficient evidence of affinity to the Nematognathi, which is 
probably a modern group. 
The Mycteropide then occupy a position between the Anti- 
archa and Marsipobranchii (Monorrhina) on the one hand and the 
_ fishes on the other. They would, with the latter, enter the 
|“ cladu” Amphirhina of Heckel, if that division be regarded as 
_ defined by the presence of two nareal orifices. But this disposi- 
tion of them would violate truer affinities to the orders without 
lower jaw and scapular arch, for which the term Agnatha 
(Haeckel) may be retained. As compared with Cephalaspidide, 
_- Mycterops approaches nearest to Didymaspis Lankester. The 
: Mycteropide may be regarded as descendants of the Pterichthy- 
didæ, and ancestors of the Placoganoidei. Since the latter occur 
€arlier in geological time (Devonian) than the Mycteropida: (Car- 
boniferous), we may suppose that Mycterops is a descendant of a 
Silurian or Devonian type with a single median nostril, which 
will be a family of Monorrhina. From this hypothetical family 
the Arrhina (Cephalaspidide, etc.) and the Marsipobranchii (lam- 
Phreys) may then be regarded as descendants. The former los 
nares by degeneracy; the latter are degenerate in other respects. 
