AMERICAN PALEOZOIC CEPHALOPODS 
207 
River of St. Joseph Island under the name Cyrtoceras huronense. 
However, too little of the living chamber is present to determine 
this with certainty. The siphuncle is in contact with that side 
of the conch which is curved convexly lengthwise. Its segments 
may be subfusiform in outline, but not enough is exposed to give 
a good idea of its structure. The general resemblance of Cyrto- 
ceras alethes to such forms as Cyrtoceras huronense Billings, 
C. isodorus Billings, C. pandion Hall and C. plebium is sufficiently 
great to make it desirable to determine with accuracy whether 
such an early origin of cyrtoceroids of this type actually took 
place. 
The conch of Cyrtoceras metellus Billings is distinctly curved 
lengthwise and the section is stated to be circular or nearly so. 
The siphuncle of the type is said to be unknown. Billings stated 
that this species resembled Cyrtoceras syphax Billings very 
closely ; in that case its siphuncle should be located on that side 
of the conch which is curved concavely lengthwise. However, in 
the collections of the Geological Survey of Canada, Specimen No. 
820a, labelled Cyrtoceras metellus, has the siphuncle in contact 
with the convexly curved side of the conch. The segments are 
strongly nummuloidal, and, owing to the concavity of the septa, 
are strongly oblique to the ventral outline of the conch. More- 
over, on this specimen the septa rise much higher on the con- 
cavely curved side of the conch than on the convex side. 
The conch of Cyrtoceras dictys Billings is distinctly curved 
lengthwise and compressed laterally. The species appears related 
to Specimen 820a, mentioned in the preceding lines, but nothing 
is known of its siphuncle. 
The siphuncle of Cyrtoceras kirhyi Whitfield is in contact with 
the ventral wall of the conch, is 2 mm. in diameter, and its seg- 
ments appear tubular, constricting slightly along the upper part 
of the camerae, widening slightly along their lower part, and 
constricting again where entering the top of the septal neck next 
beneath. Cyrtoceras beekmanense Whitfield appears closely 
related to kirbyi. 
In Cyrtoceras raei Whitfield, the siphuncle is in contact with 
the ventral wall of the conch, is 2 mm. in diameter, and its seg- 
ments appear tubular. Possibly the conch was gyroceran in 
form, the whorls being depressed dorsoventrally. 
In Cyrtoceras vassarina Dwight, the conch is laterally com- 
