130 
ITATJTILOIDEA. 
undulating. The neck ^ of each septum (as seen in section) extends 
backwards and partly overlaps the preceding one, thus forming a 
complete siphuncular tube (fig. 11, n). Siphuncle marginal, or sub- 
marginal, cylindrical, very large, sometimes attaining a diameter of 
one half of that of the shell ; commencing with a conical chamber-. 
The surface of casts of the siphuncle is encircled by a series of oblique 
shallow grooves, alternating with annulations, caused by the over- 
^ Hyatt styles the necks “ funnels ; ” but I think the term is objectionable, 
because it is used by some authors to designate another organ in the living 
Cephalopods (Division Siphonopoda). The French call the necks “ goulots,” 
the Germans “ Siphonalduten,’' 
^ This very remarkable structure has been observed in a species of 
Endoceras, minutely described and figured by Dr. Gerhard Holm, of Upsala, in 
a paper entitled “ Ueber die innere Organisation einiger silurischer Cephalopoden” 
(Dames & Kay ser’s ‘ Palseontologische Abhandlungen,’ Band iii. Heft i. 1886, 
p. l)i I reproduce here reduced outlines of some of his figures ; — 
Fig. 12. 
Endoceras belemnitiforme, Holm. — a, cigar-shaped extremity of the shell covered 
with the test : h, cast of conical initial part of the shell with two of the 
septal chambers, denuded of the test : c, longitudinal section of the initial 
part of the shell, showing conical chamber, cc, at the commencement 
of the siphuncle, with some of the septal chambers, cc, cd, calcareous 
“organic deposit” (organische Ablagerung) resembling that which is met 
with in the terminal septum of Orthoceras truncatum (ante, p. 24) : d, 
fragment of cast of a detached siphuncle, the rings being produced by the 
overlapping of the septal necks : e, side view of the same. 
Dr. Holm assumes that the large conical chamber, cc, is the “ initial chamber,” 
and that we have here therefore the largest known in any Tetrabranch ; its 
size is certainly quite anomalous. Dr. Holm remarks, however, that he has in 
many cases observed the initial chamber to be much smaller than it is in 
