609 
to reach the extreme point of degeneracy, so as to allow the aper- 
ture to become again free and complete on all sides. 
This is, of course, negative evidence and it may be, as in Ammon- 
oids, that the dorsal edge of the mantle never loses in any series 
when restored to freedom the power to resume the shell-secreting 
structures and function on the dorsal side. It can be readily seen 
that as the whorl became gradually loosened from the inner whorl 
the mantle border would extend the secreting furrows inwards from 
both sides, or, more correctly speaking, perhaps, the non-secreting 
edge of the dorsal border would be contracted and finally disappear. 
There is no antecedent improbability that this might not take place 
in any involute nautilian shell at a sufficiently degenerative substage 
of itsontogeny. The remarkable fact, however, remains that it does 
not take place so far as I know, although I have constantly been on 
the watch for some such examples. 
Ammonotdea. 
It is not necessary to give any extended notices of observations 
on special groups in this order. I have already described the 
absence of the impressed zone in the ordinal radicals Bactrites and 
in most of the Nautilinide on pp. 361, 362, 411, 413 and the 
figures and explanations of Pl. ii, and Figs. 40-42, Pl. viii, of 
Mimoceras lituum. The more specialized genera of the Goniati- 
tinge have the impressed zone, but it is strictly a contact furrow and 
appears as shown in figures of Agoniatites fecundus, one of the Nau- 
tilinidze, sometimes very late in the ontogeny. In other still more 
highly specialized species the loose coiling of the young, figured by 
Sandberger in several species of Gephuroceras, Manticoceras /att- 
dorsale of the Devonian and by the author in Glyphtoceras crents- 
tria and atratus of the Carboniferous* indicates, that this zone is 
either absent on the ananepionic dorsum, or, if present, must occur 
as a slight dorsal furrow due to tachygenesis. The larger number 
of the Goniatitinz, as shown by Branco and the author, have, how- 
ever, such closely coiled nepionic stages that, as in all Ceratitine, 
Lytoceratine and Ammonitine, so far as known, the umbilical per- 
foration is closed along the mesal line as shown in Fig. 3, Pl. iii, 
and is represented only by funnel-like lateral prolongations, which 
do not appear to have an open connection with each other. 
* Embryology of Fossil Cephalopods, PI. iii. 
PROC. AMER. PHILOS. SOC. XXXII. 148. 3 Y. PRINTED AvGuST 6, 1894, 
