286 Mr. Toulmin Smith on the Classification 
places. I am unable therefore to describe the outer plaits. The 
characters of the inner ones are_, however, so marked that those 
of the outer ones are quite unnecessary in order to establish the 
specific difference. 
The name retrusus may be considered either to express the ex- 
traordinary degree in which the inner plaits are di’awn back to 
form the projections ; or that the most marked characters of the 
species are hidden from external observation by being on the in- 
ner plaits. In either sense the name seems equally appropriate. 
7. Cephalites catenifer. PI. XIV. figs. 9, 14, 15, 16. 
Plaits broad and deep : outer plaits projecting prominently at 
irregular intervals; projections horse-shoe shaped, with one 
arm of a lower projection often linked to the hoop of the pro- 
jection above it on the same plait ; occasional points of ana- 
stomosis between adjoining plaits : inner plaits having large 
and generally oval depressions at regular intervals ; bulging 
on each side around depressions till adjoining plaits meet and 
open into each other ; processes very eonspicuous : wall very 
thick. 
Var. Annulatus. 
Plaits broad and deep : outer plaits projeeting prominently at irre- 
gular intervals; projeetions ring-shaped, and generally running 
into eaeh other on the same plait and often anastomosing with 
those on adjoining plaits so as to form connected rings over the 
whole surface ; inner plaits having large and generally oval 
depressions at regular intervals ; bulging on each side around 
depressions till adjoining plaits meet and open into eaeh other : 
processes very conspieuous : wall very thick. 
This is a singular species. The speeific name of the typical 
specimens exactly expresses the appearance of the outer surface, 
whieh looks as if several links of a ehain were hung about it, — 
sometimes disconnected, — often connected, — always, or almost 
always, open on one (and generally the same) side. 
This speeies will be readily distinguished from C. hullatus by 
the faet that the semilunar fold is eontinued down to the upper 
edge of the plait, as well as by the links being so often con- 
tinuous, and by each individual projection being much larger. 
Figs. 14, 15 and 16 of PI. XIV., all taken from the same spe- 
cimen, will probably assist in the understanding of this modifica- 
tion of the fold. Pig. 14 shows a part of the core of the matrix, 
that which filled the eentral cavity. The round spots are where de- 
pressions existed in the body itself, and where, consequently, the 
matrix projected outwards from the core. Being broken olF at 
