70 
is uncertain. My study of the animal induces me to classify 
Parmella as a distinct and well defined genus of the 1 lelicariotuiue, 
allied to Parmarion and Parmacoehlea but more closely to 
Cystopelta. The slender foot and the bag-like visceral Jiump 
give it externally a, strong resemblance to the latter. 
The length of the animal (a spirit specimen) from muzzle to 
tail is 14 mm., from muzzle to posterior end of visceral hump 
I (i mm. Visceral hump very large, detached from and overlapping 
in length and breadth the slender foot, above protected by the 
shell, which gives to it a flattened outline, beneath enclosed in a 
thin transparent membrane through which the liver and intestines 
are plainly visible. Mantle reddish-brown dotted with black, 
finely papillate entire, free anterior to the pulmonary orifice, with 
a loose fold on the neck, covering the shell for 4 mm. anteriorly 
and for - mm. on either side \ posteriorly the shell is bare to the 
periphery. A few millemetres below the shell the mantle loses 
its papillate aspect, becomes thinner and about the periphery of 
the visceral mass merges into the thin membrane. The tail is 
furnished with a terminal mucous pore not cleft to the sole and 
apparently overhung by a horn, but the shrunken specimen did 
not permit this detail to be clearly observed. A pedal line runs 
from this pore to the lips Anterior to the pore the tail is 
(apparently) keeled, and then broadens into the usual saddle-like, 
space. The tail and muzzle are reddish-brown, a black line runs 
from beneath the anterior angle of the mantle to the mucous pore 
and another along the pedal Hue. Of the shell I add a rough 
sketch, but 1 leave the description to the abler pen of Mr. Brazier. 
So limp is the fringe of epidermis that I could not extract it 
untorn, and to expand it for drawing I had to float it out in 
water. Afriearwu ater , Austen (L, k 1<\ W. Moll, of India, 
pi. 57, tig. I), possesses a similar fringe. The initial whorls were 
occupied by the testicle. 
To my great regret I destroyed the jaw in dissecting it out, 
but succeeded in preserving a remarkable rad u la. This is longer 
than broad, somewhat cordate in outline, the rows nearly flat 
until on reaching a central crest they curve to meet lit an acute 
angle. The graceful rachidian rather resembles that of DuryeUa 
khdsica, as figured by (Job God win-A listen ; the basal plate is 
hidden by the reflection which arises in an oval, then contracts 
and again expanding presents a tricuspid cutting point. 'I he 
laterals are much curved, armed with a round blunt point and a 
small proximal cusp, they retreat from the rachidian at an angle 
of l.V, and after a series of fifteen, pass through a few transition 
forms into the minute marginals. These have a straight cusp, 
and are so contorted that the base of each is beneath the blade 
of its distal neighbour. I estimated that the odontophore 
contained 145 rows, whose formula was 300.15.1.15,300, making 
