216 M. Ussow’s Zoologico-Embryological Investigations. 
tremity ; and a lower one, the perfectly straight rudiment, 
closed at the extremity, of the true rectum. The walls of these 
two tubes, as also the cesophagus, consist of a layer of cylin- 
drical cells of the introverted upper lamella, surrounded by 
one or two layers of fusiform cells of the intestino-fibrous 
layer of the middle lamella. 
The further development of the intestinal canal which takes 
place in the third period, consists in the continued growth and 
increase in depth of its parts above mentioned. ‘The stomach 
is formed at first as a dilatation of the hinder part of the 
cesophagus, which, after it has lengthened parallel to the 
dorsal part of the mantle as far as one half the length of the 
latter, bends towards the ventral surface almost at a right angle, 
and unites* with the lengthened primitive rectum, which is 
turned up towards the back. 
At the point where the prolongation of the stomach meets 
the rectum a small dilatation is produced; and from this the 
cecum is afterwards formed. At the close of the first half of 
the third period, in transverse and longitudinal sections of 
the Cephalopoda investigated by me, there are behind the ink- 
sac (which is already considerably developed), and at. first 
nearer the ventral surface of the embryo, two blind, clavate, 
thick-walled tubules, which have been developed from a dila- 
tation of the posterior part of the intestinal canal, and repre- 
sent the rudiment of the liver. It is only in the postembryonic 
period, after the nutritive vitellus is entirely absorbed, that the 
two halves of the liver enlarge very rapidly, approach each 
other, and take up their ordinary place in the dorsal part. 
The proventriculus, or so-called crop, is also developed in 
the embryo of Argonauta in the first half of the third period, 
as a dilatation of the cesophagus situated beneath the cerebral 
ganglion. The walls of all the dilatations above mentioned, 
which originate at different times, are formed from the vari- 
ous main and subordinate parts of the intestinal tract, and con- 
sist of one or two rows of fusiform cells of the intestino-fibrous 
connexion between the ink-sac and the liver, formerly described by Van 
Beneden (Joc. cit. p. 10), has no existence. 
* I have not succeeded in observing the moment of direct union; but 
from the evidence of longitudinal sections of certain stages, and, in fact, 
of embryos in which the long anterior intestine, enlarged at the extre- 
mity, extends to two thirds the height of the mantle (first half of the 
third period), and the rectum curves up towards the dorsal surface, and 
then of certain sections (from the second half of the third period) in 
which the slightly tortuous tractus intestinalis is visible in its whole 
length, I firmly believe that I may assert that this union in reality takes 
place. 
