
544 THE AMERICAN NATURALIST. | [Vor. XXXVI. 
which there appeared to be a direct connection between the 
two cavities which could not be due to poor preservation of 
the material (Fig. 18). It would seem, therefore, that in some 
cases the archenteron is extended at the expense of the seg- 
mentation cavity, as supposed to be the case by Kupffer (12) 
and by Marshall. 
A series of frontal sections through the embryo at about 
the stage of Fig. 7 will show the various steps in the forma- 












uo 
"uma. 
GU 
IOCIS 
e Cen X 3 
EN 
SAE E OR 
eo SY SP 
; CO 
SING 


‘=, 
A 
CES 
aee 






LOK GELS a 
E.A 





A 2E À 
ex) 
Y 
r5 


- 
LACAN a M ALIM A 
I RD ER 
OS 
Os 

neon 
SONES 
SITAS 
.0800: 
bi 

pee 
(s) 
aa 
v. 
que 
A 
TA 
ü 



at 
We) 




e 
iie AM 
See ec Or? 

Fic. 19. — Part of a frontal section through the middle region of an egg at the stage of Fig. 7; 
4 H CS Ss HN IS 1 A. n Ttc" 
n et: the yolk. Lettering as in previous 
figures. 

tion of the notochord. A section just in front of the dorsal 
lip of the blastopore (Fig. 20) shows that the mesoderm in 
this region forms a continuous sheath over the dorsal wall 
of the archenteron and that it is composed of three to four 
layers of small cells (Fig. 20, 77). About eight sections 
anterior to Fig. 20 there is a noticeable thickening of the 
mesoderm in the mid-dorsal region (Fig. 21), while on either 
side of this thickening the mesoderm usually consists of but 
two layers of cells. More anterior, the thickened part of the 
mesoderm layer is completely separated from the lateral sheets 

