48 MORGAN. [Vor. IX. 
and 62. The sections are taken from a longitudinal (dorso- 
ventral) series of a young worm somewhat older than Stage 6 
but not so old as Stage 7. Fig.60 is nearer to the side and 
Fig. 62 nearer the middle line. In Fig. 60 the opening of the 
collar-tube into the posterior end of the second body-cavity of 
one side is seen at f. Behind the opening a thickening of the 
wall is seen and this is found in the next section as well (not 
figured). The third section is drawn in Fig. 61 and shows that 
the thickening in the last is a tangential slice of the wall of the 
gill-pouch, and in this third section the gill-pouch opens into the 
ectodermal invagination that was spoken of in earlier stages as 
the invagination of the collar-pores. The position of the collar- 
pores is seen in this section by the line of cells running forward 
from the pit of ectoderm. Lastly in Fig. 62 is drawn the 
seventh section from the last. It shows the first gill-pouch with 
its inturned wall to form the tongue bar of later stages. By 
superposing the series of three sections it will be seen that it is 
the outer lateral portion of the pouch that protrudes towards 
the ectodermal invagination and unites with it to form the first 
gill-slit. We also see that the tongue bar is formed from 
a portion of the upper lateral wall of the endodermal gill- 
pouch. 
The second gill-pouch has not yet reached the ectoderm. 
We find all stages in the closure of the nerve-cord in larvae 
of these intermediate stages. In Fig. 63 is a cross-section, 
taken from the middle of the collar, and shows a later stage in 
the history of the cord. (The young worm had a single pair 
of gill-slits open.) The ectoderm by a process of sliding in 
from the sides has met above the median plate of ectoderm or 
nerve-cord, and some of the cells at the point of contact have 
fused in the middle line. A narrow crevice is left in the 
middle line above the fused ectoderm, and was seen at the 
surface as a longitudinal furrow. A distinct lumen is left in 
the upper portion of the cord. It is narrow dorso-ventrally, 
but wide from side to side. Thzs lumen may be traced through 
the whole length of the cord, its proportions and size varying 
at different levels. The cells at the sides of the nerve-cord 
show a tendency to push in above the lumen of the cord, and 
