BLASTOPORE, MESODERM AND METAMERIC SEGMENTATION. 19 
this deepening continues after the middle part is obliterated 
(figs. 4, 6, 11, and 12). The deepening of the groove soon 
forms a very definite pit (g). This pit, when by the growth 
of the ventral surface it has become nearly terminal, grows 
into two pouches which project into the cavity between the 
skin and gut on either side of the solid cord of cells, which 
is the persistent hind part of the original primitive sheath 
(fig. 13 and fig. 14, pd). These pouches are derived from 
cells homologous with those which have already given rise 
to mesoblast. The continuity of this posterior pair of pits 
with the anterior is kept up by the few cells (me") budded 
off in the middle of the primitive streak. The same growth 
which opened up the space between ectoblast and endoblast 
has separated the anterior and posterior mesoderm. The fact 
that in Phoronis the two ends of each mesodermic pouch are 
actually connected by an intermediate cord of cells depends on 
this formation of a primitive streak along the whole line of 
closure of the blastopore. 
Nephridia. 
The posterior pair of mesodermic diverticula open in the 
middle line to the exterior. The closure of this opening pro- 
ceeds in such a way that each pouch remains open to the 
exterior by a smallbore on either side of the middle line. I 
believe — though this fact is not established so certainly as those 
above concerning the mesoderm — that each pore persists as 
the opening of the nephridium of its own side. The nephridia 
appear coincidently with the final narrowing of the mesodermic 
pores, but I have yet no sections showing the cells in the 
neighbourhood of the pores taking on the form of the intra- 
cellularly perforate excretory cells. The formation of these 
excretory cells, which lie in a blood space of the splanchno- 
pleure and not in the body cavity, I have independently traced 
from the mesodermic cells of the posterior pouches. 
