78 BULLETIN OF THE 
ceral segment lie just in front of the bases of the cheliceree, and are like- 
wise post-oral structures, as claimed by Balfour. 
The next stage in the process is represented in Pl. III. fig. 14, in 
which the tail-lobe is much narrower and more clearly circumscribed ; a 
considerable interval now separates it from the procephalic plate. Five 
pairs of dorsal (tergal) elements belonging to the abdominal segments 
are now visible; the four anterior pairs belong to the segments bearing 
provisional appendages, and a fifth, smaller pair, has been interpolated 
between these and the tail-lobe. The tail-lobe is apparently split in the 
median line into two bands that, in passing forwards, diverge rapidly. 
These are the two bands of ectoderm which, as before mentioned, join 
each other in the head and the tail-lobes. Between these divergent 
bands of ectoderm is to be seen a part of the yolk mass covered by only 
a thin layer of ectoderm. The legs have increased in length until they 
nearly meet in the median plane (Pl. IV. fig. 22). 
In the next stage (Pl. III. fig. 15) the dorsal region is much elon- 
gated owing to the retrogression of the tail-lobe, and the rudimentary 
terga extend much further dorsad. Up to this time the only dorsal ele- 
ments developed were the five pairs belonging to the abdominal somites, 
but during this stage the dorsal elements of the limb-bearing somites 
begin a more rapid growth. The dorsal elements of the somite bearing 
the fourth pair of legs grow much more rapidly than the others. 
In a dorsal view of a somewhat later stage (Pl. III. fig. 16) the tip of 
the tail is just visible at the posterior margin of the embryo, the dorsal 
region having increased proportionately in extent. The procephalic lobes 
are closing together in the median plane. The dorsal elements of the 
somites now nearly meet in the median line of the back. In the figure 
some of the provisional appendages (pr. app.) are visible along the sides 
of the body. 
In a slightly older embryo (Pl. III. fig. 17) the tail-lobe is no longer 
visible from above ; the cephalic lobes have become fused, and the dorsal 
elements of the somites have met in the median line. Along this line a 
narrow slightly elevated ridge indicates externally the position of the 
heart. 
The much reduced caudal lobe is to be seen from below (Pl. IV. 
fig. 21) and, diverging from it in two lines, the provisional appendages. 
Owing to the wide separation of the neural bands the legs of each pair 
are far apart. Between them a part of the yolk (not the whole, as stated 
by Barrois) protrudes, forming a sort of ventral yolk sack. The rapid 
appropriation of this store of yolk causes the disappearance of the sack ; 
