486 HIROSHI OHSHIMA 
The post-oral rods grow up to four or five times the length 
of the body-rod, while the antero-lateral rod and the aboral 
spike reach more or less twice the length of the same (fig. 5). 
The post-oral rods seem in most cases to be devoid of 
fenestration in their proximal half or one-third, whereas the 
unfenestrated portion of the aboral spike is generally much 
shorter. In an extreme case in the latter the fenestration 
hegins close to the proximal end (fig. 6, ab), exactly as the 
feature seen by Krohn in an unidentified form (8, p. 210). 
The distal parts of these rods are fairly regularly serrated. 
The serration seems to begin roughly at the point where the 
fenestration also begins (fig. 5, po, ab). The posterior ends of 
both the body- and recurrent rods show towards the end of 
this stage signs of degeneration, being gradually absorbed. 
The dorsal arch makes its appearance near the end of this 
stage, on the mid-dorsal line at the level where the oesophagus 
opens to the stomach (fig. 5, da). The two arms of the star, 
which he symmetrically and are directed antero-laterally, 
increase rapidly in length, while the unpaired, posteriorly- 
directed arm remains very short, sometimes even obliterated. 
Krohn’s figure (2, Pl. vii, fig. 2) corresponds to the early 
four-armed stage. It is the dorsal view, in which the ventral 
horizontal rods and the body-rods are not shown, while the 
descending rods, which I take as the recurrent, are not 
coming to meet each other at the posterior ends. ‘The post- 
oral rods are shown as fenestrated on their distal three-fifths, 
while the aboral spike remains unfenestrated. Both these 
kinds of rods are, however, shown to have serrated edges 
along their whole length. 
The next, six-armed stage, is characterized by the appear- 
ance of the postero-dorsal arms. Previous to the appearance 
of these arms the supporting skeleton, which is called the 
postero-dorsal rod, is formed underneath each of them 
(fig. 6, pd). The rod develops in the manner similar to that 
of the other latticed rods, and as described and figured by 
Théel in Echinocyamus pusillus (11, p. 44, Pl. vi, 
fig. 88, y). The arms of the star lie in such a position that one 
