DEVELOPMENT OF CUCUMARIA ECHINATA 207 
out ventrad, bringing together with it the lobes nos. 4 and 5, 
and on reaching the mid-ventral line it turns posteriorly to give 
rise to the mid-ventral radial canal (Pl. 9, fig. 21 4, mv). The 
other lobes except no. 1 are now differentiated into cylindrical 
tubes—primary tentacles (f,-,)—connected at the base by 
a rather narrow canal, which forms a horseshoe-shaped rudiment 
of the ring canal. The lobe no. 1 eventually gives rise both 
to the remaining one of the primary tentacles and to the free 
end of the dorsal limb of the open hydrocoele ring. It remains 
for a while as an inconspicuous outgrowth. 
Radial Canals.—The free end of the rudimentary mid- 
ventral radial canal then dilates laterally to form a rhombic 
vesicle in ventral view, and then takes on a cross shape (PI. 9, 
fig. 23 4, mv). The transverse branches thus formed are the 
primary pedicel canals (rpe, Ipc), and in correspondence with each 
of them a rudiment of the primary pair of pedicels (rp, lp) is formed. 
The four radial canals, other than the mid-ventral one, are 
formed comparatively late, especially the ventral pair (rd, Id, 
rv, lw). They arise at first as small knobs on the anterior margin 
of the ring canal, one in each interval of the primary tentacles. 
The knobs then bend outwards and immediately turn pos- 
teriorly, and they are remarkably thin as compared with the 
mid-ventral one. 
Kowalewsky (17, p. 4) first described in C. kirchs- 
bergii the first appearance of the mid-ventral radial canal, 
which soon divided into two, pushing the body-wall outwards 
to form a pair of pedicels. In C. planeci Selenka (45, 
p- 171) ascribed the development of the mid-ventral radial canal 
to too late a period, stating that it made its first appearance 
soon after the closure of the hydrocoele ring, and that four 
other radial canals and the Polian vesicle followed it. As to the 
fact that the four radial canals, other than the mid-ventral one, 
appear later than -the latter, all observers are unanimous. 
Ludwig (22, p.181) further noticed that among those four the 
ventral pair are shorter and narrower than the dorsal pair, the 
difference being observable throughout the life of the young ; 
they grow to be equal much later on. 
