138 BULLETIN OF THE 
its resemblance to the tentacular loops of Brachiolaria, and from its 
position on the water-tube connecting with the water pore, I have no 
hesitation in considering to be the first tentacular loop formed.” 
Balfour considers this structure an invagination of the external surface 
of the larva, an infolding which later is to form the ventral region of 
the Echinoderm. Metschnikoff* ascribes to A. Agassiz the discovery of 
an invagination of the outer skin of the pluteus to form the body of the 
future echinus. I have not been able to find in A. Agassiz’s works, 
quoted by Metschnikoff, that he has made such a “ discovery,” and cer- 
tainly he does not give to the lettering of his figures the same interpre- 
tation which Balfour does, when he says that the structure in question, ¢, 
is a tentacular loop. 
Pl. VII. fig. 3, represents the young of Echinarachnius formed on a 
_ pluteus of the same general form as that figured in Pl. VII. fig. 2. The 
left water-tube has here formed the “rosette” of five radial tubes, 
which are seen in profile in the figure. The whole body of the pluteus 
is not represented, but a portion of the edge of the stomach of the plu- 
teus is seen on the right-hand side of the figure. The figure is a 
representation from the dorsal side. 
The five radial bodies of the “rosette,” one of which is lettered, wt, 
are the water-tubes or the ambulacral divisions which, when seen from 
one side, would form a five-rayed water system derived from the left 
water vesicle. The system communicates with a tube which passes 
through the mesodermic layer of the plutean body, and opens externally 
by an opening, ma, which is supposed to be the madreporic opening. 
This opening is at first situated near the base or proximal end of the 
posterior rods. Its communication with the water-tubes has a retort 
form, especially in older stages. The retort-shaped vesicle was observed 
in a stage a little older, Pl. VII. fig. 4, than the last to be in direct 
communication with the rosette-shaped system of five ambulacral ves- 
sels. Each of the five radiating ambulacral tubes which form the 
rosette extends outward from a central region, which is in direct com- 
munication with the retort-shaped body. These radial ambulacral tubes 
are at first simple, without lateral branches, cceca-like folds or loops, f 
with transparent walls, the outlines of which can be plainly seen 
* Op. cit. p. 41. 
+ These five loops are supposed to be homologous with the “ odd ocular tenta- 
cles” of Arbacia and Strongylocentrotus. See A. Agassiz, Report on the “ Chal- 
lenger”’ Echinoidea, p. 8. They are supposed to be the same as the “ Fiihler,” F. 
in Asterina. (See Ludwig, op. cit., Pl. VII. figs. 96, 97.) 
