DR. E. B. WILSON ON THE DEVELOPMENT OF RENILLA. 765 
below the cesophagus in later stages. In the four days’ larva (fig. 146) the septa are 
clearly seen to be arranged in pairs on opposite sides of the dorso-ventral axis. The 
dorsal septa (d.s.) are very narrow and widely separated, and have no mesenterial 
filaments on their edges, the dorso-lateral (d./.s.) are much wider, and are thickened 
at their edges to form the mesenterial filaments; the ventro-lateral septa (v./.s.) are 
widest of all, and the ventral septa (v.s.) are about equal to the dorso-lateral. 
When the septa are sufficiently far advanced to be visible from the exterior upon 
rendering the larva transparent by reagents or by compression in the fresh state, they 
are found to have a remarkable and definite arrangement. This arrangement is 
apparent at a very early stage, and remains unchanged as far as the development can 
be followed. Hence it will be convenient to describe it from a somewhat older 
specimen (figs. 103, 104, four days). The dorsal septa (d.s.) extend backwards for 
about one-fourth the length of the body, where they are joined by the dorso-lateral 
septa (d.l.s.). From their point of union the peduncular septum (p.s,) extends back- 
wards to the aboral end of the body. The ventro-lateral septa (vl.s.) extend 
backwards some distance beyond the point of union of the above-mentioned septa, 
and then bend upwards to join the peduncular septum at the point w (fig. 104). In 
some cases it is difficult to trace the septum up to the peduncular septum, especially 
when the larva is fully expanded. In fact, I completely overlooked their connexion 
in my earlier paper, and described the septum as terminating freely below. In some 
specimens this appears to be actually the case, though it is difficult to make sure of it, 
but in every case the line of longitudinal muscles accompanying the septum (see p. 780) 
is continued up to the peduncular septum. The ventral septa (v.s.) are of nearly the 
same length as the dorsal, and in some specimens appear to terminate freely below. 
Jn most cases, however, careful examination during a half-contracted state of the larva 
shows that the lower ends of the septa bend towards one another and unite in the 
median ventral line. From their point of union a band of longitudinal muscles extends 
backwards in the median line of the body. In specimens where the septa themselves 
do not actually join, the lines of accompanying muscles bend towards one another and 
unite in the same way that those of the ventro-lateral septa join the peduncular 
septum. 
The arrangement of the septa shows, therefore, a very marked bilateral symmetry, 
the septa being disposed according to their width, length, and relations to each other, 
in pairs which are symmetrically placed with reference to the dorso-ventral plane. 
c. Formation of the peduncular septum, 
The peduncular septum has a quite different mode of origin from the radial septa, 
though it is continuous with the latter at their earliest appearance. It makes its 
appearance at about the fortieth hour at the posterior end of the body, sometimes, at 
any rate, before the stomodzeum or the radial septa are formed. A longitudinal 
section through this part of the body of a forty-hour larva is shown in fig. 150. The 
MDCCCLXXXIII. 5 F 
